Home

NONMEM USERS GUIDE INTRODUCTION TO NONMEM 7.3.0

1. etc It should be pointed out that this example in which nmtemplate is used to create a random variable for substitution into ISAMPLE can easily be done in NM73 using the ISAMPEND and SELECT 3 options for EST METHOD CHAIN or CHAIN see 1 48 Method for creating several instances for a problem starting at different randomized initial positions EST METHOD CHAIN and CHAIN Records 1 63 Single Subject Analysis using Population with Unconstrained ETAs nm73 By default NONMEM performs single subject analysis by supposing that the data of the entire data file is from one subject implied by the lack of an ID item and lack of a SIGMA record but presence of a OMEGA record The help manual demonstrates another means by which one data file may contain data from all subjects to be separately analyzed using ID item as a parsing parameter over multiple single subject problems The RECS ID option is used for this purpose as given by the following example examples indestb ctl SPROB THEOPHYLLINE POPULATION DATA Analysis of Individuals Modification of CONTROL5 control steam SINPUT ID DOSE AMT TIME CP DV WT SDATA THEOPP RECS ID RECS ID Data set will be read until ID changes or end of file SSUBROUTINES ADVAN2 HETA 1 MEAN ABSORPTION RATE CONSTANT 1 HR HETA 2 MEAN ELIMINATION RATE CONSTANT 1 HR HETA 3 SLOPE OF CLEARANCE VS WEIGHT RELATIONSHIP LITERS HR KG CALING PARAMETER VOLUM
2. CONTROL MTOUCH 1 for manager to touch the worker directory to get up to date information WTOUCH 1 for worker to touch its directory MSLEEP milliseconds for manager to wait between writing its content files WSL to the remote worker directory iEP2 milliseconds for worker to wait between writing its content files I nm730 doc 153 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 to the worker directory 3 MTOUCH 1 WSLEEP 5 WTOUCH 0 MSLEEP 0 SIDRANGES USED IF PARSE TYPE 3 1 1 50 2 51 100 There is an additional record introduced here called CONTROL When working between computers on Linux with FPI some network file systems such as NFS on Unix may require that the manager touch the remote worker directory for that directory to show the up to date file information to the manager Also the process may need a period of waiting time before the signal file is created Hence the need for the SCONTROL statements After an estimation step is performed the worker processes exit For the next estimation step that follows if there is one the manager will reload the worker processes If you want worker processes to remain resident until all estimations and problems listed in the control stream file are completed then select TRANSFER TYPE 2 Running Parallel Processes in a Mixed Platform Environment Suppose the manage
3. ecce eee e eee eerte eene 149 Setting up FPL on Pli der 152 Running Parallel Processes in a Mixed Platform Environment eesssesssesssesssecesoossooseoo 154 Installing NEP TL Ome E er 154 Some Advanced Technics For Defining the PARAFILE for an MPI System 158 Special Considerations for MAC OS X e eee ecce eee eeee ee eee se eeeo se sese sete eese tese se enaoe 159 Mounting file systems on MAC OS X eee eere eerte ee eren eee eene eee ta sete ta sete ta sete ea seta seen 159 Enabling ssh with no password on MAC OS X eee eee esee ee ee eee ee ee een see en seta setas 160 Disabling Open MPI commands on MAC OS X ceres eee e esee sete eee ee sete eo seen osea 160 Installing MPICT2 on MAC OS X eeieeivescesskteue eet ea oe cupo eee ete Se eee euet ee ee Mee pe REPRE EP Qu MET Ein 160 1 54 Repeated Observation Records NM72 eeeeeeeeeeee 161 1 55 Stochastic Differential Equation Plug In NM72 163 1 56 Turning on First Derivative Assessments for EM Bayes Analysis NM72 166 1 57 Ignoring Non Impact Records During Estimation NMT3 167 1 58 table compare Utility Program NM72 eese 167 1 59 table to xml Utility Program
4. MAXIDS Largest total number of individual records subjects in a data set used in the run MAXDREC Largest number of data records in any one individual record in any one subject TOTDREC total number of data records lines in largest data set to be used NPROB Total number of problems in the control stream LVR Largest number of etas in any problem including those listed in PRIOR As of NM73 the values for MAXDREC and TOTDREC are assessed by NMTRAN and the user may take advantage of NMTRAN s evaluation by using the maxlim option to the nmfe73 script see below But NMTRAN may not always correctly assess these values Thus it is best if the user ascertains these values ahead of time by inspection of his largest data set among all of the problems to be used by the control stream file and the largest number of parameters to be used Then set the LIM values accordingly via the SIZES record nm730 doc 36 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 One can alternatively assess empirically whether file buffers are used by beginning the run allowing perhaps one iteration to transpire then from another command window do a directory search for FILE or WK for worker files in parallelization problems section 1 53 Parallel Computing NM72 If any of the FILExx do not have 0 size then they are being used Interrupt the analysis then increase the appropriate LIM value with the SIZES record delete the FILE in case
5. NOSIGMABOUNDTEST NOTHETABOUNDTEST NOTITLE NONINFETA NOPRIOR NSIG P lt P lt P lt X XS PX PX P lt X X nm730 doc 96 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Option Classical ITS NUMDER X X X X X X NUMERICAL X X B ji X T OACCEPT XX x OMITTED X X X OPTMAP X X ORDER X X gt lt gt lt X X X OSAMPLE MI OSAMPLE M2 PACCEPT PARAFILE POSTHOC PREDICTION pp X X pP sies P lt OS P lt P lt OS P lt P lt OS P lt P lt X PRINT PSAMPLE MI PSAMPLE M2 PSAMPLE M3 PSCALE MAX PSCALE MIN XXX RANMETHOD nSmP X X X X REPEAT REPEATI p REPEAT2 SEED X SIGL SIGLO gt lt gt lt X pA pA XXX SLOW P lt gt lt OS SORT STDOBJ X gt lt STIELTJES X ZERO X May be needed to suppress error messages from NMTRAN or NONMEM 1 36 When to use each method While there is some overlap in usage of the various EM methods some basic guidelines may be noted MC Importance Sampling EM IMP is most useful for sparse few data points per subject that is fewer data points than there are etas to be estimated for a given subject or rich data and complex PK PD problems with many parameters The SAEM method is mo
6. OMEGA BLOCK 2 0 04 p 0 01 f 0 04 p SIGMA 0 01 p EST METHOD ITS INTERACTION NITER 100 PRINT 1 NOABORT SIGL 8 FILE example5 ext CTYPE 3 SEST METHOD IMPMAP INTERACTION NITER 20 ISAMPLE 300 PRINT 1 NOABORT SIGL 8 EST METHOD IMP INTERACTION NITER 20 MAPITER 0 ISAMPLE 1000 PRINT 1 NOABORT SIGL 6 EST NBURN 500 NITER 500 METHOD SAEM INTERACTION PRINT 10 SIGL 6 ISAMPLE 2 EST METHOD IMP INTERACTION NITER 5 ISAMPLE 1000 PRINT 1 NOABORT SIGL 6 EONLY 1 SEST METHOD BAYES INTERACTION NBURN 2000 NITER 5000 PRINT 10 FILE example5 txt SIGL 8 EST MAXEVAL 9999 NSIG 2 SIGL 8 PRINT 10 FILE example5 ext METHOD CONDITIONAL INTERACTION NOABORT COV MATRIX R nm730 doc 195 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 70 Example 6 Receptor Mediated Clearance model with Dynamic Change in Receptors Model Desc Receptor Mediated Clearance model with Dynamic Change in Receptors Project Name nm7examples Project ID NO PROJECT DESCRIPTION PROB RUN example6 from r2compl SINPUT C SET ID JID TIME DV CONC DOSE AMT RATE EVID MDV CMT DATA example6 csv IGNORE C The new numerical integration solver is used although ADVAN 9 is also efficient for this problem SSUBROUTINES ADVAN13 TRANS1 TOL 4 MODEL NCOMPARTMENTS 3 PRIOR NWPRI NTHETA 8 NETA 8 NTHP 0 NETP 8 NPEXP 1 PK MU 1 THETA 1 MU 2 THETA 2 MU 3 THETA 3 MU 4 THETA 4 MU 5 THETA 5 MU 6 THETA 6 MU_7 THETA 7 MU_8 THETA 8 VC
7. END SUBROUTINE CONSTRAINT nm730 doc 204 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 74 Example 10 One Compartment First Order Absorption Pharmaokinetics with Categorical Data PROB F FLAGO4est2a ctl SINPUT C ID DOSE AMT TIME DV WT TYPE DATA example10 csv IGNORE SUBROUTINES ADVAN2 TRANS2 PRIOR NWPRI NTHETA 5 NETA 3 NTHP 0 NETP 3 PK CALLFL 1 MU_1 DLOG THETA 1 KA DEXP MU_1 ETA 1 MU_2 DLOG THETA 2 V DEXP MU_2 ETA 2 MU_3 DLOG THETA 3 CL DEXP MU_3 ETA 3 SC V 1000 S THETA 5 0 10 0 2 0 0 1 0 1 OMEGA BLOCK 3 0 5 0 01 0 5 0 01 0 01 0 5 prior information for Omegas OMEGA BLOCK 3 0 09 0 0 0 09 0 0 0 0 0 09 STHETA 3 FIX Because THETA 4 and THETA 5 have no inter subject variability associated with them the algorithm must use a more computationally expensive gradient evaluation for these two parameters SIGMA 0 1 SERROR Put a limit on this as it will be exponentiated to avoid floating overflow EXPP THETA 4 F THETA 5 IF EXPP GT 30 0 EXPP 30 0 IF TYPE EQ 0 THEN PK model F_FLAG 0 Y F F ERR 1 a prediction ELSE Categorical model F_FLAG 1 A DEXP EXPP B 1 A Y DV A B 1 DV B a likelihood ENDIF SEST METHOD ITS INTER LAP NITER 1000 PRINT 5 SIGL 6 NSIG 2 NOABORT NOPRIOR 1 CTYPE 3 CITER 10 CALPHA 0 05 FILE examplel10 ext Because of categorical data which can make conditional density highly non normal
8. sosehvnsdvsesevi 77 PACOEPT U 1 test touasges evoisternudatocstnensesaguaasnastxosdiesanbabnes tevsetonaelaesasue dleanedisuesvevsedonveisias 77 BPSCALE MINZQOLU NIVIETS tests Da WERE CO AER bood OR EU NOR PEs ERG 78 PSCALE M X 1000 NM73 irsini ters earen tara tope eH eh dva 78 OSAMPLE_M1 1 defaults listed eee ee eee eee ee eee teen ete tn state tasa tta tasoaeo 78 OSA MPL MAS T 5 iecit oi psit di odes E ASET 78 OACCEPESQS ii dieto M DI IA EEDMAESEN E ATE EOD ep t eoe Sese oS S DEG EDU VPr FA DL 78 INO PRIOR COLLAR preme Dee 78 1 29 A Note on Setting up Prior Information eese 78 1 30 Monte Carlo Direct Sampling NM72 eeeeeeeeeeeereneeeneee 83 EST METHOD DIRECT INTERACTION ISAMPLE 10000 NITER 50 83 1 31 Some General Options and Notes Regarding EM and Monte Carlo Methods 83 AUTOZO default GCNIMI7S sss suseiesssvcovscesucas ts picniskacsosenessuycnsbsess dovesipsenstasbusuecusvvesntisindevisaplesstesnne 83 1 32 MU Referencing usc als Deed bre ia ibtd adeb id Obi detbt s Cbrobdbride dh utei nnmnnn nennen 85 MUM MMNNMD 5i ist eti ee toi Lye poe Fel uu esa so Ee rossas ror Ee Po Eres Oo Saa orse URP EUR AM ted 91 GRDSGNONNND eicere iU DaSUE HI MPs sto ese iia AU BOUND oio e RAUM CUR Cei Tere EUM sitos 92 GRD DDDDDDSSN reserse rrii rea ateei M 93 33 Terminallon testing sreo aiaa aeaee iK UC uud eaaa a G dE Qo
9. As with SIML options STRAT and STRATF are available for the NONP BOOTSTRAP record to provide stratified selections see STRAT NM73 in 1 21 Bootstrap Selecting a Random Method and Other Options for Simulation NM73 Three files are produced providing nonparametric information root npd Each row contains information about a support point The support point number the ID from which the support point was obtained as an EBE of that subject ID is 1 if this support point was randomly generated because NSUPP NSUPPE was greater than number of subjects The eta values of the support point are listed followed by the cumulative probability CUM associated with each eta followed by the joint density probability of that support point if default or MARGINALS was selected If ETAS was selected then instead of cumulative probabilities the support point eta vector that best fits that subject ETM is listed root npe The expected value etas and expected value eta covariances ETC are listed for each problem or sub problem Because only one line is written per problem or sub problem the column header is displayed unless EST NOLABEL 1 only once for the entire NONMEM run However each line contains information of table number problem number sub problem number super problem and iteration number root npi The individual probabilities are listed in this file The header line unless EST NOLABEL 1 is written only once at the beginning o
10. ISAMPEND n NM73 For SAEM if ISAMPEND is specified as an upper integer value usually 10 then NONMEM will perform a ISAMPLE preprocess to determine the best ISAMPLE value For the SAMPLE preprocessing the used entered ISAMPLE value must be at least 2 It will perform 200 iterations during the ISAMPLE preprocess and the last 50 iterations will be used to obtain average conditional variance OMEGA eta shrinkage for each subject The largest etashrinkage fraction 10 is the ISAMPLE for that subject Thus ISAMPLE 2 ISAMPEND 10 Will assess a best ISAMPLE for each subject The ISAMPLE will not be higher than 10 or lower than 1 ISCALE_MIN 1 0E 06 defaults for SAEM BAYES NM72 ISCALE_MAX 1 0E 06 NM72 In MCMC sampling the scale factor used to vary the size of the variance of the proposal density in order to meet the IACCEPT condition is by default bounded by ISCALE_MIN of 1 0E 06 and ISCALE_MAX 1 0E 06 This should left alone for MCMC sampling but on occasion there may be a reason to reduce the boundaries perhaps to ISCALE MIN O0 001 ISAMPLE_MAX 1000 After the SAEM estimation method remember to revert these parameters back to default operation on the next EST step ISCALE MIN 100 ISCALE MAX 100 The default operation is that NONMEM sets ISCALE MIN ISCALE MAX to 0 1 10 for importance sampling as described earlier and to 1 0E 06 1 0E 06 for MCMC sampling nm730 doc 72 of 210 NONMEM Users Guide Introduction to
11. NM73 for the COV respectively COVT 12 4 iz yo 4t The second and subsequent TABL items have added to their second line the SEED at column 29 ESAMPLE value starting at position 41 RANMETHOD NM72 at position 53 WRESCHOL NM73 at position 65 and the format for the table starting at position 68 TABL 1 5 3 0 5 02094 0 19 0 20 0 0 1 1 0 1 12344 300 3 1 1PE12 5 The value of the third integer at Position 17 was originally limited to ONEHEADER 1 NOHEADER but as of NM73 has been expanded to the following bits being set where bit 0 is the first bit ONEHEADER bit 0 NOHEADER bit 1 NOTITLE bit 2 NOLABEL bit 3 The additional statistical diagnostic items have indices as follows where LNP4 may be 2000 for medium sized setups and 4000 for large setup NPRED LNP4 95 NRES LNP4 96 NWRES LNP4 97 NIWRES LNP4 98 CPRED LNP4 99 CRES LNP4 100 CWRES LNP4 101 nm730 doc 209 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 CIWRES LNP4 102 PREDI LNP4 103 RESI LNP4 104 WRESI LNP4 105 IWRESI LNP4 106 CPREDI LNP4 107 CRESI LNP4 108 CWRESI LNP4 109 CIWRESI LNP4 110 EPRED LNP4 111 ERES LNP4 112 EWRES LNP4 113 EIWRES LNP4 114 NPDE LNP4 115 ECWRES LNP4 116 NPD LNP4 117 OBJI LNP4 118 nm730 doc 210 of 210
12. CTYPE CTYPE 4 pP P lt P lt OS P lt P lt P lt P lt OS DERCONT X P lt S PS PS OS PN DF DFS CHAIN only nm730 doc 95 of 210 DIRECT IMP IMPMAP SAEM BAYES NONMEM Users Guide Introduction to NONMEM 7 3 0 Option Classical ITS DIRECT IMP IMPMAP SAEM BAYES EONLY X gt lt gt lt ETABARCHECK ETADER ETASTYPE FILE FNLETA FORMAT DELIM SIIK lt P lt gt lt GRD pP lt P lt gt lt lt lt P lt S PS P lt S P lt PS OS pP PS PS OS Xx X px GRID gt lt HYBRID Stieltjes X IACCEPT INTERACTION X ISAMPEND p P lt gt lt ISAMPLE ISAMPLE MI ISAMPLE MIA ISAMPLE M2 ISAMPLE M3 ISCALE MAX ISCALE MIN LAPLACE LIKE X X XX XI KK X KK MAPINTER MAPITER PP X Pv MAXEVAL MCETA MSFO PX MUM gt lt gt lt gt lt P lt gt lt P lt gt lt NBURN NITER NSAMPLE NOABORT X NOCOV when last estimation n gt O gt lt gt lt P P lt gt lt P lt gt lt P lt gt lt P lt P lt P lt S OS P lt S PS S OS NOHABORT NOLABEL X X P lt X X XX XX X X NOOMEGABOUNDTEST
13. DL_N where Soa and Sy are the old and new shrinkage values respectively E is the Etabar value Naia is the total number of subjects N54 is the number of subjects contributing information to that eta and Qis the omega variance diagonal element pertaining to that eta Alternatively set ETASTYPE 1 for NM73 in the SEST record and this will average shrinkage information only among individuals that provided a non zero derivative of their data likelihood with respect to that eta and will not include subjects with a non influential eta that is in which the derivative of the data likelihood is zero Furthermore you may specify eta i of particular subjects to be excluded by setting a reserved variable ETASXI i to 1 in PK or PRED or specify eta i of certain subjects to be included by setting ETASXI 1 22 ETASXI stands for eta shrinkage exclude include IF ID 3 ETASXI 1 1 IF ID 23 ETASXI 3 2 In nm73 additional shrinkage information called EBVshrink is the ETA shrinkage based on the average empirical Bayes variance the etc j j or phc j j listed in the phi or phm table ETAshrinkage 10096 1 4 1 etc j D OmegaG j ETAshrinkage 100 1 NT phc j J Omega j Where etcave j j is average etc j j among included subjects and phcaye j j is average phc j j among included subjects for eta j or phi j The results reported here refer to average eta shrinkage See the
14. ED50 HILL DOSE HILL X IPRED EPS 1 STHETA 4 1 1 Emax STHETA 6 9 2 ED50 STHETA 3 0 0 001 3 0 3 Hill STHETA 2 3 4 EO SOMEGA BLOCK 2 0 1 SOMEGA SOMEGA SIGMA 1 SEST METHOD CHAIN ISAMPLE 1 ISAMPEND 30 NSAMPLE 30 FILE anneal2 chn SESTIMATION METH SAEM INTER NBURN 4000 NITER 200 ISAMPLE 5 IACCEPT 0 3 CINTERVAL 25 CTYPE 3 NOABORT PRINT 100 SESTIMATION METH IMP INTER PRINT 1 NITER 0 ISAMPLE 10000 EONLY 1 MAPITER 0 COV MATRIX R UNCONDITIONAL nm730 doc 104 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Notice that the range of Monte Carlo search for the Hill coefficient is from 3 to 3 the specified lower and upper bound values note that theta 3 1s actually the log of the Hill coefficient See 1 48 Method for creating several instances for a problem starting at different randomized initial positions SEST METHOD CHAIN and CHAIN Records 1 41 COV Additional Parameters and Behavior Example syntax SCOV UNCONDITIONAL TOL 10 SIGL 10 SIGLO 11 NOFCOV ATOL 6 RESUME If COV is specified then for IMP IMPMAP and ITS methods standard error information will be supplied for every EST statement Standard error information for the classical methods METHOD 0 METHOD 1 will be given only if they are the last estimation method and only if NOFCOV is not specified If UNCONDITIONAL is specified then for the IMP and IMPMAP EM methods if the R information matrix is not positive definite
15. Initial OMEGA block 2 for sub population 2 OMEGA BLOCK 2 05 p 01 f 06 p nm730 doc 191 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Degrees of Freedom defined for Priors One for each OMEGA block defining each sub population STHETA 2 FIX 2 FIX Prior OMEGA block 1 Note that because the estimated OMEGA is separated into blocks so their priors should have the same block design OMEGA BLOCK 2 0 05 FIX 0 0 0 05 Prior OMEGA block 2 SOMEGA BLOCK 2 0 05 FIX 0 0 0 05 SIGMA 0 01 p EST METHOD ITS INTERACTION NITER 20 PRINT 1 NOABORT SIGL 8 FILE example3 ext CTYPE 3 CITER 10 CALPHA 0 05 NOPRIOR 1 EST NBURN 500 NITER 500 METHOD SAEM INTERACTION PRINT 10 SIGL 6 ISAMPLE 2 EST METHOD IMP INTERACTION NITER 5 MAPITER 0 ISAMPLE 1000 PRINT 1 NOABORT SIGL 6 EONLY 1 EST METHOD BAYES INTERACTION NBURN 2000 NITER 1000 PRINT 10 FILE example3 txt SIGL 8 NOPRIOR 0 SEST MAXEVAL 9999 NSIG 3 SIGL 10 PRINT 1 FILE example3 ext METHOD CONDITIONAL INTERACTION NOABORT NOPRIOR 1 COV MATRIX R UNCONDITIONAL nm730 doc 192 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 68 Example 4 Population Mixture Problem in 1 Compartment model with rate constant parameter and its inter subject variances modeled as coming from two sub populations Model Desc Population Mixture Problem in 1 Compartment model with rate constant parameter and its inter subject variances modeled as coming fro
16. Note on the t Distribution Sampling Density DF gt 0 and its Use With Sobol Method RANMETHOD S When using the t distribution sampling density DF gt 0 by default the algorithm creates a composite random vector from n independent univariate t distributed samples This is called the U algorithm and the most efficient use of the U type t distribution is when DF 1 2 4 5 8 or 10 These algorithms were designed to work well with the Sobol method s ability to reduce Monte Carlo noise 1 26 Monte Carlo Importance Sampling EM Assisted by Mode a Posteriori MAP estimation Sometimes for highly dimensioned PK PD problems with very rich data the importance sampling method does not advance the objective function well or even diverges For this the IMPMAP method may be used At each iteration conditional modes and conditional first order variances are evaluated as in the ITS or FOCE method not just on the first iteration as is done with IMP method These are then used as parameters to the multivariate normal proposal density for the Monte Carlo importance sampling step This method is implemented by EST METHOD IMPMAP INTERACTION This is equivalent to EST METHOD IMP INTERACTION MAPITER 1 MAPINTER 1 1 27 Stochastic Approximation Expectation Maximization SAEM Method As in importance sampling random samples are generated from normal distribution proposal densities However instead of always centered at the mean or mode of the posterior den
17. REWIND STHETA 2 0 2 0 4 0 4 0 Initial Thetas SOMEGA BLOCK 4 Inital Parameters for OM 0 4 0 01 0 4 0 01 0 01 0 4 0 01 0 01 0 01 0 4 SSIGMA 0 1 Ez Q pi Second problem selects sample ISAMPLE 2 for initial settings from file wexamplell chn Won t recreate the file as NSAMPLE 0 SEST METHOD CHAIN FILE wexamplell chn NSAMPLE 0 ISAMPLE 2 SEST METHOD COND INTERACTION MAXEVAL 9999 NSIG 2 SIGL 10 PRINT 5 NOABORT etcetera for samples 3 4 and 5 executed as problems 3 4 and 5 In the above example the five estimations are performed in sequence To perform these in parallel in a multi processor or multi computer environment a pre processing program could set up and execute a control stream file which would have as one of the commands SEST METHOD CHAIN FIL Ea examplel chn NSAMPLE 5 ISAMPLE 0 DF 20 A copy of this control stream file could be made and the pre processing program could make five new child control stream files with the NSAMPLE this time set to O so that it does not create a new chain file but uses the already existing one and ISAMPLE entries modified in the following five ways each differing by only the ISAMPLE number First control stream file SEST METHOD CHAIN FILE examplel chn NSAMPL second control stream file SEST METHOD CHAIN FILE examplel chn NSAMPL third control stream file
18. RFORMAT F8 0 37 1PE13 6 24 0PF7 2 1 14 SUBROUTINES New Differential Equation Solving Method As of NM7 A differential equation solver has been introduced called LSODA and is accessed using ADVAN 13 or ADVAN13 This routine is useful for stiff and non stiff equations This is similar to the LSODI routine used by ADVANO except that ADVANIS can at times execute more quickly than ADVAN9 The ADVAN 13 differential equation solver has been shown to solve problems more quickly with the new estimation methods whereas for classical NONMEM methods selecting ADVAN 6 or 9 may still be of greater advantage Example SSUBROUTINES ADVAN13 TRANS1 TOL 5 Where TOL is the number of digits accuracy desired to integrate the differential equations accuracy to within 107 The code to the differential equation solver is found in source LSODA f90 On occasion coded errors will be displayed if the algorithm is having trouble integrating the equations These errors may usually be ignored unless the error shows up frequently and ultimately results in failure for the problem to complete Typically the remedy is to increase or decrease TOL but for those who desire to understand what the error codes mean there are well documented comments on these at the beginning of LSODA f90 They are printed here for convenience ISTATE An index used for input and output to specify the the state of the calculation On input the values of i
19. RFORMAT NONE Here is an example of TABLE statements designated in a control stream file STABLE ID TIME PRED RES WRES CPRED CWRES EPRED ERES EWRES NOAPPEND ONEHEADER FILE tabstuff TAB NOPRINT FORMAT 1PE15 8 STABLE ID CL V1 Q V2 FIRSTONLY NOAPPEND NOPRINT FILE tabstuff PAR LFORMAT 4X A4 4 4X A8 RFORMAT F8 0 RFORMAT 4 1PE12 5 STABLE ID ETA1 ETA2 ETA3 ETA4 FIRSTONLY NOAPPEND NOPRINT FILE tabstuff ETA FORMAT F12 4 LFORMAT NONE RFORMAT NONE There is no NMTRAN error checking on the RFORMAT and LFORMAT records so the user must engage in trial and error to obtain a satisfactory table output you should set MAXEVAL 0 or MAXEVAL 1 for the SEST step to do a quick check so you don t spend hours on estimation only to find the RFORMAT LFORMAT were not appropriate nm730 doc 5 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 A word of caution The FORMAT descriptor 1P which means move the decimal point to the left by 1 will be in effect for all remaining FORMAT components For example in RFORMAT F8 0 37 1PE13 6 24 F7 2 the F field format that follows an E field format in which 1P was used will also have the decimal placed to the left and a 1 00 would appear as a 10 00 To prevent this from occurring revert to no decimal shift with OP
20. SEST METHOD CHAIN FILE examplel chn NSAMPLE fourth control stream file SEST METHOD CHAIN FILE examplel chn NSAMPLE fifth control stream file SEST METHOD CHAIN FILE examplel chn NSAMPL rj ll Lr ISAMPLE 1 DF 20 ll o tj CI ll e ISAMPLE 2 DF 20 rj ISAMPLE 3 DF 20 Tj ll e Tj ISAMPLE 4 DF 20 Tj ll o Tj Ea ll o ISAMPLE 5 DF 20 Each control stream file points to a different ISAMPLE position in the chn file so each would use these as the respective initial positions Each of these child control stream files could be loaded on to a job queue as separate processes If the user is running a multi core computer this would be quite straight forward nm730 doc 127 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 An existing chain file could actually be a raw output file from a previous analysis with a list of iterations In the following example SEST METHOD CHAIN FILE examplel previous txt NSAMPLE 0 ISAMPLE 1000000000 could pick up the final result of the previous analysis since ISAMPLE points to the iteration number and 1000000000 is the iteration number for the final estimate Thus the CHAIN method in this usage is really just an input command to bring in values from a raw output type file format Of
21. Select a t distribution with 4 degrees of freedom for the importance sampling proposal density SEST METHOD IMP INTER LAP NITER 1000 PRINT 1 ISAMPLE 300 DF 4 IACCEPT 1 0 SEST METHOD IMP EONLY 1 NITER 5 ISAMPLE 1000 PRINT 1 DF 4 IACCEPT 1 0 MAPITER 0 SEST METHOD SAEM EONLY 0 INTER LAP NBURN 2000 NITER 1000 PRINT 50 DF 0 IACCEPT 0 4 SEST METHOD IMP EONLY 1 NITER 5 ISAMPLE 1000 PRINT 1 DF 4 IACCEPT 1 0 MAPITER 0 For this example because thetas 1 3 are not linearly modeled in MU and theta 4 5 are not MU modeled all theta parameters are Metropolis Hastings sampled by the program But see examplel01 in the examples directory where Thetas 1 3 are linear modeled in MU and by default the program selects Gibbs sampling for them There is a 40 speed nm730 doc 205 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 improvement in doing so SEST METHOD BAYES NBURN 3000 NSAMPLE 3000 PRINT 100 FILE example10 txt DF 0 IACCEPT 0 4 NOPRIOR 0 SEST METHOD COND LAP INTER MAXEVAL 9999 PRINT 1 FILE example10 ext NOPRIOR 1 COV UNCONDITIONAL PRINT E MATRIX R SIGL 10 STABLE ID DOSE WT TIME TYPE DV A NOPRINT FILE examplelO tab nm730 doc 206 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 75 Description of FCON file The format of the FCON file produced by NMTRAN has been modified to incorporate the new features The new or modified items are as follows The LABL item contains a comma delimited list of labels beginning at position
22. V EXP LTVV ETA 3 MU 3 LTVV MU 5 LTVCL An incorrect usage of MU modeling would be MU 1 LOG THETA 1 MU_2 LOG THETA 2 MU 3 LOG THETA 3 CL EXP MU_1 ETA 2 V EXP MU_2 MU_3 ETA 1 In the above example MU 1 is used as an arithmetic mean to ETA 2 and a composite MU 2 and MU_3 are the arithmetic means to ETA 1 which would not be correct The association of MU_x ETA x must be strictly adhered to Once one or more thetas are modeled to a MU the theta may not show up in any subsequent lines of code That is the only usage of that theta may be in its connection with MU For example if CL EXP THETA 5 ETA 2 So that it can be rephrased as MU 2 THETA 5 CL EXP MU_2 ETA 2 But later suppose THETA 5 is used without its association with ETA 2 CLZ THETA 5 2 Then THETA 5 cannot be MU modeled because it shows up as associated with ETA 2 in one context but as a fixed effect without association with ETA 2 elsewhere However if MU 2 THETA 5 CL EXP MU 2 ETA 2 CLZ CL 2 Then this is legitimate as the individual parameter CL retains the association of THETA 5 with ETA when used to define CLZ That is THETA 5 and ETA 2 may not be used separately in any other part of the model except indirectly through CL in which their association is retained Suppose you have CL THETA 5 THETA 5 ETA 2 One should see this as CL THETA 5 1 ETA 2 So the way to MU model
23. is misspelled 4 SAEM terminates on some problems Cause is access violation when CONSTRAIN is called Work around for earlier versions of NONMEM is to set CONSTRAIN 0 Or set MAXOMEG using SIZES such that they are at least NEPS 1 NEPS 2 5 When defining compartments in MODEL NMTRAN does not always terminate DATA CMOD code lines properly with respect to continuation markers resulting in a failed nm730 doc 14 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 compilation of FSUBS Work around is to have more than an integer multiple of 6 compartments named for example if you have 24 compartments define a 251 compartment 6 When CHAIN record is used ISAMPLE may not be less than 1 Work around for earlier versions of NONMEM is to change the index number iteration number for a raw output file of a previous analysis of the desired record in the file to a positive number 7 When a simulation is desired using the results of a previous estimation using MSFI NONMEM sometimes prevents its use because of a flag indicating it was not properly estimated Work around for earlier versions of NONMEM use the record CHAIN FILE file ext ISAMPLE xxxx where file ext is the name of the raw output file of the previous analysis and xxxx is the iteration number typically the last iteration 8 During an estimation with FO or FOCE and the last subject in the data set has non influential etas for example with interoccasion variability
24. nm730 doc 185 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 65 Example 1 Two compartment Model Using ADVAN3 TRANSA Model Desc Two compartment Model Using ADVAN3 TRANS4 Project Name nm7examples Project ID NO PROJECT DESCRIPTION PROB RUN Example 1 from samp51 SINPUT C SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT CLX V1X QX V2X SDIX SDSX DATA examplel csv IGNORE C S SUBROUTINES ADVAN3 TRANS4 NTHETA number of Thetas to be estimated NETA number of Etas to be estimated and to be described by NETAxNETA OMEGA matrix NTHP number of thetas which have a prior NETP number of Omegas with prior PPrior information is important for MCMC Bayesian analysis not necessary for maximization methods PRIOR NWPRI NTHETA 4 NETA 4 NTHP 4 NETP 4 PK The thetas are MU modeled Best that there is a linear relationship between THETAs and Mus The linear MU modeling of THETAS allows them to be efficiently Gibbs sampled MU 1 THETA 1 MU_2 THETA 2 MU_3 THETA 3 MU_4 THETA 4 CL DEXP MU_1 ETA 1 V1 DEXP MU_2 ETA 2 Q DEXP MU_3 ETA 3 V2 DEXP MU_4 ETA 4 S1 V1 ERROR Y F F EPS 1 The Thetas are to list in order the following NTHETA of initial thetas NTHP of Priors to THETAS Degrees of freedom to each OMEGA block Prior Initial values of THETA NTHETA of them STHETA 0 001 2 0 LN CL 0 001 2 0 LN V1 0 001 2 0 LN Q 0 001 2
25. the population parameters CTYPE 3 and for less then 2 significant digits change NSIG Prior information is not necessary for ITS so NOPRIOR 1 The intermediate and final results of the ITS method will be recoded in row column format in examplel ext SEST METHOD ITS INTERACTION FILE examplel ext NITER 500 PRINT 5 NOABORT SIGL 4 CTYPE 3 CITER 10 CALPHA 0 05 NOPRIOR 1 NSIG 2 The results of ITS are used as the initial values for the SAEM method A maximum of 3000 Stochastic iterations NBURN is requested but may end early if statistical test determines that variations in all parameters is stationary note that any option settings from the previous EST carries over to the next SEST statement within a PROB The SAEM is a Monte Carlo process SO setting the SEED assures repeatability of results Each iteration obtains only 2 Monte Carlo samples ISAMPLE so they are very fast But many iterations are needed so PRINT only every 100th iteration After the stochastic phase 500 accumulation iterations will be Performed NITER to obtain good parameters estimates with little stochastic noise As a new FILE has not been given the SAEM results will append to examplel ext SEST METHOD SAEM INTERACTION NBURN 3000 NITER 500 PRINT 100 SEED 1556678 ISAMPLE 2 After the SAEM method obtain good estimates of the marginal density objective function along with good estimates of the standard errors This is best done with
26. 2 SDE CADD WILL EVALUATE THE TRUE COEFFICIENTS WS TO THE STOCHASTIC COMPONENTS H n general if you have nmcmt observation compartments then first ncmt EPS will pertain to measurement error and the second ncmt set of EPS will pertain to stochastic errors This means you cannot have L2 type correlations and proptadditive should be packaged into a single EPS For two obervations you may have F CMT 1 THEN H PRED A 1 V f W SQRT THETA 5 THETA 5 IPED IPRED THETA 6 THETA 6 Y IPRED W EPS 1 WS EPS 3 2 ENDIF F CMT 2 THEN H PRED A 2 V nm730 doc 164 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 W SQRT THETA 7 kTHETA 7 KIPED IPRED THETA 8 THETA 8 Y IPRED W EPS 2 WS EPS 4 ENDIF Number of compartments 1 number of base model derivative equations 1 LAST CALL SDE CADD A HH TIME DV CMT 1 0D 00 1 0D 00 SDE STHETA 0 2 3 1 CL STHETA 0 3 5 2 VD STHETA 0 2 4 SIGMA STHETA 0 1 SGW1 SOMEGA 0 1 1 CL SOMEGA 0 01 2 VD SIGMA 1 FIX 1 FIX PK SEST METHOD ITS INTERACTION LAPLACE NUMERICAL SLOW NOABORT PRINT 1 CTYPE 3 SIGL 5 SEST METHOD IMP INTERACTION NOABORT SIGL 5 PRINT 1 IACCEPT 1 0 CTYPE 3 SEST MAXEVAL 9999 METHOD 1 LAPLACE INTER NOABORT NUMERICAL SLOW NSIG 3 PRINT 1 MSFO sde9 msf SIGL 9 SCOV MATRIX R UNCONDITIONAL STABLE ID TIME FLAG AMT CMT IPRED IRES IWRES ONEHEADER NOPRINT FILE sde9 fit Thi
27. During this time the advance of the parameters may be monitored by observing the results in file specified by the FILE parameter described later in the Format of Output Files section and the advance of the objective function SAEMOBJ at the console may be monitored When all parameters or the SAEMOBJ do not appear to drift in a specific direction but appear to bounce around in a stationary region then it has sufficiently burned in A termination test is available described later that will give a statistical assessment of the stationarity of objective function and parameters The objective function SAEMOBJ that is displayed during SAEM analysis is not valid for assessing minimization or for hypothesis testing It is highly stochastic and does not represent a marginal likelihood that is integrated over all possible eta but rather is the likelihood for a given set of etas NSAMPLE NITER 1000 Sets maximum number of iterations in which to perform the non stochastic accumulation phase default 1000 ISAMPLE 2 defaults listed ISAMPLE M1 2 ISAMPLE M1A 0 NM72 ISAMPLE M2 2 ISAMPLE M3 2 IACCEPT 0 4 These are options for the MCMC Bayesian Metropolis Hastings algorithm for individual parameters ETAS used by the SAEM and BAYES methods For each ISAMPLE SAEM performs JSAMPLE_M1 mode 1 iterations using the population means and variances as proposal density followed by ISAMPLE MIA mode 1A iterations testing model parameters
28. If used the SIGLO option is the precision to which the individual etas are estimated The SIGL level set by the user continues to be the precision or delta setting for the finite difference algorithms in the higher level estimation process for THETAS OMEGAS and SIGMAS By default if SIGLO is not specified then SIGLO is set to the same value as SIGL and everything is evaluated in accordance with the previous paragraph Should SIGLO be used the recommended setting would be nm730 doc 57 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SIGLO TOL SIGL lt SIGLO NSIG lt SIGL 3 1 18 Alternative convergence criterion for FO FOCE Laplace NM72 Sometimes many iterations will occur with very little change in the objective function even with SIGL TOL adjustment This may occur because a parameter may oscillate at the 2 significant digit for example and NSIG was set to 3 The parameter may never settle down to a value that fluctuates at less than NSIG significant digits if its contribution to the objective function is very small Thus a minimum objective function is achieved but NONMEM s traditional convergence test based on all parameters changing by less then NSIG significant digits is never satisfied An alternative convergence test is to set CTYPE 4 in the SEST statement NONMEM will then additionally test if the objective function has not changed by more then NSIG digits beyond the decimal point over 10 iterations If t
29. MU_3 ETA 3 THETA SID_V or the last SID level NSID use the negative sum of the thetas of the other SID levels so that the sum of all thetas is 0 that is the super nested average theta is 0 K DEXP MU_1 ETA 1 THSUM_KA CL DEXP MU_2 ETA 2 THSUM_ CL V DEXP MU 3 ETA 3 THSUM V E lo rcc Il ENDIF S2 V SERROR IPRE F IF TYPE 0 Y IPRE IPRE EPS 1 IF TYPE 1 AND SID lt NSID Y THETA SID_KA EPS 2 The fitting of the pseudo data TYPE gt 0 IF TYPE 1 AND SID NSID Y THSUM_KA EPS 2 constrains the SID level thetas to be IF TYPE 2 AND SID lt NSID Y THETA SID CL EPS 3 constrained and modeled using extra IF TYPE 2 AND SID NSID Y THSUM_CL EPS 3 Sigma variances 2 4 IF TYPE 3 AND SID lt NSID Y THETA SID_V EPS 4 IF TYPE 3 AND SID NSID Y THSUM_V EPS 4 STHETA 0 2 4 2 0 1 x15 0 0 FIXED 0 1 x15 0 0 FIXED 0 1 x15 0 0 FIXED SOMEGA BLOCK 3 VALUES 0 1 0 001 SIGMA 0 1 P SSIGMA BLOCK 3 VALUES 0 3 0 001 This is the inter SID variance SEST METHOD 1 INTERACTION PRINT 1 NSIG 2 SIGL 10 FNLETA 0 NOHABORT NONINFETA 1 MCETA 20 SCOV MATRIX R UNCONDITIONAL SIGL 10 Notice the use of variable replacement mapping SABBR REPLACE short hand entries for initial thetas omegas and sigmas and that the sum of the thetas to the SID data item are fixed to nm730 doc 109 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 0 by constraining th
30. PK include nonmem reserved general Request extra information for Bayesian analysis An extra call will then be made tor accepted samples BAYES EXTRA REQUEST 1 MU 1 THETA 1 MU 2 THETA 2 MU_3 THETA 3 MU_4 THETA 4 CL DEXP MU_1 ETA 1 V1 DEXP MU_2 ETA 2 Q DEXP MU_3 ETA 3 V2 DEXP MU_4 ETA 4 S1 V1 When Bayes extra 1 then this particular set of individual parameters were accepted So you may record them if you wish IF BAYES EXTRA 1 AND ITER_REPORT gt 0 AND TIME 0 0 THEN WRITE 50 112 1X F14 0 5 1X 1PG12 5 ITER REPORT ID CL V1 Q V2 OBJI NIREC 1 ENDIF SERROR include nonmem reserved general Y F F EPS 1 IF BAYES EXTRA 1 AND ITER_REPORT gt 0 THEN WRITE 51 I12 1X F14 0 2 1X 1PG12 5 ITER REPORT ID TIME F ENDIF Initial values of THETA STHETA 0 001 2 0 LN CL 0 001 2 0 LN V1 0 001 2 0 ILN Q 0 001 2 0 LN V2 INITIAL values of OMEGA SOMEGA BLOCK 4 0 15 P 0 01 F nm730 doc 199 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 0 15 P 0 01 F 0 01 F 0 15 P 0 01 F 0 01 F 0 01 F 0 15 P Initial value of SIGMA SSIGMA 0 6 P THETA 2 0 FIX 2 0 FIX 2 0 FIX 2 0 FIX OMEGA BLOCK 4 10000 FIX 0 00 10000 0 00 0 00 10000 0 00 0 00 0 0 10000 Prior information to the OMEGAS SOMEGA BLOCK 4 0 2 FIX 0 0 0 2 0 0 0 0 0 2 0
31. This is the simplified bootstrap technique described in 8 To provide a series of simplified bootstrap analyses as an example nm730 doc 63 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SIML 12345 SUBP 100 SEST METHOD COND INTERACTION MAXEVAL 9999 NSIG 3 SIGL 10 PRINT 5 NOABORT SNONP BOOTSTRAP EXPAND In the above example BOOTSTRAP option is given in NONP along with the SIML statement without a BOOTSTRAP option On the first sub problem NONMEM will pass the original data to the estimation step EST to obtain final THETAS OMEGAS and SIGMAS with EBE s adjusted for expansion EXPAND followed by a nonparametric density analysis on the original data set On the second sub problem the estimation step is skipped but the final THETAS OMEGAS SIGMAS and EBE s from the first analysis are retained and a nonparametric density analysis is performed on a bootstrap version of the original data set For a full bootstrap analysis method as described in 8 SIML 12345 SUBP 100 BOOTSTRAP 1 SEST METHOD COND INTERACTION MAXEVAL 9999 NSIG 2 PRINT 5 NOHABORT SNONP EXPAND NSUPPE 50 In the above example 100 bootstrap analyses are performed The SIML provides a bootstrap version of the original data set for estimation by EST this is followed by EBE assessment on the original data set followed by nonparametric density assessment on the bootstrap data set STRAT STRATF NM73
32. and writes 6 digits to the right of the decimal for the II data item See Help guide for more details Times may be optionally encoded as hh mm ss instead of just hh mm For example 8 45 20 will be acceptable and incorporates the seconds values The ANNEAL record provides a means of SAEM simulated annealing to provide global search techniques for thetas that do not have Omegas associated with them See 1 40 ANNEAL to facilitate EM search methods NM73 for this additional annealing technique Population weighted residual diagnostic values can be calculated for normally distributed data even though there are also non normally distributed data values in the same subject See the MDVRES option in 1 13 TABLE Additional Statistical Diagnostics Associated Parameters and Output Format nm730 doc 11 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 When TABLE values exceed 0 3E 39 a warning is issued but the table is still produced A utility program to fill in extra records with small time increments to provide smooth plots This utility program can also fill in by various interpolation techniques missing covariate values for original records Also if an MDV is set to a value greater than or equal to 100 it is converted to that value minus 100 upon input but will also not be used at all during estimation only for table outputting This option allows you to use a data file that was enhanced with extra records for both estimat
33. if the last subject had no data during the last inter occasion the eta for that last inter occasion is non influential the estimation may become inefficient due to incorrect gradient assessments This has been corrected for some types of problems but this may still persist in other problems which may be remedied with the SLOW option For earlier versions of NONMEM another work around when possible is to reorder the subjects so that the last subject does not have one or more non influential ETA s 9 When only thetas are in a problem and there are single subject data then standard errors are printed out but covariance inverse covariance and correlation matrices are reported as 0 Work around for earlier versions of NONMEM If possible pose the problem as multi subject insert one eta as OMEGA 0 0 FIXED 10 When using DOWHILE DATA in abbreviated NMTRAN code there should be no comment on that line such as DOWHILE DATA start of dowhile 11 In abbreviated code recursion code and INFN DOWHILE DATA cannot both be present in the same control stream The error message is MUST BE DO WHILE CONDITION ENDDO Workarounds for earlier versions of NONMEM 1 avoid unnecessary recursive variables by defining them as COM 1 COM 2 etc 2 use MSF to put the INFN block in another problem 12 With large numbers of thetas and or omegas the xml file may incorrectly print out the various variance matrices of estimates covariance correla
34. obtained from the previous iteration is used as a proposal density Population parameters thetas sigmas and omegas are then updated from subjects conditional mean parameters gradients and their variances by single iteration maximization steps that are very stable and improve the objective function The population parameters converge towards the minimum of the objective function which is an accurate marginal density based likelihood exact likelihood A series of options defined at the SEST command are available to the user to control the performance of the importance sampling such as the number of Monte Carlo samples per individual ISAMPLE and scaling of the proposal density relative to the posterior density IACCEPT Termination criteria CITER CALPHA CTYPE and CINTERVAL may also be set which are explained in detail in a later section Typically 300 Monte Carlo samples are needed and 50 200 iterations are required for a randomly stationary objective function that is when the objective function does not vary in a directional manner beyond the Monte Carlo fluctuations The Importance sampling method is specified by EST METHOD IMP INTERACTION Followed by one or more of the following options NITER NSAMPLE 50 Sets maximum number of iterations default 50 Typically 50 100 iterations are need to for a problem to have a randomly stationary objective function ISAMPLE 300 Sets number of random samples per subject use
35. since the last print iteration XML Formatted Output An XML markup version of the standard results output file is automatically produced Control Stream Files may be written in mixed case User defined data labels and file names retain their case designation Stochastic Differential Equations SDE Additional data items have been added to facilitate SDE problems Specialized data labels allow repeated PRED and ERROR calls for a single record but with different EVID values XVID1 XVID2 XVID3 XVID4 XVIDS In addition a plug in routine OTHER SDE f90 is available for Monte Carlo methods but not for FOCE methods that evaluates the stochastic differential equations without requiring coding of these equations in the control stream file by the user See sections 1 54 Repeated Observation Records NM72 and 1 55 Stochastic Differential Equation Plug In NM72 nm730 doc 16 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 CHAIN statement that is applicable to the entire PROB that allows incorporation of initial parameters from raw output files or randomization and serves as parameters for simulations The EST METHOD CHAIN supplies initial parameters from raw output files or randomizations only for the estimation method See section 1 48 Method for creating several instances for a problem starting at different randomized initial positions EST METHOD CHAIN and CHAIN Records Both covariance and correlation matrices to
36. 0 LN V2 The Omegas are to list in order the following NETAxNETA of initial OMEGAS NTHPxNTHP of variances of Priors to THETAS NETPXxNETP of priors to OMEGAS matching the block pattern of the initial OMEGAS INITIAL values of OMEGA NETAxNETA of them OMEGA BLOCK 4 0 15 P 0 01 F 0 15 P 0 01 F 0 01 F 0 15 P 0 01 F 0 01 F 0 01 F 0 15 P Initial value of SIGMA SSIGMA 0 6 7 P nm730 doc 186 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Prior information of THETAS NTHP of them STHETA 2 0 FIX 2 0 FIX 2 0 FIX 2 0 FIX Variance to prior information of THETAS NTHPxNTHP of them Because variances are very large this means that the prior information to the THETAS is highly uninformative OMEGA BLOCK 4 10000 FIX 0 00 10000 0 00 0 00 10000 0 00 0 00 0 0 10000 Prior information to the OMEGAS NETPxNETP of them OMEGA BLOCK 4 0 2 FIX 0 0 0 2 0 0 0 0 0 2 0 0 0 0 0 0 0 2 Degrees of freedom to prior OMEGA matrix 1 for each Omega Prior block Because degrees of freedom is very low equal to the the dimension of the prior OMEGA this means that the prior information to the OMEGAS is highly uninformative STHETA 4 FIX The first analysis is iterative two stage maximum of 500 iterations NITER iteration results are printed every 5 iterations gradient precision SIGL is 4 Termination is tested on all of
37. 00E 00 2 00E 00 1 00E 00 6 00E 00 1 00E 00 1 00E 00 0OE 00 8 06E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 6 00E 00 2 00E 00 1 00E 00 0OE 00 8 06E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 6 00E 00 3 00E 00 1 00E 00 0OE 00 8 07E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 7 00E 00 1 00E 00 1 00E 00 0OE 00 8 07E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 7 00E 00 2 00E 00 1 00E 00 0OE 00 8 07E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 7 00E 00 3 00E 00 1 00E 00 0OE 00 8 08E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 8 00E 00 1 00E 00 1 00E 00 0OE 00 8 08E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 8 00E 00 2 00E 00 1 00E 00 0OE 00 8 08E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 8 00E 00 3 00E 00 1 00E 00 0OE 00 8 09E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 9 00E 00 1 00E 00 1 00E 00 0OE 00 8 09E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 9 00E 00 2 00E 00 1 00E 00 0OE 00 8 09E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 9 00E 00 3 00E 00 1 00E 00 0OE 00 8 10E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 00E 01 1 00E 00 1 00E 00 0OE 00 8 10E 02 0 00E 00 1 00E 12 0 00E
38. 1 Y CUMD MDVRES 1 ENDIF MDVRES stands for missing data value MDV for residual RES assessment Setting MDVRES to 1 is equivalent to temporarily declaring that data point as missing during the weighted residual assessments To incorporate LOQ data into NPDE assessments 4 use the following method as an example Here TYPE and LOQ are user defined in previous code or data item SERROR SD THETA 5 IPRED LOG F DUM LOQ IPRED SD CUMD PHI DUM IF Q 1 OR NPDE MODE EQ 1 THEN PE E F FLAG 0 Y IPRED SD ERR 1 IF TYPE EQ 2 AND NPDE MODE EQ 0 THEN F FLAG 1 Y CUMD MDVRES 1 ENDIF IF TYPE EQ 2 DV LOQ LOO By default DV LOQ is set to 1 0d 300 by the NONMEM routine that calls ERROR PRED If the users ERROR PRED sets DV LOQ to some other value and NPDE_MODE 1 then the NPDE is being evaluated during that time and this censored value is to be treated as if it is a non censored datum with value of LOQ DV_LOQ LOQ in accordance with 4 utilizing a standard F_FLAG 0 definition for Y Note that during estimation of the objective function when NPDE_MODE 0 NPDE is not being evaluated and censored values should be treated using F FLAG 1 and Y must be defined as the integral of the normal density from inf to LOQ ESAMPLE 300 Number of random samples to be used to generat
39. 1 KA TH IF OCC 2 KA TH IF OCC 3 KA TH Another Example SABBR REPLACE TH SABBR REPLACE TH SPK KA THETA SID KA CL THETA SID CL which is equivalent to SPK IF SID 1 KA THETA 4 IF SID 2 KA THETA 6 IF SID 1 CL THETA 5 IF SID 2 CL THETA 7 ETA OCC TH 0 ETA 4 7 10 ETA SID KA THETA 4 6 ETA SID CL THETA 5 7 A list of numbers may be given as SABBR REPLACE TH or by the short hand SABBR REPLACE TH nm730 doc ETA SID KA THETA 4 7 10 13 ETA SID KA THETA 4 to 13 by 3 23 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 At least one comma must appear so NMTRAN knows it is a number list not a variable name Another example Long hand SABBR REPLACE THETA SID KA THETA 4 7 10 13 25 29 33 37 Short hand SABBR REPLACE THETA SID KA THETA 4 to 13 by 3 25 to 37 by 4 Easier Inter occasion variability modeling NM73 Abbreviated code Replacement Feature and Repeated Feature of OMEGA may be combined for easier Inter occasion variability modeling For example SABBR REPLACE ETA OCC CL ETA 4 7 10 when OCC 1 eta 4 to be used when OCC 2 eta 7 to be used etc SABBR REPLACE ETA OCC V ETA 5 8 11 SABBR REPLACE ETA OCC KA ETA 6 9 12 SPK CL TVCL EXP ETA 1 ETA OCC_CL V TVV EXP ETA 2 ETA OCC V KA TVKA
40. 100 NOPRIOR 0 CTYPE 3 CINTERVAL 100 Note the use of the include file nonmem reserved general which for purposes of this example contain the following declarations of reserved variables C ITER REPORT Iteration number that is reported to output C can be negative if during a burn period C BAYES EXTRA BAYES EXTRA REQUEST used in example 8 USE NMBAYES REAL ONLY OBJI USE NMBAYES INT ONLY ITER_REPORT BAYES EXTRA REQUEST BAYES EXTRA USE PNM CONFIG ONLY PNM NODE NUMBER USE NM INTERFACE ONLY TFI TFD nm730 doc 202 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 73 Example 9 Simulated Annealing For Saem using Constraint Subroutine Model Desc Two compartment Model Using ADVAN3 TRANS4 Project Name nm7ex amples Project ID NO PROJECT DESCRIPTION PROB RUN Example 9 from samp51 SINPUT C SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT CLX V1X QX V2X SDIX SDSX DATA example9 csv IGNORE C S SUBROUTINES ADVAN3 TRANS4 OTHER ANEAL F90 PK MU_1 THETA 1 MU_2 THETA 2 MU 3 THETA 3 MU 4 THETA 4 CL DEXP MU_1 ETA 1 V1 DEXP MU_2 ETA 2 Q DEXP MU_3 ETA 3 V2 DEXP MU_4 ETA 4 S1 V1 ERROR Y F F EPS 1 Initial values of STHETA THETA 0 001 2 0 LN CL 0 001 2 0 LN V1 0 001 2 0 LN Q 0 001 2 0 LN V2 1 INITIAL values of OMEGA OMEGA BLOCK 4 0 05
41. 15 0 01 0 01 0 01 0 15 SIGMA 0 06 SETAS 0 x4 SEST METHOD 1 INTERACTION FNLETA 2 MAXEVAL 0 STABLE ID TIME DV IPRED CMT EVID MDV ETA1 ETA2 ETA3 ETA4 NOAPPEND NOPRINT NOTITLE FILE nmtemp tab Note that lt NMID gt is to be replaced with a particular NONMEM ID number by nmtemplate and the THX are to be replaced with specific values of thetas nmtemplate nmtemp nmt nmtemp ctl NMID 47 TH1 1 7 TH2 1 4 TH3 0 8 TH4 2 0 The resulting file nmtemp ctl will have the various values substituted into the various lt gt placeholders and is ready to be read by NMTRAN nmfe73 nmtemp ctl nmtemp res In the above nmtemp nmt example because FNLETA 2 then NONMEM will simply evaluate the IPRED values using the inputted etas from the ETAS record without performing an estimation Another example template file is example6 nmt listed in the util directory that you may inspect for other ideas Actually nmtemplate is a general variable substitution program and can process any text file in the manner shown above Consider a FINEDATA control stream file template util nmtemp fnt SINPUT C SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT CLX V1X QX V2X SDIX SDSX nm730 doc 178 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SDATA nmtemp csv IGNORE C SFINEDATA AXIS TIME LIN TSTOP lt TSTOP gt TSTART lt TSTART gt NEVAL lt NEVAL gt FILE nmtemp2 csv in which the tsta
42. 9 over an unlimited number of lines The first line contains the item LABL in column 1 and subsequent lines have blanks in positions 1 4 LABL ID JID TIME CONC DOSE RATE EVID MDV CMT The LBW1 item contains a comma delimited list of labels for the additional weighted residual type parameters starting at position 6 in each line LBW1 IWRS IPRD NPRED NRES IWRES NIP RED CRES D I R WRES CIE EDI RES RESI IP REDI CR WRESI CIPREDI CIRESI R W D ED ERES EWRES RES EIPRED EIRES E ECWRES NPD The CHAIN record reports its input as follows CHN 2 12345566787 123 300 0 15000E 00 20 3 120 3 CFIL myfile chn CDLM pIBELSZ8 ORDR TSOL Where the mapping for CHN is CHN CTYPE SEED ISAMPLE NSAMPLE IACCEPT DF NOTITNOLAB DFS RANMETHOD where NOTITNOLAB NOTITLE 2 NOLABEL The SIGL and SIGLO are on the second line of the EST item at position 25 and 29 ESTM 09999 7 10 0 0 1 0 1 0 0 O0 60 0 0 0 0 0 0 0 0 0 11 8 SIGL SIGLO The THTA item contains initial theta estimates in a comma delimited list of numbers starting at position 9 in each line THTA 1 100000000000000E 00 1 100000000000000E 00 1 100000000000000E 00 1 100000000000000E 00 1 100000000000000E 00 1 100000000000000E 00 1 100000000000000E 00 1 100000000000000E 00 nm7
43. C SFINEDATA tstart 0 TSTOP 700 NEVAL 500 AXIS TIME LIN CMT 1 4 file mydatab fine csv See also finel infnl infn2 in the examples section of on line help and guide VIII on using the INEN routine and finedata utility to create interpolated values 1 62 nmtemplate Utility Program NM73 The utility program nmtemplate in the util directory will perform variable substitution on appropriately tagged control stream template files and produce executable control stream files The syntax is as follows nmtemplate source template file destination file varl vall var2 val2 var3 val3 nm730 doc 177 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 where varl vall is the variable name and value to substitute in the template file The variable varl must in turn appear as varl in the template file and is case sensitive For example consider the template file util nmtemp nmt SPROB RUN Example 1 from samp51 SINPUT C SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT CLX VIX QX V2X SDIX SDSX SDATA nmtemp2 csv IGNORE C ACCEPT ID EQ lt NMID gt SSUBROUTINES ADVAN3 TRANS4 SPK MU 1 THETA MU 2 THETA MU 3 THETA MU 4 THETA 4 CL DEXP MU_1 ETA 1 V1 DEXP MU_2 ETA 2 1 2 3 D Q DEXP MU_3 ETA 3 V2 DEXP MU_4 ETA 4 S1 V1 SER IPRED F Y F F EPS 1 Initial values of THETA STHETA TH1 TH2 TH3 TH4 SOMEGA BLOCK 4 0 15 0 01 0 15 0 01 0 01 0
44. CTYP SEST METHOD BAY PRINT 50 NOPRIO SEST ME MAXEVAL 9999 NO COV MATRIX R P Note the use of informative names for the prior information see 1 29 A Note on Setting up THOD 1 INT R 0 PRIOR 1 Prior Information ES INTERACTION NOABORT NB ES 545981 5003 0 0 Prior information to the OMEGAS RINT E UNCONDITIONAL 1 39 A note on Analyzing BLQ Data NM73 Since NONMEM VI SIGMA x x has been allowed to be used on the right hand side of equations in the control stream file This has offered a means to obtaining the residual variance in code for example nm730 doc E 3 PRINT 5 NOPRIOR 1 URN 200 NITER 500 CTYPE ERACTION NSIG 3 SIGL 10 PRINT 1 NOABORT NONMEM Users Guide Introduction to NONMEM 7 3 0 IPRED F SD SORT SIGMA 1 1 IPRED Y IPRED IPRED EPS 1 SSIGMA 0 01 Whereas previously to obtain SD a theta needed to be used as the residual coefficient in place of SIGMA SERROR IPRED F SD THETA 1 IPRED Y IPRED SD EPS 1 STHETA 0 1 SSIGMA 1 0 FIXED Furthermore if some data are below level of quantitation BLQ and it is desired to use an integral of the normal density to represent that the value can be anywhere below BLQ this can be modeled using THETA as follows requiring the Laplace method ERROR E ETA 3 IPRED LOQ IPRED SD UMD PHI DUM 1 0E 30 F DV GT LOQ
45. Ea aa aaaeei 93 CFYDE PEE E E E E EE E ETAETA 93 CINTER VA Dineren e aas e OEE de FIR S XR E E E oiea soes 94 CITER o CNSA MIP i arnee coteta pe oO rte UN bN Or rore Eers Eere SES REUSS sotare eo p E 94 CALPHA EE E A E N OE NE NE EEEE 94 1 34 Use of SIGL and NSIG with the new methods sess 95 1 35 List of EST Options and Their Relevance to Various Methods 95 1 36 When to use each method eeeeeeeeeeeeeeeeeeeeeeeee nennen nennt 97 137 COMPOSITES methods sii aae ual ca cl road gai vada ate uet c aee E aA s srt ra aU Cad Vni aa debe 98 1 38 THETAI THI AND THETAR THR Records for Transforming Initial Thetas and Reporting Thetas NM73 esses 99 1 39 A note on Analyzing BLQ Data NM793 esee 101 1 40 ANNEAL to facilitate EM search methods NMY73 103 1 41 COV Additional Parameters and Behavior eese 105 TOL SIGE SIGLO NM72 ionhostortaeb eite ern een veo ee Me e PEE I IM E 105 ATOL NMT2 mr ese 106 NOECOY NMT2 i assccoseie e e ee umo oUm MeL d uan Loto ades tas aoe eion koss Sioa 106 RESUME NMJ iusiosediuc tp masen dtu Diete ia EE Oa bU a ode ei e S OQ TERRAM ORNA RUM 106 1 42 A Note on Covariance Diagnostics e
46. Guide Introduction to NONMEM 7 3 0 sss Ga Ei p i dp il XHAH x x td Q m COAUVUUBRWDND Uc UuUm me 3 d pi Ds ca n ll lt pm ERROR F F EPS 1 Km ll STHETA 2 0 2 0 4 0 4 0 Initial Thetas SOMEGA BLOCK 4 Inital Parameters for OMEGA 0 4 0 01 0 4 0 01 0 01 0 4 0 01 0 01 0 01 0 4 SSIGMA 0 1 E Set the Priors Good Idea if Doing MCMC Bayesian SOMEGA BLOCK 4 Prior to OMEGA 0 2 FIX 022 0 Qie2 04 0 070 0422 0 0 0 0 0 0 0 2 STHETA 4 FIX Set degrees of freedom of OMEGA PRior ITS Store results in sampl5 extra txt SEST METHOD ITS INTERACTION FILE samnp5l extra TXT NITER 30 PRINT 5 NOABORT MSFO msf SIGL 6 Next to SA previous same file SEST METHOD SAEM NBURN 200 NITER 500 PRINT 100 Calculate OBJF by importance sampling SEST METHOD IMP EONLY 1 NITER 5 ISAMPLE 3000 PRINT 1 Store results of Bayesian in its own file EST METHOD BAYES FILE TXT NBURN 200 NITER 3000 PRINT 100 Do an FOCE just for comparison SEST METHOD COND INTERACTION MAXEVAL 9999 NSIG 2 SIGL 6 PRINT 5 COV MATRIX R M Option settings carry over from ST by default So results are added to F P F P Qc More examples of composite analysis are given at the end of this document 1 38 THETAI THI AND THETAR THR Records for Transforming Initial The
47. Jl phc j j Omega j For subject i eta or phi j root shk NM72 This file presents composite eta shrinkage and epsilon shrinkage information the same as given in the report file between the TERM and TERE tags but in rows column format and with adjustable formatting Type 1 etabar Type 2 Etabar SE Type 3 P val Type 4 Eta shrinkage Type 5 EPS shrinkage Type 6 Eta shrinkage based on empirical Bayes Variance Type 7 number of subjects used root shm NM73 As of NM73 the shm table which stands for shrinkage map will contain information which etas were excluded in the eta shrinkage assessment The syntax is as follows For each subject sub population the value listed in column ets j contains the information about whether and how that eta was included in the etabar shrinkage calculations It is a binary value of the format x abcdef where each of the letters may be 0 or 1 If the eta is excluded from the etabar eta shrinkage summary that is recorded in the main NONMEM report file or the shk file then x 1 otherwise it is 0 The remaining binary digits after the decimal point describes conditions about this eta that were involved in deciding whether to exclude this eta a set to 1 if NONMEM assessed this eta as non influential the derivative of the data likelihood with respect to that eta is 0 This exclusion criterion is only acted on that is actually excludes this eta indicated by xz 1 if
48. LIM16 3 double precision values Buffers 1 3 4 13 and 15 are used during an estimation step To obtain the fastest analysis even when the estimation is parallelized you may want to optimize their LIM sizes 1 8 Multiple Runs As of NONMEM 7 there is decreased likelihood of early termination of runs using multiple problems and or the Super Problem feature 1 9 Improvements in Control Stream File input limits 1 By default there may be up to 50 data items per data record In NM72 set PD in SIZES record to change this 2 Data labels may be up to 20 characters long 3 Numerical values in the data file may now be up to 24 characters long 4 ID values in the data file may be up to 14 digits long 5 The numerical values in THETA SOMEGA and SIGMA may be each up to 30 characters long and may be described in E field notation 6 By default you may have up to 50 items printed in tables In NM72 set PDT in SIZES record to change this 1 10 Issuing Multiple Estimations within a Single Problem A sequence of two or more EST statements within a given problem will result in the sequential execution of separate estimations This behavior differs from NONMEM VI where two sequential SEST statements acts as the continuation of defining additional options to a single estimation For example nm730 doc 39 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 STHETA 0 3 0 5 6 SOMEGA 0 2 0 2 0 SSIGMA 0 2 First es
49. NM72 eeeeeeeeeeeeeenennn nnn 168 1 60 xml compare Utility Program and its Use for Installation Qualification NM72 P 169 1 61 finedata Utility Program NMYT 3 eeeeeeeeeeeeeereneeeeeee nennen nnn 172 1 62 nmtemplate Utility Program NMY793 eeeeeeeeeeeeeennneen 177 1 63 Single Subject Analysis using Population with Unconstrained ETAs nm73 180 1 64 Referentes Mp 184 1 65 Example 1 Two compartment Model Using ADVAN3 TRANS4 186 1 66 Example 2 2 Compartment model with Clearance and central volume modeled with covariates age and gender cceceeeseeeeeeeeeeeeeeeeeeeeeeneeeeeeeeeeeeeeeeeennees 189 1 67 Example 3 Population Mixture Problem in 1 Compartment model with Volume and rate constant parameters and their inter subject variances modeled from TWO sub poDUlaHOHRS c ihein exce rti Gids qai eoa uve cp Fa ku wd qua Roo uE cV C DUM X PXM IN Ma SERIE Ve Riad E 191 1 68 Example 4 Population Mixture Problem in 1 Compartment model with rate constant parameter and its inter subject variances modeled as coming from two SUD POPDUIALIONS t H 193 1 69 Example 5 Population Mixture Problem in 1 Compartment model with rate constant parameter mean
50. OMEGAs and SIGMAs are now printed in the NONMEM report file Also all correlation matrices whether to OMEGAS and SIGMAS or pertaining to the correlation matrix of estimates are printed out with diagonal elements equal to the square root of diagonal element of covariance matrix standard error Allow user to input OMEGAs and SIGMAs as standard deviations and or correlations or Cholesky format See Alternative Inputs for 0MEGA and SIGMA Values VARIANCE CORRELATION CHOLESKY NM72 in section l 4 Expansions on Abbreviated and Verbatim Code NM72 NM73 New options for EST SIGLO MAPINTER MAPITER NOHABORT ORDER METHOD DIRECT ISCALE MIN ISCALE MAX CONSTRAIN FNLETA ATOL See the following sections 1 16 Controlling the Accuracy of the Gradient Evaluation and individual objective function evaluation I 17 The SIGLO level NM72 1 25 Monte Carlo Importance Sampling EM 1 26 Monte Carlo Importance Sampling EM Assisted by Mode a Posteriori MAP estimation I 27 Stochastic Approximation Expectation Maximization SAEM Method 1 28 Full Markov Chain Monte Carlo MCMC Bayesian Analysis Method 1 30 Monte Carlo Direct Sampling NM72 1 32 MU Referencing 1 33 Termination testing 1 34 Use of SIGL and NSIG with the new methods New options for COV SIGLO ATOL NOFCOV See section 1 41 COV Additional Parameters and Behavior TABLE has two new special output variables OBJI and NPD OBJI is individual objective function same as g
51. RANMETHOD should be specified only on the first STABLE command The RANMETHOD set in the STABLE command does not propagate to EST or CHAIN As of NM73 the Sobol sequences with scrambling may be requested RANMETHOD nISImIP where n is the random number generator type S is Sobol sequence and m is the Sobol scrambler and P may be specified to retain separate seed patterns for each subject so that the random pattern is retained regardless of single or parallel processing See the description of RANMETHOD under 1 25 Monte Carlo Importance Sampling EM Among the Sobol sequence methods the S2 method appears to provide the least biased random samples that is nearly uniform distribution with good mixing in multi dimensional spaces NOLABEL NM73 Do not print column labels It may be combined with ONEHEADER to print only the title at the beginning of each table NOTITLE NM73 Do not print table titles It may be combined with ONEHEADER to print only the column labels at the beginning of each table NOLABEL NOTITLE is equivalent to NOHEADER nm730 doc 49 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 FORMAT 1PG13 6 This parameter defines the delimiter and number format for the present table and subsequent tables until a new FORMAT is specified The first character defines the delimiter which may be s for space t for tab or the comma The default format is s1PE11 4 The syntax for the number format is Fortran ba
52. a population mean clearance per unit weight which is constant with time observation record and is more universal among subjects The MU variables may vary with inter occasion but not with time Suppose we have a situation where WT has an unknown power term associated with it modeled as THETA 3 in this example CL THETA 2 WT THETA 3 EXP ETA 1 Normally we could efficiently linear model this as follows MU_1 THETA 2 THETA 3 LOG WT CL EXP MU_1 ETA 1 with THETA 2 transformed into the log of clearance domain However if WT changes record by record within the individual then LOG WT may not be in the Mu modeling We would then remove the THETA 3 LOG WT term from MU 1 MU IZLOG THETA 2 CL WT THETA 3 EXP MU_1 ETA 1 And THETA 3 itself would not be MU modeled For NONMEM 7 2 0 NMTRAN is programmed to detect some MU modeling errors Nonetheless the user should verify that these rules are followed Examples at the end of the document show examples of MU modeling for various problem types Study these examples carefully When transposing your own code begin with simple problems and work your way to more complex problems At this point one may wonder why bother inserting MU references in your code MU referencing only needs to be done if you are using one of the new EM or Gibbs sampling methods to improve their efficiency The EM methods may be performed without MU references but it will be several fold slower than th
53. as acceptable to use A copy of the nonmem reserved general file is in the util directory It needs to be placed in the present run directory so NMTRAN has access to it You could opt to copy only part of the list in nonmem reserved general according to need into any file with name starting with nonmem reserved A list of useful variables and their meanings are listed in guides Wiseful variables pdf Be careful in its use as you have the ability to change the values of these reserved variables and this could crash the system if you change the wrong thing Note also that the nonmem reserved general file may contain function declarations such as TFI and TFD which are convenient functions to easily convert an integer to text text from integer TFI or double precision value to text text from double TFD This is quite useful so that the compiler can catch a misuse of that function s arguments If you wish to define your own function and have the information about its proper use of arguments be conveyed upon its execution so the compiler may detect errors then one method is to package the definition of the function in a USE module such as is done in the following example Myfuncmodule f90 defines the functions mymin and mymax MODULE MYFUNCS contains function mymin a b c d e integer mymin integer a b c d e mymin min a b c d e end function function mymax a b c d e integer mymax integer a b c d e mymax max a b c d
54. at the NONMEM command line Modernized Code All code has been modernized from Fortran 77 to Fortran 90 95 The IMSL routines have also been updated to Fortran 90 95 Furthermore machine constants are evaluated by intrinsic functions in FORTRAN which allows greater portability between platforms All REAL variables are now DOUBLE PRECISION 15 significant digits Error processing is more centralized 1 4 Expansions on Abbreviated and Verbatim Code NM72 NM73 FORTRAN 95 Considerations The greatest changes as of NONMEM 7 1 are the renaming of many of the internal variables and their repackaging from COMMON blocks to Modules Whereas formerly a variable in a common block may have been referenced using verbatim code as COMMON PROCM2 DOSTIM DDOST 30 D2DOST 30 30 Now you would reference a variable as follows USE PROCM REAL ONLY DOSTIM And you may reference only that variable that you need without being concerned with order In addition FORTRAN 95 allows you to use these alternative symbols for logical operators Example Fortran 77 nm730 doc 19 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 IF ICALL EQ 3 THEN WRITE 50 CL V ENDIF Fortran 95 IF ICALL 3 THEN WRITE 50 CL V ENDIF The list of operators are Name of logical operator Fortran 77 Fortran 95 Equal to EQ Not equal to NE Greater than GT gt Greater than or equal to GE gt Less t
55. computer or its IP address This is followed by a and a share name of an accessible directory For this example the computer name is any computer and the share name of the directory is share so enter any_computer share Thus from the manager side drive w will be associated with any_computer share which is in fact c share as seen by the worker computer You may be asked to enter username and password Setting up FPI on Windows A versatile loading program called psexec exe freeware from www sysinternals com supplied with the NONMEM installation in the run directory can be used that allows one to load processes locally or on other computers You may choose alternative loading programs Copy psexec exe from the NONMEM s run directory to your managers run directory From a DOS console window type Psexec to see the parameters options for this launching program nm730 doc 143 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 To test that your manager computer can load the NONMEM program on the worker computer if different from manager copy a computername exe from NONMEM s run directory we shall assume it is named NONMEM 7 2 0 to the network mapped directory that is local to the worker Copy nonmem7 2 0 run computername exe w share Then type from the manager console window Psexec any_computer c share computername exe remember these are just example names of computers and network sha
56. course users may have the chain file created by any program not just NONMEM so long as it has the raw output file format with delimiter specified by DELIM FORMAT which is space by default NM73 If the option ISAMPEND is set to a value greater than ISAMPLE then NONMEM will evaluate the objective function using FOCEI method for each sample between numbers ISAMPLE and ISAMPEND in the file and then select the one with the smallest objective function For example SEST METHOD CHAIN FILE random txt NSAMPLE 20 ISAMPLE 1 ISAMPEND 20 randomly creates 20 sets of initial parameters and selects the one with the lowest objective function If METHOD CHAIN is used it must be the first SEST command in the particular PROB Furthermore because the settings it uses for FILE NSAMPLE ISAMPLE IACCEPT CTYPE and DF are functionally different from the way the other EST methods use them these settings from METHOD CHAIN are not passed on to the next EST command which must be an estimation method However other parameters such as DELIM FORMAT SEED AND RANMETHOD will be passed on as default delimiter format to the next EST command However the RANMETHOD does not propagate to the CHAIN record DFS 1 DEFAULT NM73 As of NM73 the SIGMA matrix may be randomly created with an inverse Wishart distribution centered about the initial SIGMA values with degrees of freedom DFS for dispersion If DFS 1 which is the defaul
57. delimited by commas between column items and an amp at the end of a line breaks the record across multiple lines The second file is delimited by spaces between column items and an amp breaks a record across multiple lines table compare mytablel tab mytable2 tab amp c S amp c myprecision xtl mydifferences txt In the above example the first file is delimited by commas between column items and an amp at the end of a line breaks the record with a c at the beginning of the next line The second file is delimited by spaces between column items and an amp at the end of a continuing line and a c at the beginning of the next line table compare mytablel tab mytable2 tab amp SSc myprecision xtl mydifferences txt In the above example the first file is delimited by commas between column items and an amp at the end of a line breaks the record The second file is delimited by spaces between column items and no special character at the end of a continuing line the S serves as a place holder for line contination markers since apace is too ambiguous as a continuator and a c at the beginning of the next line It is useful to redirect difference results to a file in this example mydifferences txt For example the user may desire that only relative differences greater than 0 01 be reported A very simple control file could be PRECISION ALL 0 01 0 003 stating that all columns be compared with a relative difference of
58. drive Consider the following example From the worker computer create a share directory such as mkdir home myself share Next use your editor and sudo privilege to modify the etc hosts file nm730 doc 149 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 sudo gedit etc hosts And map IP address to computer names 127 0 0 1 localhost 192 168 1 3 my manager 192 168 1 2 any computer Then save and exit Use your editor to edit etc exports sudo gedit etc exports Add the following line home myself share 192 168 1 0 24 rw sync Which allows IP addresses 192 168 1 0 through 192 168 1 255 to access this share directory Then exit sudo exportfs a Stop and restart NFS system this is for Ubuntu the command may differ on your computer sudo etc init d nfs kernel server Stop sudo etc init d nfs kernel server restart Go to the manager computer and also place computer names to IP address mapping in etc hosts 127 0 0 1 localhost 192 168 1 3 my manager 192 168 1 2 any computer Then create a mount drive for the remote directory mkdir mnt share sudo gedit etc fstab Enter the mount drive entry for the remote directory any computer home myself share mnt share nfs rw sync 0 0 and exit the editor Then sudo mount mnt share nm730 doc 150 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Test by copying a file from the manager to the worker cp myfile mnt share Next the s
59. etastype 1 b set to 1 if NONMEM excluded this eta for this sub model sub population for this subject because this was not the best fitting sub model for this subject Thus all etas of that subject for all sub models that are not the optimally fitting will have this bit set and only the optimal sub model will have B cleared 0 for all its etas c set to 1 if NONMEM determined that this eta had no influence for this sub model This bit is not set to 1 if bit B is 1 This bit is not set to 1 for non population mixture models Also this exclusion criterion is set and acted upon when FOCE Laplace are used but is not set or acted on for the Em methods IF NONINFETA is set to 1 then FOCE Laplace behave similarly to EM methods and will not set this bit even if the eta has no influence d set if the eta is excluded based on selecting the hybrid option in EST e Set if the user requested an exclusion based on ETASXI i 1 setting in PK or PRED for eta i f Set if the user requested an inclusion based on ETASXI i 2 setting in PK or PRED for eta i Be careful about using this as it over rides all other exclusion criteria except bit B The F bit is the only one that indicates inclusion when set rather than exclusion nm730 doc 122 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 root grd NM72 This file contains gradient values for classical NONMEM methods The format of these files are subject to FORMAT ORDER NOLABEL an
60. fine SEST METHOD COND INTERACTION MAXEVAL 9999 NSIG 3 SIGL 10 PRINT 5 NOABORT NOPRIOR 1 FILE examplel ext Time for the standard error results You may request a more precise gradient precision SIGL that differed from that used during estimation COV MATRIX R PRINT E UNCONDITIONAL SIGL 12 Print out results in tables Include some of the new weighted residual types STABLE ID TIME PRED RES WRES CPRED CWRES EPRED ERES EWRES NOAPPEND ONEHEADER FILE examplel TAB NOPRINT STABLE ID CL V1 Q V2 FIRSTONLY NOAPPEND NOPRINT FILE examplel PAR nm730 doc 187 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 STABLE ID ETA1 ETA2 ETA3 ETA4 FIRSTONLY NOAPPEND NOPRINT FILE examplel ETA nm730 doc 188 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 66 Example 2 2 Compartment model with Clearance and central volume modeled with covariates age and gender Model Desc Two Compartment model with Clearance and central volume modeled with covariates age and gender Project Name nm7examples Project ID NO PROJECT DESCRIPTION PROB RUN example2 from sampc SINPUT C SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT GNDR AGE DATA example2 csv IGNORE C SSUBROUTINES ADVAN3 TRANS4 NTHETA number of Thetas to be estimated NETA number of Etas to be estimated and to be described by NETAxNETA OMEGA matrix NTHP number of thetas which have a prior NETP number of Omegas with prior Prior information is important f
61. following The Sigma values are in their Cholesky format as this is the form in which convergence of these values are tested The Alpha are those based on ones actually used for convergence test of that parameter or which would have been used on that parameter if CTYPE were of proper type The alpha may be bonferoni corrected because of multiple comparisons depending on number of parameters that were tested or would have been tested Objective function alphas are not bonferoni corrected For importance sampling and iterative two stage the average objective function listed in root cnv could be used as an alternative to the final objective function for likelihood ratio tests nm730 doc 123 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 root smt NM72 S matrix if COV step failed root rmt NM72 R matrix if COV step failed root imp NM73 The root imp file is produced if the user selects importance sampling with option IACCEPT 0 0 In such cases this file lists the final ACCEPT and DF values that NONMEM selected for each subject Three files are produced providing nonparametric information root npd NM73 Each row contains information about a support point The support point number the ID from which the support point was obtained as an EBE of that subject ID is 1 if this support point was randomly generated because NSUPP NSUPPE was greater than number of subjects The eta values of the support point are lis
62. following example control stream file portion STHETA 2 0 2 0 4 0 4 0 Initial Thetas SOMEGA BLOCK 4 Inital Parameters for OMEGA 0 4 0 01 0 4 0 01 0 01 0 4 0 01 0 01 0 01 0 4 SIGMA 0 1 PRIOR NWPRI NTHETA 4 NETA 4 NEPS 1 NTHP 4 NETP 4 NEPP 1 Prior information of THETAS NTHP 4 of them nm730 doc 79 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 THETA 2 0 FIX 2 0 FIX 2 0 FIX 2 0 FIX Variance to prior information of THETAS NTHPxNTHP 4x4 of them Because variances are very large this means that the prior information to the THETAS is highly uninformative Note that the order of STHETA values among the THETA records and the order of SOMEGA values among the OMEGA records is very important But STHETAs and SOMEGAs can be interspersed SOMEGA BLOCK 4 10000 FIX 0 00 10000 0 00 0 00 10000 0 00 0 00 0 0 10000 Prior to OMEGA NETPxNETP 4x4 if them SOMEGA BLOCK 4 Set degrees of freedom of OMEGA Prior one value per OMEGA block Uninformative Omega prior is designated by having a DF that is equal to the dimension size of the Omega block THETA 4 FIX Prior to SIGMA NEPPxNEPP 1x1 if them SIGMA 0 05 FIX Set degrees of freedom of SIGMA Prior one value per SIGMA block Uninformative SIGMA prior is designated by having a DF that is equal to the dimension size of the Sigma block THET
63. hat The square root of the matrix V may be evaluated by using the square root of the eigenvalues or by Cholesky decomposition when WRESCHOL option is used see below Similarly the CIPRED is the individual predicted valuef fj at the conditional mode or mean eta and CIRES DV f When INTERACTION is not set then CIWRES V n 0 y f ip is evaluated that is the variance portion is evaluated using f q 0 However CIWRESI conditional individual weighted residual with interaction is always evaluated as except for FO see below CIWRESI V 8 y f i regardless of the INTERACTION setting nm730 doc 46 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 For FO the conditional individual weighted residual will not differ from the non conditional weighted residual That is for FO the CIWRES and CIPRED are evaluated using F eta 0 for numerator and denominator terms since this is what is done during estimation and no EBE eta hat 1s evaluated CIWRES V n 0 y f q 0 CIWRESI Even for FO with interaction the predicted function numerator and residual variance denominator is still evaluated at eta 0 so CIWRESI CIWRES The interaction contribution is accounted for with additional first order Taylor terms to make a linear projection of the contribution of eta eps interaction While it would be inappropriate to add these Taylor terms to CIWRESI these Taylor terms are added to the pop
64. importance sampling IMP performing the expectation step only EONLY 1 so that final population parameters remain at the final SAEM result Five iterations NITER should allow the importance sampling proposal density to become stationary This is observed by the objective function settling to a particular value with some stochastic noise By using 3000 Monte Carlo samples ISAMPLE this assures a precise assessment of standard errors SEST METHOD IMP INTERACTION EONLY 1 NITER 5 ISAMPLE 3000 PRINT 1 SIGL 8 NOPRIOR 1 The Bayesian analysis is performed While 10000 burn in iterations are requested as a maximum because the termination test is on CTYPE lt gt 0 set at the first SEST statement and because the initial parameters are at the SAEM result which is the maximum likelihood position the analysis should settle down to a stationary distribution in Several hundred iterations Prior information is also used to facilitate Bayesian analysis The individual Bayesian iteration results are important and may be need for post processing analysis So specify a separate FILE for the Bayesian analysis SEST METHOD BAYES INTERACTION FILE examplel txt NBURN 10000 NITER 10000 PRINT 100 NOPRIOR 0 Just for old times sake let s see what the traditional FOCE method will give us And remember to introduce a new FILE so its results won t append to our Bayesian FILE Appending to examplel ext with the EM methods is
65. is an optional input and is normally 500 TO continue the user may simply reset ISTATE to a value gt 1 and call again the excess work step counter will be reset to 0 In addition the user may increase MXSTEP to avoid this error return see below on optional inputs 2 Means too much accuracy was requested for the precision of the machine being used This was detected before completing the requested task but the integration was successful as far as T To continue the tolerance parameters must be reset and ISTATE must be set to 3 The optional output TOLSF may be used for this purpose Note If this condition is detected before taking any steps then an illegal input return ISTATE 3 occurs instead 3 Means illegal input was detected before taking any integration steps See written message for details Note If the solver detects an infinite loop of calls to the solver with illegal input it will cause the run to stop 4 Means there were repeated error test failures on one attempted step before completing the requested task but the integration was successful as far as T The problem may have a singularity or the input may be inappropriate 5 Means there were repeated convergence test failures on one attempted step before completing the requested task but the integration was successful as far as T This may be caused by an inaccurate jacobian matrix if one is being used 6 Means
66. it can hide a serious ill posed problem so use with care 1 16 Controlling the Accuracy of the Gradient Evaluation and individual objective function evaluation In classical NONMEM methods First order First order conditional Laplace the user specifies SIGDIGIT or NSIG to indicate the number of significant digits that population parameters are to be evaluated at the maximum likelihood If NSIG 3 the default then the problem would be optimized until all of the parameters varied by less than 3 significant digits This same NSIG value would also be used to specify relative step size h to each THETA SIGMA and OMEGA for evaluating the partial derivative of the objective function with respect to the parameter Such partial derivative evaluations are needed to set up gradients to determine the direction the search algorithm must travel to approach the minimum of the objective function The forward finite difference of the partial derivative of O the objective function with theta 1 would be evaluated as O 8 A 8 Oh Numerical analysis of forward finite difference methods 6 recommends that the ideal relative step size h for the parameter theta 1 should be no greater than SIGL 2 where SIGL is the significant digits to which the objective function is evaluated If h is set to a precision of SIGL 2 which for the present discussion we mean it is set to 105 82 then the resulting derivative itself will have approximately SIGL 2 precision a
67. lines for launching the worker nodes is completely dependent on your computing and parallel distribution environment and the syntax requirements of the launching program The psexec exe program located in the run directory of the NONMEM folder is available for Windows to launch a program on the same computer as with the first 2 worker nodes or on a remote computer last worker node An alternative launching program may be used The w option in psexec specifies the working directory as the worker identifies it from which the NONMEM programs is to be launched nm730 doc 138 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 The index numbers that begin an item in a list 1 2 etc are optional If present it refers to node 1 manager node 2 node 3 etc If not present the item number is determined by the order in which the item was listed It is best to use them for greater clarity In SDIRECTORIES the directory names must follow syntax rules of the particular operating system The SDIRECTORIES record is optional If left out or if a directory name is not given for a process Then the default values are NONE for common directory position 1 workerl for the first worker position 2 worker2 for the second worker position 3 etc These are interpreted as sub directories to the present run directory There is no need to create the worker directories ahead of time although its parent directory whether local or netwo
68. manager s computer the worker only needs certain of the parameters from the command line 2 wdir cd Nworkerl hosts 1 localhost 1 noprompt nonmem exe This launches a worker process on a separate computer 3 wdir c share worker3 n 1 host any worker noprompt continued on same line c share worker3 nonmem exe e DIRECTORIES nm730 doc 148 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 NONE FIRST DIRECTORY IS THE COMMON DIRECTORY 2 workerlN NEXT SET ARE THE WORKER directories 3 w share worker3 SIDRANGES USED IF PARSE TYPE 3 1 1 50 2 51 100 An additional setting in GENERAL is introduced called COMPUTERS By default COMPUTERS is equal to 1 However if you are running MPI method on Windows and you have at least one of the worker processes on another computer and your LIM values are not maximized so that some file buffers are being used then you may need to set COMPUTERS 2 If you obtain a read write error on FILE10 or other FILEXX error then set COMPUTERS 2 Unlike FPI the MPI system can only use the starting parallel pnm file specified at the command line and it may not be easily switched later in the control stream All processes remain resident throughout the entire job although it will honor requests of parafile off or parafile on at individual EST records which allows you to have control of which estimation method will use parallel processin
69. modeled for two sub populations but its inter subject variance is the same in both sub populations eere 195 1 70 Example 6 Receptor Mediated Clearance model with Dynamic Change in alelo olol E pe RE RRREER ERRIRERTERERERERERE 196 1 71 Example 7 Inter occasion Variability euren 198 1 72 Example 8 Sample History of Individual Values in MCMC Bayesian Analysis 199 1 73 Example 9 Simulated Annealing For Saem using Constraint Subroutine 203 nm730 doc 7 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 74 Example 10 One Compartment First Order Absorption Pharmaokinetics with Galegorical Date iis cactus nonien ds ae Da EM ace Deve Cau dvo dua uiu a UI UR Kan CO Mo ove DU IG RR Saaie 205 1 75 Description OF FCON TI Igis uou bii dide e bec ie ena edil ld oh bw bud ibn d Rena 207 nm730 doc 8 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 1 What is new in NONMEM Version 7 3 0 versus NONMEM 7 2 0 The main new features of NONMEM 7 3 compared to NONMEM 7 2 0 are as follows Execution script nmfe73 offers more control in discerning location of compiler and mpi system This option can facilitate execution of NONMEM in which there can be potential conflict with other software that may use alternative compilers and mpi systems See section l 5 Invoking NONMEM and the locfile op
70. of the variance of the proposal density in order to meet the IACCEPT condition is in NM72 by default bounded by ISCALE_MIN of 0 1 and ISCALE_MAX 10 0 On very rare occasions the importance sampling objective function varies widely and the scale factor boundary may need to be reduced perhaps ISCALE_MIN 0 3 ISAMPLE MAX 3 After the importance sampling estimation remember to revert these parameters to default operation on the next EST step ISCALE_MIN 100 ISCALE_MAX 100 Note the values to ISCALE MIN and ISCALE MAX for the IMP method in NONMEM 7 1 and earlier were 0 01 100 respectively and were not changeable by the user EONLY 1 Evaluate the objective function by performing only the expectation step without advancing the population parameters default is 0 population parameters are updated When this method is used NITER should equal 5 to 10 to allow proposal density to improve with each iteration since mean and variance of parameters of normal or t distribution proposal density are obtained from the previous iteration Also it is good to get several objective function values to assess the Monte Carlo noise in it SEED 14456 default The seed for random number generator used in Monte Carlo integration is initialized default seed is 11456 MAPITER 1 default NM72 By default MAP estimation is performed only on the first iteration to obtain initial conditional values modes and approximate variances to be used for the
71. oiii aep Ur DA ER DM ERO ob MOIS RU nSt S Ss 73 Obtaining the Objective Function for Hypothesis Testing After an SAEM Analysis 74 1 28 Full Markov Chain Monte Carlo MCMC Bayesian Analysis Method 75 EST METHOD BAYES INTERACTION eere en eese ense teen ene te easet teens seta eas enata 76 NBURN 4000 T H A 76 NSAMPEE NITER 10000 odit or bebe Rm took DI eO EN CHIA EUR cassachovbousanseoneausnsesasinns 76 ISAMPLE M1 2 defaults listed scscccccccsssstcsescccsecccisseveacesscosvassetesssdecsasesteteooscebesesssscstussccssces 76 TSAMPLE M1AZU NMT72 5 ecscceiesenenss ice bcidnetor ere petuneedes dno tes duret ees eder Ue ea caase tisano sage ade eros ENS 76 ISAMPLE M2 Diss ss o icones nestrive tbe eee esee eee prie e Ve ea eva pedi Fiebre CL P VI S s SeS 76 ISAMPLE M32 piain erroe vio st rider L et souti ORDER Un be kV VE ERA Eo erede iden ES 76 IXCCEPT 04 EEEE A EE A EAA EE E E 76 ISCALE_MIN 1 0E 06 defaults for SAEM BAYES NM72 esssessssossesescsccocscsoesesssesosee 77 ISCALE MAX 1 0E 06 NM aosserascceseecn ausa deca eso aea aod deaur ou vi daran are sino deris 77 PSAMPLE_M1 1 defaults listed cossssscscosssescsssossessecsenssascosssssesseesessonsessensoasessnsssossens 77 BSAMPEE M25 I easi cuneta REP QUUM NO Be iR ERU ERA A e EE PATUE RASEN UE REA IRURE 77 nm730 doc 4 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 DA UPS EL Mc
72. options for the configure step If SETUP72 installed 64 bit binaries configure prefix usr local mpi64 CFLAGS m64 FFLAGS m64 enable f90 disable cxx amp tee c txt If SETUP72 installed 32 bit binaries configure prefix usr local mpi32 enable f90 amp tee c txt Either way continue with make l amp tee m txt make install l amp tee mi txt Then replace libmpich a in the NONMEM 72 directory e g if 32 bit was installed cd opt nm72 mpi mpi ling cp libmpich a libmpich a orig cp usr local mpi32 lib libmpich a libmpich a 1 54 Repeated Observation Records NM72 To assist in specialized methodologies such as stochastic differential equations 14 15 16 a record in a data file may be set up for repeated calls to PK and ERROR Each time the same record is passed through PK and or ERROR but with a different EVID The user s control stream model in PK or SERROR may then take advantage of executing certain code conditional on the EVID value For this to occur the user must introduce one or more of the following data items in the data file with these names XVIDI XVID2 XVID3 XVID4 XVID5 These stand for extra EVID s On the first call to PK ERROR the EVID is set to the value given in XVIDI On the second call the EVID is set to that in column XVID2 etc up to XVIDS Only as many XVID s as are required are needed to be defined All the other items in the record do not change except that if the presen
73. prediction residual and weighted residual data items NPDE EWRES etc these data items comprise the residual record The default size of buffer 2 is related to the number LIM2 of residual records stored in memory at any one time The size of buffer 2 has been set to allow LIM2 100 000 residual records for up to 100 000 data records The least number of residual records allowable must exceed the largest number of data records used with any one subject Each residual data record consists of 19 eight byte double precision computer words The allocation of memory for buffer 2 is 19 LIM2 3 8 bytes Buffer 3 holds a number of contiguous subject header records for input data The size of buffer 3 is related to the number LIM3 of subject header records stored in memory at any one time The default size of buffer 3 has been set to allow LIM3 1000 subject header records Each subject header record consists of four 8 byte computer words The allocation of memory for buffer 3 is 4 LIM3 1 8 bytes Buffer 4 holds a number of contiguous ETA records For each subject NONMEM generates values for ETA variables The size of buffer 4 1s related to the number LIM4 of ETA records stored in memory at any one time The size of buffer 4 has been set to allow LIM4 1000 ETA records Each ETA record consists of MMX LVR 8 byte double precision computer words The allocation of memory for buffer 4 is MMX LVR LIM4 43 8 Buffer 5 holds a number of contiguous mixtur
74. problem when run singly or in parallel will report a similar cpu time This is in contrast with elapsed time which is improved with parallelization 1 48 Method for creating several instances for a problem starting at different randomized initial positions EST METHOD CHAIN and CHAIN Records The METHOD CHAIN option of the SEST command allows the user to create a series of random initial values of THETAS and OMEGAS or for reading in initial population parameters from a file of rectangular rows column format Consider the following example SEST METHOD CHAIN FILE examplel chn DELIM NSAMPLE 5 CTYPE 0 ISAMPLE 3 DF 100 SEED 122234 RANMETHOD 2 IACCEPT 0 5 In this example NSAMPLE random samples of THETAS and OMEGAS will be generated and written to a file specified by FILE using comma as a delimiter SEED sets the starting seed for the random samples By default CTYPE 0 random values of theta are generated from a uniform distribution spanning from lower bound theta to upper bound theta specified in the THETA statement If a nm730 doc 125 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 boundary for a theta is not specified then 1 IACCEPT THETA is used for a lower bound and 1 IACCEPT THETA is used for an upper bound For the SIGMA values their Cholesky decomposed values are uniformly varied between 1 ACCEPT SIGMA and 1 IACCEP
75. sampling density Subsequently nm730 doc 67 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 the Monte Carlo assessed conditional means and variances from the previous iteration are used as parameters to the sampling density However the user can select the pattern by which MAP estimations are intermittently done and their conditional statistics used for the sampling density MAPITER n means the first n iterations are to use MAP estimation to assess parameters for the sampling density After these n iterations the conditional means and variances of the pervious iteration are used for the sampling density parameters of the present iteration If MAPITER O then the first iteration will rely on conditional means and variances that are in memory These may have come from an MSF file or from a previous estimation step MAPINTER 0 default NM72 Every nth iteration the MAP estimation should be used to provide parameters to the sampling density Thus if MAPITER 20 and MAPINTER 5 then for the first 20 iterations MAP estimation is used and thereafter every 5 iteration the MAP estimation is used If MAPINTER 1 NM73 then mapinter will be turned on only if the objective function increases consistently over several iterations Setting an option to 100 will force NONMEM to select the default value for that parameter DF 4 The proposal density is to be t distribution with 4 degrees of freedom Default DF 0 is normal density The
76. ssmultidose ctl for additional examples Subscripted Variables Enhancement NM73 Subscripts may be used with user defined variables that are declared to be arrays using the ABBR DECLARE record and also with certain reserved variables such as THETA Subscripts may be integer variables and expressions For example SABBR DECLARE INTEGER IND SABBR DECLARE X 10 SPK IND 1 X IND THETA IND 1 Autocorrelation CORRL2 NM73 Correlation of residual variables using CORRL2 may now be written in abbreviated code For example examples ar1mod ctl SABBR DECLARE T NO SABBR DECLARE DOWHILE J SABBR DECLARE INTEGER I SERROR IF NEWIND NE 2 I 0 IF MDV EQ 0 THEN I I41 T I TIME J 1 DOWHILE J I CORRL2 J 1 EXP THETA 4 TIME T J J J41 ENDDO ENDIF Simulation with autocorrelation is also possible A new example is provided examples arlnewsim ctl MOD Function NM73 The Fortran intrinsic function MOD may now be used in abbreviated code k MOD i j MOD returns the remainder when i is divided by j The variables i and j must be either both integer or both real However this function should not be involved in evaluation of the objective function nm730 doc 25 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 MIN MAX Functions NM73 The Fortran intrinsic functions MIN and MAX ma
77. status text in the report file For example an iterative two stage analysis may be requested followed by an MCMC Bayesian method followed by an FOCEI method The theta sigma and omega results of the iterative two stage method will be passed on as initial values for the MCMC Bayesian method to facilitate the MCMC Bayesian analysis which in turn can supply initial values for the FOCEI method Each of these intermediate analyses will provide output to the NONMEM report file and will be identified by unique text for that method To allow a program to consistently find the appropriate positions in the file without having to search for specific words in the text the report file is augmented with special tag labels that remain constant regardless of the method used The tags always begin with followed by four letters to indicate the tag type followed by a colon The following tags are presently defined SPARA NM72 This tag identifies the parallelization file and number of nodes used if parallel estimation is performed TBLN NM72 This tag specifies that following it on the same line will be found an integer that refers to the number of this estimation method This number is also the table number listed in the title to tables in the various output files raw output file cov cor etc The table number is incremented for each EST statement across all problems in the control stream file METH This tag specifies that f
78. symbolically in the mpd hosts for example as nm730 doc 155 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 MY MANAGER COMPUTER WORKER A COMPUTER WORKER B COMPUTER So long as these symbolic names are listed in the etc hosts file with the IP address The number of computers is number of worker computers not cores plus the manager computer If loading just on one computer then mpdboot n 1 To unload MPI after your last NONMEM run mpdallexit See section 5 of mpich2 1 2 1 userguide pdf for a full description of using the man MPI program mpiexec or mpirun Once you have an MPI system set up for a quick test on a single multi core computer try the following Copy foce parallel ctl and examplel csv from the NONMEM examples directory mpilinux8 pnm from the NONMEM run directory and psexec exe from the NONMEM run directory into your standard run directory Then execute the following from your standard run directory Nmfe73 foce parallel ctl foce parallel res parafile mpilinux8 pnm nodes 4 where the values of nodes should be no greater than the number of cores available on your computer A typical structure of a pnm file for running NONMEM MPI Linux note TRANSFER TYPE 1 is as folllows GENERAL NODES 2 PARSE TYPE 2 PARSE NUM 50 TIMEOUTI 100 TIMEOUT 10 PARAPRINT 0 TRANSFER_TYPE 1 NODES number of nodes that is process whether cores or computers SINGLE node NODES 1 MULTI node node mea
79. the program will modify the matrix to be positive definite will report that it has done so and provide the standard errors The user should use the standard error results with caution should a non positive definite flag occur The ITS and SAEM methods can only evaluate the S matrix and will do so even if MATRIX R is requested The banner information will show what type of variance was evaluated The BAYES method always supplies standard errors correlation matrix and covariance matrix even when COV step is not requested as these results are a direct result of summarizing the accumulated NITER samples Furthermore the matrices are always positive definite and therefore always successful To obtain the eigenvalues to the correlation matrix even for the BAYES method a COV step must be issued with the PRINT E feature TOL SIGL SIGLO NM72 The TOL used by PREDPP when differential equations are integrated and SIGL and SIGLO may be set specifically for the COV step distinct from those used during EST This special option for COV is not so important for the new EM or BAYES methods which are able to obtain suitable standard errors using SIGL SIGLO and TOL that are also used for estimation but classical NONMEM methods in particular can require a different significant digits level of evaluation usually more stringent during the COV step than during EST Keep in mind that when evaluating the R matrix SIGL and TOL should be at lea
80. theta parameters it does not have to re evaluate the predicted function which can be computationally expensive especially if one of the differential equation solver ADVAN s are used An alternative method for specifying GRD modeled parameters is by using the following syntax GRDzt vi ni t2v2 n2 t3v3 n3 Where t refers to a parameter type T for theta S for SIGMA v refers to a letter S D or N and n refers to a number list For example to specify thetas 3 5 through 8 to be Gibbs samples theta 4 is sigma like and sigmas 1 3 are to be Metropolis Hastings processed GRD TG 3 5 8 TS 4 SN 1 3 Thetas and sigmas not specified are given a default D designation The SN designation is also used by EM methods to not determine the derivatives of the objective function with respect to the Sigmas analytically which is faster but numerically 1 33 Termination testing A termination test is available for importance sampling iterative two stage burn in phase of SAEM and the burn in phase of MCMC Bayesian It is during burn in that one wishes to know when the sampling has reached the stationary distribution for SAEM and BAYES The second sampling stage in SAEM and BAYES still is determined by how many samples NITER or NSAMPLE are desired to contribute to the final answer so convergence does not apply there There are four parameters set in the SEST statement to specify the termination options CTYPE CTYPE 0 no terminatio
81. this expression could be transformed to LOG ABS X 1 0E 300 to avoid arguments to LOG that are non positive nm730 doc 133 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 If you have an expression which is ultimately exponentiated then there is a potential for floating point overflow An expression such as EXP X Which is likely to cause a floating point overflow could be filtered with IF X gt 100 0 EXIT EXP X Again if the calculation must complete such as when evaluating a user defined likelihood then you can place a limiting value taking care that it causes little first derivative discontinuity EXPP THETA 4 F THETA 5 Put a limit on EXPP as it will b xponentiated to avoid floating overflow IF EXPP GT 40 0 EXPP 40 0 FLAG 1 Categorical data IF EXPP gt 40 then A gt 1 0d 17 A B approaches 1 1 B approaches 0 and Y is approximately DV D 1 DEXP EXPP A DV A B 1 DV B a likelihood If your code uses SQRT phrases the expression within parentheses should be always positive Sometimes expressions are calculated to near zero but slightly negative values such as 1 1234444555E 16 Such values may legitimately be 0 but square rooting a negative number could result in failure of analysis In such cases the difficulty is due to the finite precision of the computer e g rounding error causing a value to be negative that would be non negative o
82. this is MU 2 1 0 CL THETA 5 MU_2 ETA 2 Which would mean that in the end THETA 5 is not actually MU modeled since MU_2 does not depend on THETA 5 One would be tempted to model as follows nm730 doc 87 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 MU 2 THETA 5 CL MU 24MU 2 ETA 2 But this would be incorrect as MU 2 and ETA 2 may not show up together in the code except as MU_2 ETA 2 or its equivalent Thus THETA 5 cannot be MU modeled In such cases remodel to the following similar format CL THETA 5 EXP ETA 2 So that THETA 5 may be MU modeled as MU 2 LOG THETA 5 CL EXP MU_2 ETA 2 Again for EM methods better to re parameterize as MU 2 THETA 5 CL EXP MU_2 ETA 2 And log transform the initial value of THETA 5 Sometimes a particular parameter has a fixed effect with no random effect such as Km THETA 5 with the intention that Km is unknown but constant across all subjects In such cases the THETA 5 and Km cannot be Mu referenced and the EM efficiency will not be available in moving this Theta However one could assign an ETA to THETA 5 and then fix its OMEGA to a small value such as 0 0225 0 1542 to represent 15 CV if OMEGA represents proportional error This often will allow the EM algorithms to efficiently move this parameter while retaining the original intent that all subjects have similar although not identical Km s Very often inter subject var
83. to NONMEM 7 3 0 IPRED E0 EMAX DOSE HILL ED50 HILL4DOSE HILL Y IPRED EPS 1 STHETA 4 1 1 Emax STHETA 6 9 2 ED50 STHETA 0 001 3 Hill STHETA 2 3 4 EO SOMEGA BLOCK 2 0 1 0 01 0 1 SOMEGA 0 1 SOMEGA 0 0 FIXED SANNEAL 4 0 3 SIGMA 1 SESTIMATION METH SAEM INTER NBURN 1000 NITER 500 ISAMPLE 5 IACCEPT 0 3 CINTERVAL 25 CTYPE 0 NOABORT PRINT 50 CONSTRAIN 5 SIGL 8 SESTIMATION METH IMP INTER PRINT 1 NITER 0 ISAMPLE 10000 EONLY 1 CONSTRAIN 0 MAPITER 0 DF 4 COV MATRIX R UNCONDITIONAL The user may modify the subroutine CONSTRAINT that performs the simulated annealing algorithm The source code to the CONSTRAINT subroutine is available from the source directory as constraint f90 and the user may copy this to their run directory and as convenient to rename it Then specify OTHER name of source f90 in the SUBROUTINE record as shown in example 9 The subroutine CONSTRAINT may also be used to provide any kind of constraint pattern on any parameters Another technique is to use an initial Monte Carlo search method using EST METHOD CHAIN ISAMPEND and then use the standard gradient method for SAEM as follows SPROB Emax model with hill 3 SINPUT ID DOSE DV SDATA anneal dat IGNORE PRED MU 1 THETA 1 EMAX EXP MU_1 ETA 1 MU 2 THETA 2 ED50 EXP MU_2 ETA 2 MU 3 THETA 4 EO EXP MU_3 ETA 3 MU_4 THETA 3 HILL EXP MU_4 ETA 4 IPRED EQ EMAX DOSE HILL
84. useful in combination with the SIML record SCHAIN FILE test2 chn ISAMPLE 3 ISAMPEND 10 NSAMPLE 10 SEED 6234 SSIML 112345 334567 NORMAL SUBP 4 SEST METHOD IMP INTERACTION NITER 40 PRINT 1 NOABORT SIGL 4 CTYPE 3 CITER 10 In the above example for the first subproblem a file called test2 chn is created and stores to NSAMPLE 10 randomly created sets of thetas omegas and sigmas numbered 1 to NSAMPLE Then a sample of parameters is selected from this file uniformly randomly between ISAMPLE 3 and ISAMPEND 10 and these parameters are used to create a data set for the first sub problem and an estimation is performed For the second sub problem a new file of parameters does not need to be created but another sample is selected randomly uniformly between samples 3 and 10 from which a new data set is created and estimation analysis performed The parameter file may already exist perhaps as a raw output file from a previous MCMC Bayesian analysis and it is desired to randomly selected sets of parameters SCHAIN FILE examplel chn ISAMPLE 0 ISAMPEND 10000 NSAMPLE 0 SEED 6234 SSIML 112345 334567 NORMAL SUBP 100 In the above example NSAMPLE 0 so this means the file examplel chn already exists which is in fact the raw output file examplel txt from the MCMC Bayesian analysis of examplel Samples from 0 to 10000 the stationary distribution range are selected ra
85. user may copy this to their run directory and as convenient to rename it Then specify OTHER name of source f90 in the SUBROUTINE record as shown in example 9 nm730 doc 73 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 As of NM73 when CONSTRAIN gt 4 simulated annealing is also performed on diagonal elements of OMEGAS that are fixed to 0 to facilitate estimation of any associated thetas See 1 40 ANNEAL to facilitate EM search methods for this additional annealing technique The subroutine CONSTRAINT may also be used to provide any kind of constraint pattern on any parameters The mapping of parameters between Monolix and NONMEM SAEM is as follows Monolix NONMEM SAEM Number of Chains ISAMPLE KO CONSTRAINT subroutine may be user modified to provide any constraining pattern on any population parameters Kl NBURN K2 NITER Auto K1 CTYPE 1 2 3 Population Parameter settings menu rho IACCEPT ml ISAMPLE MI m2 ISAMPLE MIA m3 ISAMPLE M2 m4 ISAMPLE M3 No simulated annealing CONSTRAIN 0 Simulated Annealing CONSTRAIN 1 2 3 User may also define algorithm SEED SEED Obtaining the Objective Function for Hypothesis Testing After an SAEM Analysis After the analysis suitable objective functions for hypothesis testing and second order standard errors can be obtained by importance sampling at the final population parameter values Thus one could issue this sequ
86. user may simply desire to obtain the minimum maximum values obtained nm730 doc 132 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 51 Imposing Thetas Omegas and Sigmas by Algebraic Relationships Simulated Annealing Example Additional algorithmic constraints may be imposed upon the model parameters by use of the subroutine CONSTRAINT This feature is available only for the EM and Bayesian algorithms One use would be to slow the rate of reduction of the diagonal elements of the OMEGA values during the burn in phase of the SAEM method This is shown in example 9 where a user supplied annealing algorithm is used to replace the built in one described earlier By specifying OTHER ANEAL f90 where ANEAL f90 was originally derived from a template of CONSTRAINT 90 in the source directory the user supplied CONSTRAINT subroutine can be incorporated into the model In example 9 whenever iteration number ITER_NO changes a new OMEGA is evaluated that is larger than what was determined by the SAEM update Typically this expansion algorithm should be such that its impact decreases with each iteration 1 52 Stable Model Development for Monte Carlo Methods The Monte Carlo EM and Bayesian methods create samples of etas from multi variate normal or t distributions Because of this some extreme eta values may be randomly selected and sent to the user developed model specified in PK SPRED DES and or ERROR Usually these extreme
87. values but using the final THETAS and SIGMAS It is recommended that the initial OMEGAs have inflated values relative to the final OMEGAS which is usually the case to allow the outlier subjects to be fitted with little constraint from the population distribution For each subject the EBE that provides the highest individual likelihood value not the highest posterior density whether from the final fit EBE or the expanded OMEGA EBE is selected as a support point This is the inflated variance recommendation from 7 NPSUPP NM73 NONP NPSUPP 50 Number of total support points to be used If NPSUPP gt number of subjects then extra support points are randomly created from the final OMEGAS even when EXPAND is selected for the base EBE support points This is the extended Grid Method as described in 7 NPSUPPE NM73 NONP NPSUPPE 50 Number of total support points to be used If NPSUPPE gt number of subjects then extra support points are randomly created from the initial presumably inflated OMEGAS even when EXPAND is not selected for the base EBE support points BOOTSTRAP NM73 NONP BOOTSTRAP The original data set is fitted during the parametric estimation SEST and the eta support points from the original data set are used for the nonparametric version However a bootstrap sample with subjects uniformly randomly selected with replacement from the original data set is used for the nonparametric distribution analysis
88. worker 1 WK2 FILE for worker 2 etc This way even if the workers and manager share the same directory as a scratch pad their files will be uniquely named and there won t be a file clobber An alternative method of launching mpi processes is to use its multiple process launch option n xx where xx is the number of processes to launch GENERAL NODES 8 PARSE TYPE 2 TRANSFER TYPE 1 PARAPRINT 0 COMPUTERS 2 COMMANDS 1 mpiexec wdir PWD n 1 nonmem 2 wdir SPWD n 3 host MY MANAGER COMPUTER nonmem wnf 3 wdir SHOME n 4 host MY WORKER COMPUTER nonmem wnf S DIRECTORIES 1 8 NONE 3 mnt workerl Command 2 launches 3 processes and command 3 launches 4 processes so there are still 8 processes launched Special Considerations for MAC OS X Mounting file systems on MAC OS X It is easier to use afp Apple Filing Protocol than nfs To export a file system or folder to another Mac Select the Apple menu System Preferences Sharing File Sharing Under shared folders click and select the folder e g mydir Under users click and select the users nm730 doc 159 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 To mount a file system or folder from another Mac Open a finder window You should see the hostname of the other computer listed under Shared Click on it Click on connect as Enter the username and password Click on the folder e g mydir The file syste
89. 0 0 0 0 0 0 2 THETA 4 FIX SEST METHOD BAYES INTERACTION FILE example8 ext NBURN 10000 NITER 1000 PRINT 100 NOPRIOR 0 CTYPE 3 CINTERVAL 100 Note that the contents is written to file fort 50 and fort 51 If parallelization is used then fort 50 and fort 51 files in each of the worker directories will be created and must be collected after the run to obtain records for all of the subjects Alternatively specific file names may be given the names being created according to the node number However care must be given the specific directory location is valid for a given run example8b Model Desc Two compartment Model Using ADVAN3 TRANS4 Project Name nm7examples Project ID NO PROJECT DESCRIPTION PROB RUN Example 8 from samp51 SINPUT C SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT CLX V1X QX V2X SDIX SDSX DATA example8 csv IGNORE C abbr DECLARE INTEGER FIRST WRITE INTEGER FIRST WRITE2 SUBROUTINES ADVAN3 TRANS4 PRIOR NWPRI NTHETA 4 NETA 4 NTHP 4 NETP 4 PK include nonmem reserved general nm730 doc 200 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Request extra information for Bayesian analysis An extra call will then be made for accepted samples BAYES EXTRA REQUEST 1 MU_1 THETA 1 MU_2 THETA 2 MU_3 THETA 3 MU_4 THETA 4 CL DEXP MU_1 ETA 1 V1 DEXP MU_2 ETA 2 Q DEXP MU_3 ETA 3 V2 DEXP MU_4 ETA 4 S1 V1 When Bayes extra 1 then thi
90. 0 01 and absolute difference of 0 003 Precision crietria for specific columns in the tables may also be given PRECISION ALL 0 01 0 003 WRES 0 1 0 2 CL 0 05 0 02 The equation for comparison is if ABS X Y gt R MAX ABS X ABS Y A then the difference is reported where R is relative difference tolerance and A is absolute difference tolerance 1 59 table_to_xml Utility Program NM72 The utility table_to_xml program in the NONMEM util directory can be used to convert additional NONMEM output tables produced during the SEST step into XML formatted files The syntax is as follows as an example table to xml my results cov my results cov xml nm730 doc 168 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 where the delimiter may be t or s for comma tab or space Default delimiter is space The rules schema document type definition by which the xml file is constructed are given in tables xsd and tables dtd which are in the run or util directory table to xml my results cov my results cov xml amp c specifies that the table file may have line continuator characters amp and c as described in the table compare section 1 60 xml compare Utility Program and its Use for Installation Qualification NM72 The utility program xml compare will compare the contents of two NONMEM report XML files that are produced by NONMEM The syntax to the command line is xml compare myresult1 xml myresult2 xml myprec
91. 0 characters to allow for the largest range of objective function values and the greatest expression of precision The iteration number which is the first value in every line is typically positive but also may be negative under the following conditions 1 The burn in iterations of the MCMC Bayesian analysis are given negative values starting at NBURN the number of burn in iterations requested by the user These are followed by positive iterations of the stationary phase 2 The stochastic iterations of the SAEM analysis are given negative values These are followed by positive iterations of the accumulation phase 3 Iteration 100000000 negative one billion indicates that this line contains the final result thetas omegas and sigmas and objective function of the particular analysis nm730 doc 118 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 4 Iteration 100000001 indicates that this line contains the standard errors of the final population parameters 5 Iteration 100000002 indicates that this line contains the eigenvalues of the correlation matrix of the variances of the final parameters 6 Iteration 100000003 indicates that this line contains the condition number lowest highest Eigen values of the correlation matrix of the variances of the final parameters 7 Iteration 100000004 indicates this line contains the OMEGA and SIGMA elements in standard deviation correlation format 8 Iteration 100000005 in
92. 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 00E 01 2 00E 00 1 00E 00 0OE 00 8 10E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 00E 01 3 00E 00 1 00E 00 OOE 00 8 11E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 10E 01 1 00E 00 1 00E 00 0OE 00 8 11E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 10E 01 2 00E 00 1 00E 00 0OE 00 8 11E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 10E 01 3 00E 00 1 00E 00 OOE 00 8 12E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 20E 01 1 00E 00 1 00E 00 0OE 00 8 12E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 20E 01 2 00E 00 1 00E 00 0OE 00 8 12E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 20E 01 3 00E 00 1 00E 00 0OE 00 8 13E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 30E 01 1 00E 00 1 00E 00 0OE 00 8 13E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 30E 01 2 00E 00 1 00E 00 0OE 00 8 13E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 30E 01 3 00E 00 1 00E 00 OOE 00 8 14E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 40E 01 1 00E 00 1 00E 00 OOE 00 8 14E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 40E 01 2 00E 00 1 00E 00 OOE 00 8 14E 02 0 0
93. 00 3000 100 100 100 100 00000 3000 4000 100 1 1556678 0 0 3 5 0 05000 Default values are designated 100 or 100 0 The parameters are right justified in their respective fields and are identified as follows BEST method psample ml psample m2 psample m3 paccept osample mi Osample m2 osample m3 oaccept isample isample ml isample m2 isample m3 iaccept nsample nburn df eonly Seed noprior nohead ctype citer calpha Cinterval mapiter mapinter isample mla iscale min iscale max Constrain atol fnleta Ranmethod mceta noninfeta isampend etastype auto stdobj numder pscale min pscale max where Method 1 any classical NONMEM method nm730 doc 208 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 lt thod 10 DIRECT thod 11 BAYES thod 12 ITS thod 13 IMP thod 14 IMPMAP thod 15 SAEM Method 16 CHAIN nohead notitle 2 nolabel Sees e e e e e e lt BEST is followed by the following items which contain text starting at position 9 BFIL examplel chn BDLM 1PE12 5 BMUM DDMMX BGRD NNGGD ORDR PFIL Where BFIL contains the FILE name given in SEST BMUM contains MUM BGRD contains GRD ORDR NM72 contains order pattern for output to additional results file and PFIL NM72 contains parafile name After a COVR item there is a COVT item with two integers starting at position 9 and spaced 4 positions apart They are the SIGL TOL SIGLO ATOL NM72 NOFCOV NM72 RESUME
94. 0E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 40E 01 3 00E 00 1 00E 00 OOE 00 8 15E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 50E 01 1 00E 00 1 00E 00 0OE 00 8 15E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 50E 01 2 00E 00 1 00E 00 0OE 00 8 15E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 50E 01 3 00E 00 1 00E 00 0OE 00 8 16E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 60E 01 1 00E 00 1 00E 00 0OE 00 8 16E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 60E 01 2 00E 00 1 00E 00 0OE 00 8 16E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 60E 01 3 00E 00 1 00E 00 OOOOOO00000000000000505050000500505000000500000000000000000 The idea in doing this is to cause the following term to be added to the objective function Nap gt 0 2 0 log Z i l nm730 doc 110 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Where 0 1s the vector of SID thetas and 21s the variance among the SID thetas For the above example 0 is a 3x1 vector one element each for KA TYPE 1 CL TYPE 2 and V TYPE 3 for i 1 to NSID where NSID is the number of possible values of SID which in this example NSID 16 The X matrix is the 3x3 block matrix to Epsilons 2 3 and 4 NONMEM is fooled into constructing the above term by use of the additional d
95. 1 PSAMPLE M2 1 PSAMPLE M3 1 OSAMPLE M1 1 OSAMPLE M2 1 OACCEPT 0 5 ISAMPLE M1 2 ISAMPLE M1A 0 ISAMPLE M2 2 ISAMPLE M3 3 METHOD SAEM INTERACTION CTYPE 3 NITER 1000 NBURN 4000 ISAMPEND 10 NOPRIOR 1 CITER 10 CINTERVAL 0 CALPHA 0 05 IACCEPT 0 4 ISCALE MIN 1 0E 06 ISCALE MAX 1 0E 06 ISAMPLE M1 2 ISAMPLE M1A 0 ISAMPLE M2 2 ISAMPLE M3 2 CONSTRAIN 1 EONLY 0 ISAMPLE 2 METHOD ITS INTERACTION CTYPE 3 NITER 500 NOPRIOR 1 CITER 10 CINTERVAL 1 CALPHA 0 05 METHOD IMP INTERACTION CTYPE ER 500 ISAMPLE 300 ISAMPEND 10000 NOPRIOR 1 CITER 10 CINTERVAL 1 CALPHA 0 05 IACCEPT 0 0 ISCALE MIN 0 1 ISCALE MAX 10 DF 0 MCETA 3 EONLY 0 MAPITER 1 MAPINTER 1 ll RW Z H Ei I METHOD IMPMAP INTERACTION CTYPE 3 NITER 500 ISAMPLE 300 ISAMPEND 10000 NOPRIOR 1 CITER 10 CINTERVAL 1 CALPHA 0 05 IACCEPT 0 0 ISCALE MIN 0 1 ISCALE MAX 10 DF 0 MCETA 3 EONLY 0 I The AUTO option is ignored by the FO FOCE Laplace methods The AUTO setting itself transfers to the next SEST within the same PROB just like any other option settings explicitly set by the user in the control stream file so AUTO remains on or off until then next AUTO option specified For example in the following example EST METHOD ITS AUTO 1 PRINT 10 EST METHOD SAEM AUTO 1 PRINT 50 EST METHOD IMP PRINT 1 EONLY 1 NITER 5 ISAMPLE 1000 EST METHOD BAYES AUTO 1 FILE auto txt PRINT 100 N
96. 2 AND SDE EQ 4 THEN AHT1 0 PHT1 A2 ENDIF IF A OFLG EQ 1 THEN A 0 1 AHT A 0 2 PHT ENDIF SDES DADT 1 CL VD A 1 0 DADT 2 CL VD A 2 CL VD A 2 SGW1 SGW1 SERROR OBS ONLY IPRED A 1 VD IRES DV IPRED W SQRT A 2 1 VD 2 THETA 3 2 IWRES IRES W Y IPRED W EPS 1 SEST MAXEVAL 9999 METHOD 1 LAPLACE NUMERICAL SLOW INTER NOABORT SIGDIGITS 3 PRINT 1 MSFO sde8 msf COV MATRIX R STABLE ID TIME FLAG AMT CMT IPRED IRES IWRES EVID ONEHEADER NOPRINT FILE sde8 fit nm730 doc 162 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 With the following fragment of the data file ID TIME DV AMT CMT FLAG MDV EVID SDE XVID1 XVID1 XVID3 1 0 0 1000 1 0 1 1 2 me 1 1 1 0 5 24 317 0 1 1 0 0 2 0 2 3 1 1 18 469 0 1 1 0 0 2 0 2 3 1 d 18 018 0 1 T 0 0 2 0 2 3 is 2 18 728 0 ab 1 0 0 2 0 2 3 1 2 59 13 445 0 1 1 0 0 2 0 2 3 1 3 14 924 0 1 1 0 0 2 0 2 3 1 3 5 11 846 0 jT 1 0 0 2 0 2 3 1 4 10 691 0 1 1 0 0 2 0 2 3 1 4 5 9 9394 0 T a 0 0 2 0 2 3 1 5 9 9075 0 dt 1 0 0 2 0 2 3 1 949 10 7 0 1 1 0 0 2 0 2 3 1 6 8 9861 0 1 1 0 0 2 0 2 3 1 3 7 2274 0 T jT 0 0 2 0 2 3 1 8 6 4909 0 1 1 0 0 2 0 2 3 1 9 3 7281 0 1 1 0 0 2 0 2 3 1 10 1 9238 0 ab 1 0 0 2 0 2 3 1 11 2 172 0 1 1 0 0 2 0 2 3 1 12 1 0763 0 1 1 0 0 2 0 2
97. 2 THETA 2 MU 3 THETA 3 MU 4 THETA 4 CL DEXP MU 1 ETA 1 V1 DEXP MU_2 ETA 2 Q DEXP MU_3 ETA 3 V2 DEXP MU_4 ETA 4 S1 2V1 ETA 5 SLEVEL SID 5 11 Let us suppose that the data item named SID is the site ID NONMEM needs to know that SID is to be associated with eta 5 and in turn eta 1 is nested within eta 5 The data file need not be sorted for super ID values The LEVEL record gives this information a such that SID is a super ID data item associated with eta 5 inter site eta and eta 1 nests within eta 5 5 1 NONMEM will then perform appropriate summary statistics for eta 5 and make the appropriate constraints on eta 5 so eta 5 changes by site that is by every SID value change and not by every ID value change You may have additional parameters having site variability etas and their suitable nesting etas such as for V1 Q and V2 SPK MU_1 THETA MU 2 THETA MU 3 THETA MU 4 THETA 4 CL DEXP MU_1 ETA 1 ETA 5 V1 DEXP MU_2 ETA 2 ETA 6 1 2 3 nm730 doc 107 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Q DEXP MU 3 ETA 3 ETA 7 V2 DEXP MU_4 ETA 4 ETA 8 S1 V1 SLEVEL SID 5 11 6 21 7 3 1 8 4 Perhaps in addition to SID you have country ID let s call that data item CID Perhaps there are several sites belonging to one country some other sites belonging to another country etc This would provide a nesting level of 2 above that of
98. 3 2 0 0 1000 1 0 1 1 2 lt zl 2 0 5 17 586 0 1 1 0 0 2 0 2 3 2 1 13 758 0 1 1 0 0 2 0 2 3 2 125 9 6241 0 1 1 0 0 2 0 2 3 2 2 9 6419 0 1 1 0 0 2 0 2 3 2 2 9 8 5945 0 1 1 0 0 2 0 2 3 2 3 6 3709 0 1 1 0 0 2 0 2 3 2 34 5 7 7656 0 1 1 0 0 2 0 2 3 2 4 4 5152 0 1 1 0 0 2 0 2 3 2 4 5 5 0167 0 1 i 0 0 2 0 2 3 2 5 4 6339 0 1 1 0 0 2 0 2 3 2 55 4 2107 0 1 1 0 0 2 0 2 3 2 6 3 1452 0 1 1 0 0 2 0 2 3 2 7 2 0888 0 a a 0 0 2 0 2 3 2 8 2 4506 0 1 1 0 0 2 0 2 3 2 9 0 001 0 1 a 0 0 2 0 2 3 2 10 1 1174 0 1 1 0 0 2 0 2 3 2 11 0 001 0 1 1 0 0 2 0 2 3 2 12 0 001 0 1 1 0 0 2 0 2 3 Compare this data file with sde7 csv with its repeated data record and see its control stream file examples sde7 ctl which is the traditional way of programming an SDE problem in NONMEM The examples sde6 ctl control stream file is the problem without an SDE component 1 55 Stochastic Differential Equation Plug In NM72 An alternative method to evaluating stochastic differential equation problems is to utilize the plug in routine SDE f90 in the NONMEM examples directory which numerically evaluates the SDE equations without requiring in line coding into the control stream An example control stream file is as follows examples sde9 ctl nm730 doc 163 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SPROBLEM PK ODE HANDS ON ONE SINPUT ID TIME DV AMT CMT FLAG MDV SDE SDATA Sde9 csv
99. 30 doc 207 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Similarly items LOWR lower bound thetas UPPR upper bound thetas BLST block variances elements and DIAG diagonal variance elements are formatted the same as THTA BLST and DIAG may have additional integer indicators in positions 5 8 on their first line as before The ANNL NM73 contains parameters to the SANNEAL record with omega element followed by its starting value ANNL 3 4 The SIML record has attached to it starting at position 57 the simulation RANMETHOD The OLEV NM73 contains parameters to the SLEVEL command The data column name pertaining to the level is in columns 9 to 28 and the level description begins at position 29 OLEV SID 3 1 4 2 OLEV CID 5 3 6 4 The NOMSFTEST NM73 option to MSFI is recorded as a 1 in column 32 of the FIND record FIND o 0 1 0 0 1 The NOREPLACE NM73 and BOOTSTRAP NM73 option settings are in positions 41 and 45 to the SIML record respectively SIML 0 1 0 10 0 0 0 0 1 50 The nonparametric NM73 bootstrap option at postion 25 expand options at position 29 1 3 EPXAND 2 4 NSUPPE option number of supplementary points NSUPP E begins at column 33 NONP 1 0 0 0 il 50 The item BEST contains parameters for the additional parameters to the SEST command The values begin at position 5 and are 12 spaces apart 6 parameters per line BEST 11 100 100 100 100 00000 1 10 O 100 000
100. 6 telling NONMEM to report the objective function of each subject as a data likelihood without an eta population density or an integral over all etas component added This is called POPULATION WITH UNCONSTRAINED ETAS analysis versus the standard SINGLE SUBJECT or POPULATION and will be labeled as such in the NONMEM report file under ANALYSIS TYPE For this example all thetas are fixed to 0 as well so that the etas contain the full values of the individual parameters to which they are associated KA K CL and residual variance W1 squared Since thetas are no longer in play in indestm initial etas become relevant so the ETAS record is used to introduce them and MCETA 1 assures that these initial etas as well as etas 0 are tested at the beginning of the etas curve fitting the MAP estimation as viable starting positions Also since all of the traditional population parameters THETAS SIGMAS and OMEGAS are fixed only a single evaluation MAXEVAL 0 is necessary To compare the results of indestm with those of indestb note that the four etas in indestm phi match with the final three theta parameters and OMEGA 1 1 listed in indestb ext or indestb res and notice that the individual objective functions of subjects listed in indestm phi match with the final objective function of each of the 12 single subject analyses in indestb ext Furthermore the variance covariance etas ETC listed in indestm phi match with the variance covariance of
101. 7 3 0 recent PARAFILE file specification If parafile off is given at the command line then no parallelization is done for the entire control stream regardless of PARAFILE options within the control stream file The format of the parallel file is best shown by this example which is heavily commented to describe the meanings of the records and options available This parafile example is set up for FPI method on Windows GENERAL NODES 2 PARSE_TYPE 3 PARSE_NUM 200 TIMEOUTI 60 TIMEOUT 10 PARAPRINT 0 TRANSFER_TYPE 0 NODES number of nodes that is process whether cores or computers SINGLE node NODES 1 MULTI node node means process whether cores or computers NODES gt 1 WORKER node NODES 0 parse num number of subjects to give to each node parse type 0 give each node parse num subjects parse type 1 evenly distribute numbers of subjects among available nodes parse type 2 load balance among nodes parse type 3 assign subjects to nodes based on idranges parse type 4 load balance among nodes taking into account loading time This setting of parse type will assess ideal number of nodes If loading time too costly will eventually revert to single CPU mode timeouti seconds to wait for node to start if not started in time deassign node and give its load to next worker until next iteration timeout minutes to wait for node to compelte if not completed by then deassign no
102. 730 doc 166 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 57 Ignoring Non Impact Records During Estimation NM73 Typically users may produce data files that are augmented with additional non dose non observation records in order to output predicted values at additional times to create high resolution curves However too many of such records tend to slow down the estimation analysis As of NM73 if an MDV is set to a value greater than or equal to 100 it is converted to that value minus 100 upon input but will not be used during estimation or covariance assessment only for table outputting This option allows you to use the same file for estimation and table outputs without significantly slowing down the estimation So if MDV 101 it will be converted to 1 upon use for final evaluations and the records will be ignored during estimation The subroutines in NONMEM that ignore MDV 100 and MDV 101 records are OBJ all estimation and covariance steps OBJ2 parametric OBJ3 non parametric and OS initial estimates of omegas and sigmas Care must be taken in using MDV gt 100 in that during estimation covariate data items of these records are not used which can have a slightly different interpolation impact than what is finally recorded in the tables where they are used You may specifically request that any one of these routines not ignore the MDV gt 100 records by setting MDVII 1 for OBJ to include MDV gt 100 records MD
103. A 1 FIX By default the number of prior experiments is 1 However perhaps you have more than one previous study and you wish to average their contribution forming a composite average set of prior parameters to influence the present analysis In this case add NEXP n to the NWPRI record above where n is the number of experiments Then add the prior information of each additional study with additional STHETA SOMEGA and SIGMA statements The order is then THETA records list the parameters in order the following NTHETA of initial thetas Exp 1 NTHP of Priors to THETAS Degrees of freedom to each OMEGA block Prior Degrees of freedom to each SIGMA block Prior Exp 2 NTHP of Priors to THETAS Degrees of freedom to each OMEGA block Prior Degrees of freedom to each SIGMA block Prior nm730 doc 80 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 The 0MEGA records list the variances in order the following NETAXNETA of initial OMEGAS Exp 1 NTHPxNTHP of variances of Priors to THETAS NETPxNETP of priors to OMEGAS matching the block pattern of the initial OMEGAS Exp 2 NTHPxNTHP of variances of Priors to THETAS NETPxNETP of priors to OMEGAS matching the block pattern of the initial OMEGAS The SIGMA records list the variances in order the following NEPSxNEPS of initial SIGMAS Exp 1 NEPPxNEPP of priors to SIGMAS matching the block pattern of the initial SIGMAS Exp 2 NEPPxNEPP of priors to SIGMAS
104. CNIVETS sic usdieiecide eov obi Fue cet lesbcudivedccusnestelubaieas toacduencetinedsoeaSeepecetisban teas lestestinendousiecdes 124 root npe NMTS adsueti deci iesmied dvs sdatedscccvelaviudsssbudamiteviusatscdeasustevadlicaivduiiel Miadetesseeseubsstadeas 124 TOOL TPL NMTS sewer sessssivscecdiuvas sep Vasa Eur eY dan t akin adiuvat expo exque ed Pav V ia iv ss oddsen si epa V seasons Seaseaavadecdes 124 root feh NM T3 aoc Gibt S P cess cub oO Geen ec PANE bv Gules IR RN MI MINI Maid 125 root dohi INMTO iaiia i asiota nieou RR FE su aos Dat eu Rz obe D Ceca asasan aeaaeai OSES a oaasi teaei east 125 FOOL CDU UNIVETS seosciticesistantsbeletisactccanccsucccarendecstesietatusdeveeisactutacosuacscevedpatudisustcaaasdecdidivesdesaiesens 125 1 48 Method for creating several instances for a problem starting at different randomized initial positions EST METHOD CHAIN and CHAIN Records 125 DES 1 DEFAULT NINITS csiexsnsvonnuasbassexsvetesnetisonhevenslenvedenusslesticuantvassesssseadeousavassuaveseuevgvons 128 SCHAIN Record cacao Re ca PU E casa ote ERG ER UR ODER ES FRE A NIU CON RAS Sci Roch 128 SELECT 0 DEFAULT NM73 sccccssssnsissascicecsisenasveucesiashesasvensovanisacnadeessehsecnasbeassuboccoanseubencs 130 1 49 ETAS and PHIS Record For Inputting Specific Eta or Phi values NM73 130 1 50 Obtaining individual predicted values and individual parameters during MCMC Bayesian Analysis cieswssesecoseaiantnarandaaddeaduiitaaacaitanstaw
105. D 1 FIX intial parameters to sigma SIGMA 0 1 Set degrees of freedom of OMEGA Prior one value per OMEGA block SOMEGAPD 4 FIX Prior to OMEGA NETPxNETP 4x4 if them SOMEGAP BLOCK 4 Variance to prior information of THETAS NTHPxNTHP 4x4 of them THETAPV BLOCK 4 10000 FIX 0 00 10000 0 00 0 00 10000 0 00 0 00 0 0 10000 Prior to SIGMA NEPPxNEPP 1x1 if them SSIGMAP 0 05 FIX Informative prior information may come from a previous study Typically they are used as follows The theta priors for the present analysis are obtained from the estimates of thetas from the previous study The variance covariance to theta priors of the present analysis are obtained form the variance covaraince submatrix pertaining to the theta estimates from the previous study The omega priors of the present analysis are obtained from the estimates of omegas from the previous study The degrees of freedom to the omega priors of the present analysis are at most the total number of subjects in the previous study Dr Mats Karlsson has proposed the following formula for selecting degrees of freedom DF 2 Omega estimate of previous analysis SE of omega of previous analysis Or DF 2 Omega estimate of previous analysis SE of omega of previous analysis 41 nm730 doc NONMEM Users Guide Introduction to NONMEM 7 3 0 to adjust for degrees of freedom loss in the estimate of Omeg
106. D MDV CMT WT SDATA finetest csv IGNORE C SFINEDATA NEVAL 1 AXIS TIME LIN MISSING 99 WT LIN file finetest7 csv SPROB RUNS example6 from r2compl SINPUT C SET ID JID TIME DV CONC DOSE AMT RATE EVID MDV CMT WT SDATA finetest7 csv IGNORE C SFINEDATA tstart 0 TSTOP 50 NEVAL 250 AXIS TIME LIN CMT 1 3 WT LIN PREV MISSING 99 file finetest7a csv A scheme to determine how to supply values to various data items for these inserted records may also be given For example to specify that the value of the next original record should be used to supply the value for WT in the inserted record SFINEDATA WT NEXT The following values may be given NEXT When inserting records between two consecutive original records of time tl PREV and t2 NEXT the PREDPP s default of using the covariate value of the t2 NEXT record is used for the inserted records NEXT is the default PREV When inserting records between two consecutive original records of time t1 PREV and t2 NEXT the covariate value of the t1 PREV record is used for the inserted records LAST may be coded instead of PREV to be consistent with the options of the BIND record Note that the BIND record is not used by finedata nm730 doc 174 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 LIN or LINLIN A covariate linear time linear interpolation is used for the covariate value for the inserted records LINT or LINLINT T for truncate produces trunca
107. D4 records stored in memory at any one time The size of buffer 7 has been set to allow LIM7 2 NMPRDA records which is generally fewer than the number of NMPRD4 records existing for any given subject Each NMPRD4 record consists of LIM7 2 LNP4 8 byte double precision computer words The default allocation of memory for buffer 7 is 4 LNP4 8 bytes nm730 doc 38 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 The memory allocation of Buffer 8 is LVR 1 LIM8 3 double precision values Buffer 11 holds a number of contiguous problem header records The size of buffer 11 is related to the number LIM11 of problem header records stored in memory at any one time The size of buffer 11 has been set to allow LIM11 25 problem header records Each problem header record consists of forty two 8 byte integer computer words The allocation of memory for buffer 11 is 42 LIM11 3 8 9408 bytes The memory allocation of Buffer 13 is 404 LIM13 3 double precision values After NONMEM VI there are also buffers 15 and 16 The sizes of these buffers are related to constants LIM15 and LIM16 These buffers are used in DAT15 and DATI6 If LIM16 is not adequate NONMEM will produce error messages such as the following TOT NO OF RESIDUAL RECS IN BUFFER 16 IS LESS THAN NO OF DATA RECS WITH SOME INDIVIDUAL The memory allocation of Buffer 15 is LCM110 LIM15 3 double precision values The memory allocation of Buffer 16 is MMX 4
108. E 00 1 00E 00 3 00E 00 1 00E 00 0OE 00 8 02E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 2 00E 00 1 00E 00 1 00E 00 0OE 00 8 02E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 2 00E 00 2 00E 00 1 00E 00 0OE 00 8 02E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 2 00E 00 3 00E 00 1 00E 00 0OE 00 8 03E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 3 00E 00 1 00E 00 1 00E 00 0OE 00 8 03E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 3 00E 00 2 00E 00 1 00E 00 0OE 00 8 03E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 3 00E 00 3 00E 00 1 00E 00 0OE 00 8 04E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 4 00E 00 1 00E 00 1 00E 00 0OE 00 8 04E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 4 00E 00 2 00E 00 1 00E 00 0OE 00 8 04E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 4 00E 00 3 00E 00 1 00E 00 0OE 00 8 05E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 5 00E 00 1 00E 00 1 00E 00 0OE 00 8 05E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 5 00E 00 2 00E 00 1 00E 00 0OE 00 8 05E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 5 00E 00 3 00E 00 1 00E 00 0OE 00 8 06E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0
109. E 01 0OE 00 1 00E 00 1 00E 02 1 42E 06 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 10E 01 1 00E 00 0 00E 00 1 10E 01 0OE 00 2 00E 00 0 00E 00 0 00E 00 1 00E 00 0 00E 00 1 00E 00 1 00E 00 2 00E 00 1 20E 01 1 00E 00 0 00E 00 1 00E 00 0OE 00 2 00E 00 1 00E 01 2 73E 01 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 30E 01 1 00E 00 0 00E 00 2 00E 00 0OE 00 2 00E 00 2 00E 01 2 79E 01 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 40E 01 1 00E 00 0 00E 00 3 00E 00 0OE 00 2 00E 00 5 00E 01 2 68E 01 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 50E 01 1 00E 00 0 00E 00 4 00E 00 0OE 00 2 00E 00 1 00E 00 2 32E 01 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 60E 01 1 00E 00 0 00E 00 5 00E 00 0OE 00 2 00E 00 2 00E 00 1 74E 01 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 70E 01 1 00E 00 0 00E 00 6 00E 00 0OE 00 2 00E 00 5 00E 00 1 30E 01 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 80E 01 1 00E 00 0 00E 00 7 00E 00 OOOOoo0o0o0000000o0o0ooon0Q Added data portion TYPEz1 2 3 to provide variance constrained among the SID values and bind it to the inter SID SIGMA variance ID TIME DV AMT RATE EVID MDV CMT ROWNUM SID TYPE L2 0OE 00 8 01E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 00E 00 1 00E 00 1 00E 00 0OE 00 8 01E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 00 1 00E 00 2 00E 00 1 00E 00 0OE 00 8 01E 02 0 00E 00 1 00E 12 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00
110. E WT SINCE DOSE IS WEIGHT ADJUSTED CALLFL 1 A THETA 1 THETA 2 L THETA 3 SC CL K Q STHETA 0 001 3 0 001 2 0 001 1 SOMEGA 2 For single subject data OMEGA is residual variance nm730 doc 180 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SERROR Y F ERR 1 ERR must be used instead of EPS SEST MAXEVAL 450 PRINT 5 SCOV SPECIAL MATRIX R PRINT E SPECIAL is required to obtain the variance covariance matrix for single subject data STABLE ID DOSE WT TIME NOPRINT ONEHEADER FILE indestb tab NOTITLE STABLE ID KA K CL SC NOPRINT FIRSTONLY NOAPPEND FILE indestb par NOTITLE ONEHEADER INCLUDE indestb txt 11 INCLUDE Inserts copies of the file named indestb txt for each additional individual which performs the analysis for the first subject and the accompanying include file performs analysis on the subsequent subjects SPROB THEOPHYLLINE POPULATION DATA Analysis of Individuals SINPUT ID DOSE AMT TIME CP DV WT SDATA THEOPP RECS ID NOREWIND NOREWIND data set will be read starting after the previous individual STHETA 0 001 3 0 001 2 0 001 1 SOMEGA 2 For single subject data OMEGA is residual variance SEST MAXEVAL 450 PRINT 5 SCOV SPECIAL MATRIX R PRINT E SPECIAL is required to obtain the variance covariance matrix for single subject data STABLE ID DOSE WT TIME NOPRINT FORWARD NOHEADER FILE indestb tab STABLE ID KA K CL SC NOPRINT FIRSTONLY FORWARD NOAPPEND NOHEAD
111. E definition in an include file that begins with the name NONMEM RESERVED case insensitive at the beginning of the section you want to use it For example NONMEM RESERVED GENERAL in the util directory has many quite useful variables listed including ITER REPORT in the form of C ITER REPORT Iteration number that is reported to output C can be negative if during a burn period C BAYES EXTRA BAYES EXTRA REQUEST used in example 8 USE NMBAYES REAL ONLY OBJI USE NMBAYES INT ONLY ITER REPORT BAYES EXTRA REQUEST BAYES EXTRA USE PNM CONFIG ONLY PNM NODE NUMBER USE NM INTERFACE ONLY TFI TFD The user may use any one of these variables such as shown in example 8 PK include nonmem reserved general BAYES EXTRA REQUEST 1 MU 1 THETA 1 MU 2 THETA 2 MU 3 THETA 3 MU 4 THETA 4 CL DEXP MU_1 ETA 1 nm730 doc 26 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 V1 DEXP MU_2 ETA 2 Q DEXP MU_3 ETA 3 V2 DEXP MU_4 ETA 4 S1 V1 IF BAYES EXTRA 1 AND ITER_REPORT gt 0 AND TIME 0 0 THEN WRITE 50 ITER REPORT ID CL V1 Q V2 ENDIF Note the lack of needing to begin a line with when using ITER REPORT BAYES EXTRA REQUEST or BAYES EXTRA because NMTRAN read the nonmem reserved general file and listed the variables declared in there
112. E eai e db e Pra eR REEL sne 66 y Ws E BLUT C E TE 66 ISAMPEND n STDOBJI d NM73 oie sesustkiutucceUtk one bRpH VER Rea HeHRRPM ae UR RE USE I LIN PREIS RR EPUM RR een 66 nm730 doc 3 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 TA CCE PT LLL PT sdususens 67 IACCEPT 0 0 NM7 3 cssvelacses eivsatersndatecsanencdcatesesncstxosdiesantabnes Oevsetonselacesasuedddeavedesusssevscdoeeinans 67 ISCALE_MIN 0 1 defaults for IMP NM72 scscssssssssssosssscsssessescesssscessscssasssssessesers 67 ISCALE MAX 10 0 NM72 sivscssisscsdesoteosvssbssoaaitecssnsiundstaasondepetsvisoannabesesunguacabiasendeseaioscasons ns 67 EONLY ST aint Glsabstnn nono EMG Oe AE E IRD NR NIE 67 SEEDs14456 defa lt zie eres dni diit orinse gt vti ba ene subcarecsseesssbasterdsnsidesessaubevetiosaes 67 MAPITER 1 default NIVIT2 sscccesccsiceassunccceaevdacecstesncConceddhessdrvssrenssosscesteoacdenscdeaesseiecoeenevouccs 67 MAPIN FER 0 default NM772 icit siodcosecun mt un ca peicu sana ce E ceni ara acam tain era eara ep dea end 68 nr c MH 68 RANMETHOD nISImIP NM72 default n 3 e eere eee eee eee eee eene ee eee en en ueo 68 Note on the t Distribution Sampling Density DF gt 0 and its Use With Sobol Method RANME THODES D C 70 1 26 Monte Carlo Importance Sampling EM Assisted by Mode a Posteriori MAP cili zii or ce 70 EST METHOD IMPMAP INTE
113. ER FILE indestb par Another method now available in NM73 is for NONMEM to treat all the subjects as part of a population analysis but if all OMEGA diagonals are set to 1 0E 06 FIXED this is a key value to indicate to NONMEM that there is no population density constraint for etas associated with the posterior density effectively making the posterior density strictly a data likelihood In the following example the indestb problem was restructured to implement this method as shown here in examples indestm ctl SPROB THEOPHYLLINE POPULATION DATA SINPUT ID DOSE AMT TIME CP DV WT SDATA THEOPP SSUBROUTINES ADVAN2 SPK THETA 1 MEAN ABSORPTION RATE CONSTANT 1 HR THETA 2 MEAN ELIMINATION RATE CONSTANT 1 HR THETA 3 SLOPE OF CLEARANCE VS WEIGHT RELATIONSHIP LITERS HR KG SCALING PARAMETER VOLUME WT SINCE DOSE IS WEIGHT ADJUSTED CALLFL 1 A THETA 1 ETA 1 K THETA 2 ETA 2 CL THETA 3 ETA 3 SC CL K STHETA 0 0 FIXED X4 SOMEGA 1 0E 06 FIXED X4 nm730 doc 181 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SETAS 3 08 04 0 2 SERROR W1 SQRT ABS THETA 4 ETA 4 IPRED F Y F W1 EPS 1 SSIGMA 1 0 FIXED SEST METHOD 1 INTERACTION LAPLACE MAXEVAL 0 PRINT 5 NOHABORT FNLETA 0 MCETA 1 STABLE ID DOSE TIME DV IPRED W1 NOAPPEND NOPRINT FILE INDESTM TAB STABLE ID KA K CL NOAPPEND FIRSTONLY NOPRINT FILE INDESTM PAR Notice in the above example that OMEGA diagonals are set to 1 0E 0
114. ES CIWRESI NMYT3 e eere eere ee eene eene testen setas seas sete sena 46 NIDVRES O NM73 default ei sse ton born su pre bak ee EF eY supera en tbe PEE EEPPATE EE QRPP SEDEM REDE 47 ESAMPLUES3O00 EEA T boe iV E DER SERRE E AREE ERRAT TA E 48 nm730 doc 2 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 WRESCHOL NMJT3 asscscesieisexes cionivtetivtveenseN inet prie to ere rad Rd uan e REO Roh e riisin stese 48 lU D D Sesterce RR 49 RANMETHOD nISImIP NM72 default n 3 eere eee esee eee esee en aenean tn enana 49 NOL ABEL NM73 P ieaie 49 NOTITLE NM T3 istis ES DISD EE RSEN er SERM eH QAM MR RH OR S 49 FORMAT IPGID O tod iti riore dior ect ta nbn sotisa serasi sebisa korsas dea de oed iae eU Aa cR 50 LEORMAT RFORMAT NM72 lt ccccssessisesscensedsosssonensensnessonsnsensecensuctcechoeenddentedosceseuassennsscosas 51 1 14 SUBROUTINES New Differential Equation Solving Method 52 ATOL NM72 P P 53 MXSTEP NMZ3 csc sesloceasaavtaceeaseascesmease esa veacecan seas Uu OR rate tid eI AOI NU THEN HN M UE 53 1 15 EST Improvement in Estimation of Classical NONMEM Methods 54 1 16 Controlling the Accuracy of the Gradient Evaluation and individual objective function evaluation irse kan Rn ha Rak CY RR a E XR ERRARE RERO X E XR XH RI ERR Ia 54 1 17 The SIGLO level NM72 ciuthuen un e
115. EWT I became zero for some I during the integration Pure relative error control ATOL I 0 0 was requested on a variable which has now vanished The integration was successful as far as T 7 Means the length of RWORK and or IWORK was too small to proceed but the integration was successful as far as T This happens when DLSODA chooses to switch methods but LRW and or LIw is too small for the new method Note Since the normal output value of ISTATE is 2 it does not need to be reset for normal continuation Also since a negative input value of ISTATE will be regarded as illegal a negative output value requires the user to change it and possibly other inputs before calling the solver again ATOL NM72 An option when using ADVANIS is the absolute tolerance The ATOL for ADVAN13 by default is 12 that is precision is 10 Usually the problem runs quickly when using ADVANIS with this setting On occasion however you may want to reduce ATOL usually set it equal to that of TOL and improve speeds of up to 3 to 4 fold ATOL may be set at the EST or COV command The absolute tolerance is set to the same ATOL for all compartments As of NM73 ATOL also acts on ADVAN O s differential equation solver where by default absolute significant digits accuracy absolute tolerance is 12 The relative tolerance for ADVANIS is still set by TOL by the SSUBROUTINES COV or TOL record just as it is for the other di
116. EXP ETA 3 ETA OCC_KA SOMEGA BLOCK 3 0 1 0 01 0 1 0 01 0 01 0 1 SOMEGA BLOCK 3 0 03 0 001 0 03 0 001 0 001 0 03 SOMEGA BLOCK 3 SAME 2 Repeat OMEGA BLOCK 3 SAME twice In the above example the NMTRAN parses the variable name OCC CL at the underscore and determines that there is a data item called OCC with which to associate the variable with the etas listed DO WHILE enhancement NM73 DOWHILE may now be used in all blocks of abbreviated code If a variable is used as a DOWHILE loop variable it must be declared SABBR DECLARE DOWHILE I Recursive random variables dowhile recursive variables may be computed in DOWHILE blocks as well as in ordinary abbreviated code A new example examples sumdosetn ctl uses DOWHILE for dose super imposition in a transit compartment and includes the following Sabbr declare dosetime 100 dose 100 Sabbr declare dowhile i Sabbr declare dowhile ndose SPK CALLFL 2 IF NEWIND 2 NDOSE 0 IF AMT gt 0 and cmt 1 THEN NDOSE NDOSE 1 dosetime NDOSE TIME DOSE NDOSE AMT ENDIF SDES INPT 0 nm730 doc 24 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 I 1 DOWHILE I NDOSE IPT 0 IF T gt dosetime I IPT DOSE I T dosetime 1 NN EXP KTR T dosetime I INPT INPT IPT I I 41 ENDDO See also ssaddl ctl ssonedose ctl and
117. EXP MU_1 ETA 1 K10 EXP MU_2 ETA 2 K12 EXP MU_3 ETA 3 K21 EXP MU_4 ETA 4 VM EXP MU_5 ETA 5 KMC EXP MU_6 ETA 6 KO3 EXP MU_7 ETA 7 K30 EXP MU_8 ETA 8 S3 VC Si Vc KM KMC S1 F3 K03 K30 DES DADT 1 DADT 2 DADT 3 K104K12 A 1 K21 A 2 VM A 1 A 3 A 1 KM K12 A 1 K21 A 2 VM A 1 A 3 A 1 KM K30 A 3 K03 ERROR CALLFL 0 ETYPE 1 IF CMT NE 1 ETYPE 0 IPRED F Y F F ETYPE EPS 1 F 1 0 ETYPE EPS 2 STHETA Initial Thetas 4 0 MU_1 2 1 MU 2 0 7 MU 3 0 17 MU 4 2 2 MU 5 0 14 MU_6 3 7 MU 7 70 7 MU 8 degrees of freedom for OMEGA prior 8 FIXED dfo Initial Omegas nm730 doc 196 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 OMEGA BLOCK 8 0 2 p 0 0043 0 2 p 0 0048 0 0023 0 2 p 0 0032 0 0059 0 0014 0 2 p 0 0029 0 002703 0 00026 0 0032 0 2 p 0 0025 0 00097 0 0024 0 00197 0 0080 0 2 p 0 0031 0 00571 0 0030 0 0074 0025 0034 2 p 00973 00862 0041 0046 00061 0 0056 0 0056 0 2 p ooooooo0o o Omega prior OMEGA BLOCK 8 FIX 0 oooooooo oooooooN oooooo ooooooN oooooo
118. GG IMTGG MTPTR K 1 1 2 What is new in NONMEM Version 7 2 0 versus NONMEM 7 1 2 The main new features of NONMEM 7 2 compared to NONMEM 7 1 2 are as follows Dynamic Memory Allocation No need to modify SIZES for unusually large problems Memory is automatically sized according to the number of parameters and number of subjects User may override computer generated values using a SIZES statement as the first executed line of the control stream Often for moderate sized problems this results in much smaller memory usage compared to the standard memory usage in NONMEM 7 1 Particularly helpful for parallel computing when using multiple cores on a single computer Please see section l 6 Dynamic Memory Allocation NM72 and 7 Changing the Size of NONMEM Buffers Parallel Computing The computation of a single problem that can take many hours or days may be distributed over two or more cores and or computers to complete in a shorter time After the primary installation of standard NONMEM described below parallel computing may require additional setup in order to implement which can be very specific to the operating system and Fortran compiler used In addition you may need assistance from your IT administrator Please read the installation notes below and Section 1 53 Parallel Computing NM72 MSF file system fully expanded to Monte Carlo Methods Seamless resumption of expectation maximization and Bayesian methods in case of sudden interruption
119. GL 3 With these options the algorithm sets up the following For forward finite difference h is set to SIGL 2 precision For central finite difference h is set to SIGL 3 precision For forward second order difference h is set to SIGL 3 precision The individual fits for evaluating optimal eta values will be maximized to a precision of the user supplied SIGL value Optimization of population parameters occurs until none of the parameters change by more than NSIG significant digits For the COV step the step size for evaluating the R matrix central difference second derivative is set to SIGL 4 which according to numerical analysis yields the optimal precision of SIGL 2 for the second derivative terms If only the S matrix is evaluated central difference first derivative then the step size for it is set to SIGL 3 But see COV Additional Options and Behavior for a way to set SIGL and TOL for COV distinct from the option for the SEST command If the user sets NSIG gt SIGL 3 and specifies SIGL then the optimization algorithm will do the following which is a less than optimal setup For forward finite difference h is set to NSIG precision For central finite difference h is set to NSIG precision For forward second order difference h is set to NSIG precision The individual fits for evaluating optimal eta values will be maximized to a precision of the user supplied SIGL value Optimization of population parameters occurs unt
120. ID and is expressed as follows for example examples Wuperid2 ctl SPK MU_1 THETA MU 2 THETA MU 3 THETA MU 4 THETA 4 CL DEXP MU_1 ETA 1 ETA 5 ETA 9 V1 DEXP MU_2 ETA 2 ETA 6 ETA 10 Q DEXP MU_3 ETA 3 ETA 7 ETA 11 V2 DEXP MU_4 ETA 4 ETA 8 ETA 12 S1 V1 1 2 3 SLEVEL SID 5 1 6 2 7 3 8 4 CID 9 5 10 6 11 7 12 8 Thus for clearance eta 9 is the country variability that has nested in it the site variability eta 5 which in turn has nested in it the subject variability the standard ID data eta 1 When performing FOCE with LEVEL you must use the SLOW option in EST and MATRIX R for the covariance step COV should be selected Nesting below the subject ID as for previous versions of NONMEM as shown for inter occasion variability example 7 The above method using LEVEL is a linearized approximation at the super ID level and takes advantage of a dual run for each OBJ function call freely allowing all etas to vary on the first run then averaging the SID etas fixing them to these averages and going through another run to allow the subject ID etas to be assessed This approximation method works very well for the EM and Monte Carlo methods and reasonably well for the FOCE Laplace methods To perform an exact analysis separate thetas must be defined for each value pertaining to a super ID data item so that theta is shared only by the subjects with the particu
121. IGNORE SSUBROUTINE ADVAN6 TOL 9 DP OTHER SDE f90 nde number of base equations ncmt number of observation compartments SABBR DECLARE SGW 3 need at least ncmt of these SMODEL COMP CENTRAL there are nde base states COMP DFDX1 need ncmt observation compartments COMP DPDT11 Will need ndet 1 nde 2 of these SPK IF NEWIND NE 2 OT 0 MU THETA 1 CL EXP MU_1 ETA 1 MU_2 THETA 2 VD EXP MU_2 ETA 2 SGW1 THETA 4 SDES FIRSTEM 1 DADT 1 CL VD A 1 NEXT DERIVATIVES ARE ACUALLY PREDICTIVE VALUES FOR COMPARTMENTS 1 AND 2 RESPECTIVELY Derivatives of these with respect to A will be calculated symbolically by DES routine created by NMTRAN DADT 2 A 1 VD DUMMY PLACEMENT FOR DERIVATIVES OF THE STOCHASTIC ERROR SYSTEM THESE ARE FILLED OUT BY SDE DER SGW 1 SGW1 the DA array THEN contains all derivatives of DADT DXDT with respect to A X number of base model derivative equations nde 1 Number of compartments nomt 1 DA is a reserved array dimensioned DA IR LAST CALL SDE_DER DADT A DA IR SGW 1 0d 00 1 0d 00 SERROR OBS ONLY IPRED A 1 VD IRES DV IPRED W THETA 3 IWRES IRES W WS 1000 0 CENTRAL COMPARTMENT PLASMA LEVELS EPS 1 USER MODEL ERROR CONTRIBUTION EPS 2 STOCHASTIC ERROR CONTRIBUTION THE WS IS JUST A PLACEHOLDER COEFFICIENT SDE CADD WILL REPLACE THIS WITH THE CORRECT VALUE X IPRED W EPS 1 WS EPS
122. IMI LIM2 LIM3 LIM11 LIMI3 LIMI5 and LIMI6 LIMI2 and LIM14 are not used Enclosing the option in quotes maxlim 1 2 3 11 16 is required for some operating systems For sizing MAXRECID use the number 17 Setting maxlim 1 17 is equivalent to maxlim 3 whereas maxlim 3 means to have NMTRAN size only LIM3 Description of Buffers A number of contiguous data records are stored in memory at any one time in buffers If a large enough memory area can be made available for this purpose then the entire data set can be stored in memory throughout the NONMEM run and computing costs can be decreased The following discussion of NONMEM buffers should not be confused with I O buffers which are used by the operating system The size of buffer 1 is related to the number LIMI of data records stored in memory at any one time A large proportion of data sets will consist of no more than 10000 data records Consequently the size of buffer 1 has been set to allow LIM1 10000 data records The least number of data records allowable must exceed the largest number of data records used with any one subject which rarely will be as large as 10000 Each data record consists of PD 8 byte double precision computer words and the allocation of memory for buffer 1 is PD LIM1 3 8 bytes nm730 doc 37 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Buffer 2 holds a number of contiguous residual records For each data record NONMEM generates
123. ITER 1000 T S S S S the IMP statement also has AUTO 1 However for the following example EST METHOD ITS AUTO 1 PRINT 10 EST METHOD SAEM AUTO 1 PRINT 50 EST METHOD IMP PRINT 1 EONLY 1 NITER 5 ISAMPLE 1000 AUTO 0 EST METHOD BAYES AUTO 1 FILE auto txt PRINT 100 NITER 1000 Xn xr xr our the AUTO setting is turned off for IMP and turned back on for BAYES Any option settings implicitly set by the AUTO feature does not transfer to the next EST statement Also when using AUTO 1 the transfer of any options settings explicitly set by the user from previous EST statements may or may not occur for those options set by the AUTO option depending on the situation nm730 doc 84 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 The mapping of parameters between S ADAPT and NONMEM is as follows S ADAPT NONMEM Pmethod 4 IMPMAP Pmethod 8 IMP Pmethod 1 ITS Pmethod 6 DIRECT Npopiter NITER Npopc ISAMPLE Npop MCETA optmethod OPTMAP covest ETADER Gefficiency IACCEPT Gamma min ISCALE MIN Gamma max ISACLE MAX DFRAN DF Popconv test CTYPE Popconv rows CITER Popconv alpha CALPHA Ndelpar MAPINTER Poperr type 3 COV MATRIX S Poperr type 8 COV MATRIX R Poperr_type 9 COV POPFINAL subroutine CONSTRAINT subroutine may be us
124. MA and OMEGA matrices are listed in lower triangular order row wise 1 2 2 456 78910 You may change the order in which these are displayed by specifying the ORDER option The THETAS are referenced with a T SIGMAS with S OMEGAS with O lower triangular with L upper triangular with U The first three letters given in the ORDER option refer to which parameters are listed in order T S O and the fourth letter is U or L to indicate matrix element order for sigmas and omegas Thus ORDER TSOL Is the default ordering This is different from the ordering that is given in the report file for displaying the variance matrix which is TOSU In TOSU ordering Thetas are listed first in the raw output file followed by omegas followed by sigmas and the omegas and sigma elements are listed in row wise upper triangular order or column wise lower triangular order nm730 doc 120 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 ON WW CON A 1 47 EST Additional Output Files Produced The following files are created automatically with root name based on the root name of the control stream file root cov Full variance covariance error matrix to thetas sigmas and omegas root cor Full correlation matrix to thetas sigmas and omegas root coi Full inverse covariance matrix Fischer information matrix to thetas sigmas and omegas root phi Individual phi parameters phi i mu i eta i for ith parameter and their var
125. MEM Users Guide Introduction to NONMEM 7 3 0 6 Non positive semidefinite and non singular means there is at least one negative eigenvalue and no zero valued eigenvalue Alternative diagnostic matrices may be outputted by NONMEM 7 Negative definite means there are only negative eigenvalues 8 Non negative definite means there is at least one eigenvalue that is greater than or equal to Zero NONMEM tests for conditions 1 5 and 6 and outputs appropriate result matrices or diagnostic matrices as it is able Alternative expressions would be unsuitable to describe the condition of the matrices For example non positive definite 2 does not mean the same as positive semi definite 3 Similarly non positive definite 2 is not exactly the same as non positive semidefinite 4 The set of non negative definite matrices 8 includes matrices that are positive definite 1 positive semi definite 3 and a subset of non positive semidefinite 4 not including those with all negative eigenvalues 1 43 Adding Nested Random Levels Above Subject ID NM73 Suppose you wish to model inter site variability or inter trial variability so that several subjects belong to a trial An easy albeit slightly approximate method would be to use the LEVEL feature Consider the following control stream fragment which in addition to inter subject variability eta 1 for clearance CL there is inter site variability eta 5 SPK MU_1 THETA 1 MU
126. METHOD IMP INTERACTION ISAMPLE 1000 NITER 50 MAPITER 0 Notice that since MAPITER O the first iteration of IMP method relies on starting parameters for its sampling density that came from the DIRECT sampling method 1 31 Some General Options and Notes Regarding EM and Monte Carlo Methods AUTOZ 0 default NM73 If option AUTO 1 is selected then several options will be set by NONMEM that will allow best settings to be determined The user may still over ride those options set by AUTO by specifying them on the same EST record For example EST METHOD ITS AUTO 1 PRINT 10 EST METHOD SAEM AUTO 1 PRINT 50 EST METHOD IMP PRINT 1 EONLY 1 NITER 5 ISAMPLE 1000 EST METHOD BAYES AUTO 1 NITER 1000 FILE auto txt PRINT 100 Minn Ww The settings of AUTO for each method are as follows nm730 doc 83 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 METHOD DIRECT INTERACTION ISAMPLE 1000 CTYPE 3 NITER 500 STDOBJ 10 ISAMPEND 10000 NOPRIOR 1 CITER 10 CINTERVAL 0 CALPHA 0 05 EONLY 0 METHOD BAYES INTERACTION CTYPE NOPRIOR 0 C 3 NITER 10000 NBURN 4000 I TER 10 CINTERVAL 0 CALPHA 0 05 IACCEPT 0 4 ISCALE MIN 1 0E 06 ISCALE MAX 1 0E 06 PACCEPT 0 5 PSCALE MIN 0 01 PSCALE MAX 1000 PSAMPLE Ml1
127. N 0 01 NM73 PSCALE MAX 1000 NM73 In MCMC sampling the scale factor used to vary the size of the variance of the proposal density population parameters theta sigma that are not Gibbs sampled in order to meet the PACCEPT condition is by default bounded by PSCALE MIN of 0 01 and PSCALE MAX 1000 This should left alone for MCMC sampling but on occasion there may be a reason to expand the boundaries perhaps to PSCALE_MIN 1 0e 06 PPAMPLE_MAX 1 0E 06 OSAMPLE_M1 1 defaults listed OSAMPLE M2 1 OACCEPT 0 5 These are the options for the MCMC Metropolis Hastings algorithm for OMEGA sampling If OSAMPLE_M1 lt 0 default then the OMEGA s are Gibbs sampled using the appropriate Wishart proposal density and the other options OSAMPLE M2 and OACCEPT are not relevant Otherwise for each iteration a matrix of OMEGAs are generated using a Wishart proposal density that has variance based on the previous samples done OSAMPLE MI times Next a matrix of OMEGAS are generated using a Wishart proposal density at the present OMEGA values postion and degrees of freedom dispersion factor for variances scaled to have samples accepted with OACCEPT frequency This is done OSAMPLE M2 times if OSAMPLE M2 0 then program performs this as many times as there are non fixed omega elements The final OMEGA matrix is kept Usually these options do not need to be changed from their default values listed above NOPRIOR 0 1 If prior information was
128. NONMEM 7 3 0 NOCOV 0 1 nm73 If covariance estimation is not desired for a particular estimation step set NOCOV 1 It may be turned on again for the next estimation step with NOCOV 0 If NOCOV 1 is set for an FOCE Laplace FO method this is equivalent to SCOV NOFCOV setting For ITS and IMP covariance estimation can take some time for large problems and you may wish to obtain only the objective function such as in the case of SEST METHOD IMP EONLY 1 after an SAEM estimation NOCOV has no effect on BAYES analysis as no extra time is required in assessing covariance for BAYES By default standard error information for the classical methods FO FOCE Laplace will be given only if they are the last estimation method even if NOCOV 0 for an intermediate estimation step If NOCOV 1 for the FOCE LAPLACE FO method and it is the last estimation step then standard error assessment for it will be turned off DERCONT 0 1 NM73 By the default value of the derivative continuity DERCONT is 0 When it equals 1 the partial derivative of the objective function with respect to thetas will perform an additional test to determine if a backward difference assessment is more accurate than a forward difference assessment The forward difference assessment can differ greatly from the backward difference assessment in cases of extreme discontinuity when varying certain thetas by even just a small amount in the model results in a large change in objective f
129. NONMEM USERS GUIDE INTRODUCTION TO NONMEM 7 3 0 Robert J Bauer ICON Development Solutions Hanover Maryland September 18 2014 Copyright of ICON Development Solutions Hanover MD 21076 2013 All rights reserved NONMEM Users Guide Introduction to NONMEM 7 3 0 TABLE OF CONTENTS 1 1 What is new in NONMEM Version 7 3 0 versus NONMEM 7 2 0 9 1 2 What is new in NONMEM Version 7 2 0 versus NONMEM 7 1 2 16 I 3 Introduction to NONMEM 7 and higher eese 18 1 4 Expansions on Abbreviated and Verbatim Code NM72 NM723 19 FORTRAN 95 Considerations aeeeieeres ere eeu oen oret tup t ee vn e sdoneuededaccavadcsnenevovedesvedaeverevorserassoesse 19 Continuation indicator is allowed in abbreviated code non verbatim lines NM73 21 Alternative Inputs for 0MEGA and SIGMA Values VARIANCE CORRELATION CHOLESKY NIVET2 sscccistsassncsseonctadsssecchiccaceesstussvucansosctanendessniscosusagsebsvanivaveseniddonsitivesstassianeeas 21 Repeated SAME BLOCK for 0MEGA and SIGMA Records NM73 22 Repeated Value Inputs for STHETA OMEGA and SIGMA NMY723 22 A BBR DECLARE feature for abbreviated code NM73 s ccscscssssssscssscssssscsscsseseseees 23 A BBR REPLACE feature for abbreviated code NM73 s cccccsscsssssssscssscsssscsssssescees 23 Easier Inter occasion variabil
130. NONMEM with MPI requires its share library libgfortran so 3 available for the worker process and in the path designated by the manager s LD LIBRARY PATH setting LD LIBRARY PATH SHOME gcc trunk lib SHOME libg SLD LIBRARY PATH export LD LIBRARY PATH where HOME gcc trunk lib is the library path for the manager s gfortran and HOME libef is the path on the worker computer containing at least the file libgfortran so 3 You may place these lines in the bashrc file Therefore if upon loading NONMEM on the worker computer a message is displayed indicating that certain share files are missing etc then you may need to either install gfortran or selectively make the share file available In addition the MPI system needs certain executable files available on the worker computer These are obtained from the bin directory of the MPICH2 system mpdlib py mpdman py mpd py Place these files in a directory on the worker computer that has the same path as MPICH2 is installed in the manager s computer For example if the manager s MPICH2 bin path is HOME MPICH2 LINUX mpich2 install bin then this should be where the worker computer s py files are Upon booting up before executing your first NONMEM run load up the mpi system mpdboot n number of computers f mpd hosts as instructed in the install guide The mpd hosts file contains a list of IP addresses one per line of the worker and manager computers They could be referenced
131. NOPRIOR 1 CTYPE 3 GRD TS 11 Results of ITS serve as initial parameters for the IMP method SEST METHOD IMP INTERACTION EONLY 0 MAPITER 0 NITER 100 ISAMPLE 300 PRINT 1 SIGL 8 The results of IMP are used as the initial values for the SAEM method SEST METHOD SAEM NBURN 3000 NITER 2000 PRINT 10 ISAMPLE 2 CTYPE 3 CITER 10 CALPHA 0 05 After the SAEM method obtain good estimates of the marginal density objective function along with good estimates of the standard errors SEST METHOD IMP INTERACTION EONLY 1 NITER 5 ISAMPLE 3000 PRINT 1 SIGL 8 SEED 123334 CTYPE 3 CITER 10 CALPHA 0 05 The Bayesian analysis is performed SEST METHOD BAYES INTERACTION FILE example2 TXT NBURN 10000 NITER 3000 PRINT 100 NOPRIOR 0 CTYPE 3 CITER 10 CALPHA 0 05 Just for old times sake lets see what the traditional FOCE method will give us And remember to introduce a new FILE so its results wont append to our Bayesian FILE EST METHOD COND INTERACTION MAXEVAL 9999 FILE example2 ext NSIG 2 SIGL 14 PRINT 5 NOABORT NOPRIOR 1 COV MATRIX R UNCONDITIONAL nm730 doc 190 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 67 Example 3 Population Mixture Problem in 1 Compartment model with Volume and rate constant parameters and their inter subject variances modeled from two sub populations Model Desc Population Mixture Problem in 1 Compartment model with Volume and rate constant parameters and their inter subject variances modeled from two sub popul
132. NPRED NRES NWRES These are non conditional no eta epsilon interaction pred res and wres values These are identical to those issued by NONMEM V as PRED RES and WRES PREDI RESI WRESI These are non conditional with eta epsilon interaction pred res and wres values These are identical to those issued by NONMEM VI as PRED RES and WRES The WRESI will not differ from NWRES if INTERACTION was not selected in the previous EST command CPRED CRES CWRES These are conditional no eta epsilon interaction pred res and wres values as described in 1 The conditional mode etas from FOCE or ITS also known as conditional parametric etas CPE empirical bayes estimates EBE posthoc estimates of etas or mode a posteriori MAP estimates or conditional mean etas from Monte Carlo EM methods will be referred to as 1 eta hat must be available from a previous EST MAXEVAL gt 0 command The conditional weighted residuals are estimated based on a linear Taylor series approximation that is extrapolated from the conditional mean or mode or posthoc eta estimates rather than about eta 0 CPRED f 8 g WN using the nomenclature of Guide I Section E2 Then CRES y CPRED The population variance covariance of observed data described in Guide I E 2 is also evaluated at eta hat C 1 CWRES C f y CPRED Because of the linear back extrapolation it is possible for some CPRED values to be negative User
133. OCK 2 3 p 01 f 3 p S OMEGA BLOCK 1 1 pl OMEGA BLOCK 1 SAME OMEGA BLOCK 1 SAME SIGMA 0 1 p Degrees of freedom for Prior Omega blocks S THETA 2 0 FIXED 1 0 FIXED Prior Omegas OMEGA BLOCK 2 14 FIX 0 0 125 OMEGA BLOCK 1 0164 FIX SOMEGA BLOCK 1 SAME SOMEGA BLOCK 1 SAME EST METHOD ITS INTERACTION FILE example7 ext NITER 10000 PRINT 5 NOABORT SIGL 8 CTYPE 3 CITER 10 NOPRIOR 1 CALPHA 0 05 NSIG 2 EST METHOD SAEM INTERACTION NBURN 30000 NITER 500 SIGL 8 ISAMPLE 2 PRINT 10 SEED 1556678 CTYPE 3 CITER 10 CALPHA 0 05 NOPRIOR 1 EST METHOD IMP INTERACTION EONLY 1 MAPITER 0 NITER 4 ISAMPLE 3000 PRINT 1 SIGL 10 NOPRIOR 1 EST METHOD BAYES INTERACTION FILE example7 txt NBURN 10000 NITER 10000 PRINT 100 CTYPE 3 CITER 10 CALPHA 0 05 NOPRIOR 0 EST METHOD COND INTERACTION MAXEVAL 9999 NSIG 3 SIGL 10 PRINT 5 NOABORT NOPRIOR 1 FILE example7 ext COV MATRIX R PRINT E UNCONDITIONAL nm730 doc 198 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 72 Example 8 Sample History of Individual Values in MCMC Bayesian Analysis Model Desc Two compartment Model Using ADVAN3 TRANS4 Project Name nm7examples Project ID NO PROJECT DESCRIPTION PROB RUN Example 8 from samp51 SINPUT C SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT CLX V1X QX V2X SDIX SDSX DATA example8 csv IGNORE C SUBROUTINES ADVAN3 TRANS4 PRIOR NWPRI NTHETA 4 NETA 4 NTHP 4 NETP 4
134. P 0 01 F 0 05 P 0 01 F 0 01 F 0 05 P 0 01 F 0 01 F 0 01 F 0 05 P Initial value of SI SIGMA 0 6 P GMA EST METHOD SAEM INTERACTION FILE example9 ext NBURN 5000 NITER 500 PRINT 10 NOABORT SIGL 6 CTYPE 3 CINTERVAL 100 CITER 10 CALPHA 0 05 File Aneal f90 SUBROUTINE CONSTRAINT T HETAS NTHETAS SIGMA2 NSIGMAS OMEGA NOMEGAS ITER_NO USE SIZES ONLY ISIZE DPSIZE INCLUDE nm TOTAL INC INTEGER KIND ISIZE NTHETAS NSIGMAS NOMEGAS ITER NO INTEGER I J ITER OLD DATA ITER OLD 1 REAL KIND DPSIZE OMEGA MAXOMEG MAXOMEG THETAS MAXPTHETA SIGMA2 MAXPTHETA REAL KIND DPSIZE OMEGO MAXOMEG SAVE bum m mm nL E IF SAEM MODE 1 AND IMP MODE 0 AND ITS MODE 0 AND ITER NO 200 THEN IF ITER NO ITER OLD OR ITER NO 0 THEN During burn in phase of SAEM and when a new iteration occurs iter old iter no store the present diagonals of omegas ITER OLD ITER NO DO I 1 NOMEGAS OMEGO I OMEGA I I nm730 doc 203 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 ENDDO ENDIF IF ITER NO 0 THEN DO I 1 NOMEGAS Use whatever algorithm needed to slow down the reduction of Omega The expansion of Omega should be less with each iteration OMEGA I I OMEGO I 1 0D 00 10 0D 00 ITER_NO ENDDO ENDIF ENDIF RETURN
135. P EXP CLR 1 0 CLR V1R 1 0 V1R QOR 1 0 QQR V2R 1 0 V2R 1 58811E 00 28 12694E 01 2 37435E 00 1 0 FIXED 1 0 FIXED 1 0 FIXED METHOD ITS INTERACTION MAXEVAL 9999 PRINT 5 NOHABORT SIGL 9 CTYPE 3 NITER 200 NONINFETA 1 MCETA 10 EST METHOD IMP INTERACTION MAXEVAL 9999 PRINT 1 NOHABORT ISAMPLE 3000 NITER 200 SIGL 9 DF 2 RANMETHOD 3S1P CTYPE 3 MCETA 10 EST METHOD 1 INTERACTION MAXEVAL 9999 PRINT 1 NOHABORT NSIG 3 SIGL 9 NONINFETA 1 SLOW MCETA 30 COV MATRIX R UNCONDITIONAL Note that constructions such as CL EXP MU_1 ETA 1 SQRT EXP CLR 1 0 CLR violate the strict MU_x ETA x rule recommended for EM analysis because the term nm730 doc 114 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SQRT EXP CLR 1 0 CLR is multiplied by ETA 1 Nonetheless for this example the importance sampling works quite well Note also that SORT EXP CLR 1 0 CLR approaches 1 as NU approaches infinity and therefore the random effect of CL approaches normality 1 45 Format of NONMEM Report File The format of the NONMEM report file has been slightly modified with improvements to allow third party software to more easily identify portions of the result file As described above the user has now the ability to request a series of classical or new estimation methods within the same problem if he so chooses Each of the new methods produces slightly different banner text and termination
136. R 1 EST METHOD BAYES INTERACTION NBURN 2000 NITER 5000 PRINT 10 FILE example4 txt SIGL 6 NOPRIOR 0 EST MAXEVAL 9999 NSIG 3 SIGL 12 PRINT 1 METHOD CONDITIONAL INTERACTION NOABORT FILE example4 ext NOPRIOR 1 COV MATRIX R UNCONDITIONAL SIGL 10 nm730 doc 194 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 69 Example 5 Population Mixture Problem in 1 Compartment model with rate constant parameter mean modeled for two sub populations but its inter subject variance is the same in both sub populations Model Desc Population Mixture Problem in 1 Compartment model with rate constant parameter mean modeled for two sub populations but its inter subject variance is the same in both sub populations Project Name nm7examples Project ID NO PROJECT DESCRIPTION PROB RUN example5 from adltrinm4t SINPUT C SET ID JID TIME CONC DV DOSE AMT RATE EVID MDV CMT VC1 K101 VC2 K102 SIGZ PROB DATA example5 csv IGNORE C S SUBROUTINES ADVAN1 TRANS1 MIX P 1 THETA 4 P 2 21 0 THETA 4 NSPOP 2 PK Q 1 IF MIXNUM EQ 2 Q 0 MU_1 THETA 1 Note that MU 2 can be modeled as THETA 2 or THETA 3 depending on the MIXNUM value Also we are avoiding IF THEN blocks MU 2 Q THETA 2 1 0 Q THETA 3 V DEXP MU 1 ETA 1 K DEXP MU_2 ETA 2 S1 V SERROR Y F F EPS 1 STHETA 1000 0 4 3 1000 0 MU 1 1000 0 2 9 1000 0 MU 2 1 1000 0 0 67 1000 0 MU 2 2 0 0001 0 667 0 9999 P 1
137. RACTION e eeeee eere ee eoe teen eoe te ense etn ea see tn ease nea 70 EST METHOD IMP INTERACTION MAPITER 1 MAPINTER7CI 70 1 27 Stochastic Approximation Expectation Maximization SAEM Method 70 EST METHOD SAEM INTERACTION eee eee teen eee to enses ene tn ete en sene en ense een eE aeneae 71 NBURNS200 ict onu toS R Erb EE ta EEER DIE EE CUOMS EN UI ARES IURE eat 71 NSAMPLE NITER 1000 em POTINS 71 ISAMPLE 2 defa lts listed creto isa Se ever kt vto ee PV MEM OP Erie Len Sio 71 ISAMPLE M D EAE E on do Te nive e iR Ve eh Detektei ins urat E EEEE 71 ISAMPLE M1A 0 NW7A2 4 etenccooke stans GHen iedea coi erase pn oet sd ot deed cox ondes Tes tee sa mess dodo s on tuPuade 71 ISAMPEE M22 oieshid usub nh uso E po toes DURER YET E REESE M DTE RUE RC RAV CEA rop LUE dus 71 ISAMPLE M3232 ewe oa b tbe ova oai Ui ei ei e edi vk e DR VEL CPU Un ER e aeo keh ava 71 EX i Oi ud A PETRI EN TR 71 ISAMPENDSi NM73 o uecsceps s rao dwxes Cete Pudor Auer RA iR Cope et Ud Sc Cottae fuia aT odose oir siese 72 ISCALE_MIN 1 0E 06 defaults for SAEM BAYES NM72 e eee ee ee ee eee ene enetnee 72 ISCALE_MAX 1 0E 06 NM icaneeasstoecte eps ee ba icoPertb ecu e eee endec ut dn pu raa eoo eH oe ues etae duse de uns 72 NOCOV 0 1 0mM73 o o 2 eeoduid eet ona pi cob o lat te OCC IA SP uD Ge DEORUM sosea eiS soe us 73 DERCONT 0 RET a E oec nr 73 CONSTRAINST NMT2 Gioiniicbatieb
138. RES PRED RES WRES NPDE PDERR ECWRES NOPRINT NOAPPEND FILE myfile tab ESAMPLE 1000 SEED 1233344 LFORMAT RFORMAT NM72 An alternative format description to FORMAT is RFORMAT and LFORMAT RFORMAT where R real numbers describes the full numeric record of a table so that formats for specific columns may be specified LFORMAT where L label specifies the format of the full label record of a table The formats must be enclosed in double quotes and and have valid Fortran format specifiers The RFORMAT and LFORMAT options can be repeated if the format specification is longer than 80 characters Multiple RFORMAT and LFORMAT entries will be concatenated to form a single format record specification For example LFORMAT 4X A4 4 4X A8 RFORMAT F8 0 RFORMAT 4 1PE12 5 Will result in the following formats submitted to a Fortran write statement LFORMAT 4X A4 4 4X A8 for the table s label record and RFORMAT F8 0 4 1PE12 5 For the table s numeric records If RFORMAT and LFORMAT are given then the FORMAT option will be ignored By default FORMAT RFORMAT LFORMAT specifications will be passed on to the next STABLE record in a given problem unless new ones are given To turn off an RFORMAT LFORMAT specification in a subsequent table and therefore use FORMAT instead set LFORMAT NONE
139. T CNT P1 P2 REAL KIND DPSIZE ONE TWO W DATA ONE TWO 1 00D 00 2 00D 00 SAVE IF I LE 1 RETURN W Y 1 Y 1 Y 1 THETA 3 ONE THETA 3 CALL CELS CNT P1 P2 IER1 IER2 Y 1 W CNT CNT TWO THETA 3 ONE LOG Y 1 RETURN END Continuation indicator is allowed in abbreviated code non verbatim lines NM73 In NONMEM 7 3 0 extra long lines may be continued using an amp at the end of the line CL EXP THETA 1 WERT amp EPS 1 The total number of characters in the resulting concatenated line may not exceed FSD default set to 67000 in sizes f90 In fact the continuation marker amp may be used on record lines as well If the ampersand at the end of a line is not to be interpreted as a continuation marker but as a part of the record then place a after it For example FORMAT s1PE15 8 160 amp Alternative Inputs for 0MEGA and SIGMA Values VARIANCE CORRELATION CHOLESKY NM72 In NONMEM 7 2 0 OMEGA and SIGMA elements may be entered in forms other than the default variance diagonal elements and covariance off diagonal elements Diagonal elements may also be entered as standard deviation and off diagonal elements may be entered as correlation values Options are VARIANCE STANDARD to indicate form of diagonal elements COVARIANCE CORRELATION to indicate form of off diagonal elements CHOLESKY for inputting blocks of OMEGAS or SIGMAS in their Cholesky f
140. T SIGMA but see below for the option DFS as of NM73 If CTYPE 1 then regardless of lower and upper bound designations on the THETA statements all thetas are uniformly varied using the IACCEPT factor If CTYPE 2 then the random values of theta are created based on a normal distribution with the initial THETA in the control stream file as the mean and the second set of OMEGAs as the variance if there is a PRIOR command with NTHP non zero This is the best way and most complete way to define the sampling density for the THETAs Otherwise if NTHP 0 the variance for THETA is obtained from the first set of OMEGA and requires that the THETA s be MU modeled and those THETAs not MU modeled will be varied by the uniform distribution method as described for CTYPE 0 The omega values are sampled using a Wishart density of variance listed in the 0MEGA command and DF is the degrees of freedom for randomly creating the OMEGAS If DF 0 then the dimensionality of the entire OMEGA matrix is used as the degrees of freedom As of NM73 if DF gt one million then OMEGA elements are fixed at their initial values The format of the chain file that is created is exactly the same as the raw output files including iteration numbers In the above example after the 5 random samples are made ISAMPLE 3 the third randomly created sample is selected and brought in as the initial values If ISAMPLE 0 then the initial values are not set to any of the randomly g
141. THE FLAG 0 IPRED SD ERR 1 FU Y ELSE F FLAG CUMD MDVRES 1 ENDIF SIGMA 1 0 FIXED STHETA 2 3 4 2 0 3 When performing an EM analysis such as importance sampling remember to designate the THETA that serves as the residual coefficient as a sigma like parameter by setting GRD appropriately SEST METHOD IMP LAPLACE INTERACTION CTYPE 3 NOHABORT GRD TS 3 PRINT 1 If you are using SIGMA instead then code as follows SERROR IPRED F SD SORT SIGMA 1 1 IPRED LOQ 0 1 DUM LOQ IPRED SD CUMD PHI DUM 1 0E 30 IF DV gt LOQ THEN nm730 doc 102 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 FLAG 0 IPRED IPRED EPS 1 ELSE F FLAG 1 Y CUMD MDVRES 1 ENDIF STHETA 2 3 4 2 SSIGMA 0 1 In this case the SIGMA is not being used purely as a scale parameter in a normal density variance matrix but is also being used as a parameter in another distribution the integrated normal density When using an EM or Bayes method it is best to indicate that this SIGMA should not be estimated using the usual analytical method for calculating SIGMA derivatives but using numerical derivatives by designating the GRD appropriately SEST METHOD IMP LAPLACE INTERACTION CTYPE 3 NOHABORT GRD SN 1 PRINT 1 1 40 ANNEAL to facilit
142. Users Guide Introduction to NONMEM 7 3 0 setting up a problem for the new EM methods you should start out with some trial runs and a limited number of iterations and observe its behavior Here are some starting points for the various methods EST METHOD ITS NITER 100 EST METHOD SAEM NBURN S500 NITER 500 EST METHOD IMP NITER 100 ISAMPLE 300 The convergence tests should not be used during trial runs The convergence tests for the EM methods can be fooled into running excessively long or ending the problem prematurely For example the iterations of SAEM are Markov chain dependent and therefore certain parameters may meander slowly The convergence tester if CITER and CINTERVAL are not properly set to span these meanderings may never detect stationarity for all the parameters and therefore may never conclude the analysis For IMP the parameters between iterations are less statistically correlated and the convergence tester is a little more reliable for it NMTRAN does some checking of MU statements If you wish to turn this off checking mu statements can take a long time for very large control stream files then include the NOCHECKMU option on the ABBR record ABBR NOCHECKMU MUM MMNNMD These options allow the MU reference equations for each theta to be optionally used or not used By default if a theta parameter is MU referenced it will be used to facilitate theta parameter estimation However the user may turn off spec
143. VI2 1 for OBJ2 to include MDV gt 100 records MDVI3 1 for OBJ3 to include MDV gt 100 records in a PK or PRED section for example SPK include nonmem_reserved_general MDVI1 1 MDVI2 1 MDVI3 1 1 58 table_compare Utility Program NM72 The utility program table_compare will compare the numerical values between two table files produced by the NONMEM TABLE record and the user may specify the tolerance for the comparison The syntax is table compare mytablel tab mytable2 tab myprecision xtl gt mydifferences txt where delimiter is t s for comma tab space and myprecision xtl is a precision specification or control file Default delimiter is space and default control file is table_compare xtl table compare mytablel tab mytable2 tab S myprecision xtl gt mydifferences txt In the above example the first file is comma delimited and the second one is space S delimited If a second character is given to a delimiter then this is for detecting a continuation marker at the end of a line that is to be continued If a third character is given as a delimiter this for detecting a continuation marker at the beginning of the continuing line Some examples are table compare mytablel tab mytable2 tab amp S amp myprecision xtl mydifferences txt nm730 doc 167 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 double quotes may be needed for DOS commands In the above example the first file is
144. _3 THETA 9 MU 4 THETA 10 CL DEXP MU_1 ETA 1 V1 DEXP MU_2 ETA 2 Q DEXP MU_3 ETA 3 V2 DEXP MU_4 ETA 4 S1 V1 SERROR CALLFL 0 Option to model the residual error coefficient in THETA 11 rather than in SIGMA SDSL THETA 11 W F SDSL Y F W EPS 1 IPRED F IWRES DV F W nm730 doc 189 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Initial THETAs STHETA 7 LCLM 7 LCLF CLAM CLAF LV1M LV1F V1AM V14F MU 3 0 7 MU 4 0 3 SDSL ON NOON NI oO uoouuo RAR AR ARR ARR we se Initial OMEGAs OMEGA BLOCK 4 0 5 I pl 0 001 f 0 5 pl 0 001 f 0 001 f 0 5 pl 0 001 f 0 001 f 0 001 f 0 5 pl Degrees of freedom to OMEGA prior matrix STHETA 4 FIX Prior OMEGA matrix SOMEGA BLOCK 4 0 01 FIX 0 0 0 01 0 0 0 0 0 01 0 0 0 0 0 0 0 01 SIGMA is 1 0 fixed serves as unscaled variance for EPS 1 THETA 11 takes up the residual error scaling SIGMA 1 0 FIXED The first analysis is iterative two stage Note that the GRD Specification of GRD is that theta 11 is a Sigma like parameter This will allow NONMEM to make efficient gradient evaluations for THETA 11 which is useful for later IMP IMPMAP and SAEM methods but has no impact on ITS and BAYES methods SEST METHOD ITS INTERACTION FILE example2 ext NITER 1000 NSIG 2 PRINT 5 NOABORT SIGL 8
145. a of the previous study For an OMEGA block use the smallest DF calculated among the OMEGA diagonal estimates in that block A similar formula would apply for SIGMA priors with the proviso that the DF be no larger than the total number of data points that apply for that sigma in the previous study for example if there are two sigmas one for PK data and another for PD data then the sigma for PK data gets no more than total number of PK data points in the previous study 1 30 Monte Carlo Direct Sampling NM72 On rare occasions direct Monte Carlo sampling may desired This method is the purest method for performing expectation maximization in that it creates completely independent samples unlike MCMC and there is no chance of causing bias if the sampling density is not similar enough to the conditional density unlike IMP However it is very inefficient requiring ISAMPLE values of 10000 to 300000 to properly estimate the problem The method can be implemented by issuing a command such as EST METHOD DIRECT INTERACTION ISAMPLE 10000 NITER 50 On occasion it can have some use in jump starting an importance sampling method especially if the first iteration of importance sampling fails because it relies on MAP estimation and the problem is too unstable for it Thus one could perform the following where just a few iterations of direct sampling begin the estimation process EST METHOD DIRECT INTERACTION ISAMPLE 10000 NITER 3 EST
146. a root ext file where root is base name of control stream file which may also be monitored by a text editor during the run If you run NONMEM from PDx POP Bayesian sample histories of the population parameters can be viewed after analysis is done The sample history file is written to that specified by the EST FILE option which can be also monitored by a text editor during or after the run Sometimes NONMEM does not respond to user input This may occur during a parallel distribution run using MPI or if the user began NONMEM with the background switch The user may open another console window copy the program sig exe from the NONMEM installed util directory to your run directory then enter any one of these commands Print toggle monitor estimation progress Sig J Sig R Sig P Paraprint toggle monitor parallel processing traffic Sig B Sig A Sig PA Sig PP Next move on to next estimation mode or next estimation sig K sig N Stop end the present run cleanly Sig E Sig S Subject print toggle sig T sig U sig SU Alternatively you may execute the sig program from another directory if you specify the run directory in which you want the signal file created nm730 doc 41 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 sig next nonmem run Make sure you terminate the directory name with a directory parse symbol appropriate for the operating system 1 12 COV Unconditional Evaluation T
147. age to perform the estimation using the original data file followed by table generation using the enhanced data file The FNLETA 2 setting comes in handy for this purpose SPROB RUN example6 from r2compl SINPUT C SET ID JID TIME DV CONC DOSE AMT RATE EVID MDV CMT SDATA example6 csv IGNORE C original data file used SSUBROUTINES ADVAN13 TRANS1 TOL 4 SMODEL NCOMPARTMENTS 3 SP SDES SERROR CALLFL 0 ETYPE 1 IF CMT NE 1 ETYPE 0 IPRED F Y F F ETYPE EPS 1 F 1 0 ETYPE EPS 2 SEST METHOD ITS INTERACTION SIGL 4 NITER 25 PRINT 1 FILE example6 ext NOABORT MSFO example6 msf ATOL 4 FNLETA 0 SPROB RUN example6 from r2compl SINPUT C SET ID JID TIME DV CONC DOSE AMT RATE EVID MDV CMT SDATA example6b csv IGNORE C enchanced data file SMSFI example6 msf SEST METHOD 1 FNLETA 2 ATOL 4 Because FNLETA 2 no estimation us actually done The etas loaded from the MSF file are used without modification to compute individual model parameters Since no analysis is performed setting METHOD 1 is sufficient regardless of what method was used in the earlier analysis Because ATOL 4 in the previous analysis good idea to retain this setting to yield identical evaluations from the differential equation solver STABLE ID TIME CONC IPRED CMT MDV EVID NOAPPEND NOPRINT FILE example6b fin FORMAT 1PE12 5 ONEHEADER nm730 doc 176 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 As of NM73 if an MDV is set t
148. an alternative number of nodes on the command line nmfe73 mycontrol ctl mresults res parafile fpiwini8 pnm nodes 4 in which case the first 4 nodes or node numbers 1 2 3 4 listed in COMMANDS and DIRECTORIES would be executed To also make distinct commands easy to write when launching many processes number list substitution can also be performed For example GENERAL NODES 8 PARSE TYPE 4 PARSE NUM 200 TIMEOUTI 600 TIMEOUT 1000 PARAPRINT 0 TRANSFER TYPE 1 NAMES Give a name to each node which is displayed 1 MANAGER 2 8 WORKER 10 17 COMMANDS each node gets a command line used to launch the node session cd refers to current directory Beyond the first position a will not be interpreted as a comment for commands l mpiexec wdir cd hosts 1 localhost 1 nonmem exe 2 8 wdir cd wk 1 hosts 1 localhost 1 nonmem exe nm730 doc 141 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 DIRECTORIES 1 NONE FIRST DIRECTORY IS THE COMMON DIRECTORY 2 8 wk 1 NEXT SET ARE THE WORKER directories In the above example the name of processes 2 through 8 are given as 2 8 WORKER 10 16 In this case each number represented in the list within the braces is expanded and matched with the process number so this line is equivalent to 2 WORKER10 WORKER11 WORKER12 WORKER13 WORKER14 WORKER15 WORKER16 0 10Y 01 BW Make sure that the number of items
149. aplace by either NMTRAN or the user which makes it easier to write user supplied subroutines Particularly useful for general stochastic differential equation analysis See OPTMAP and ETADER in section 1 20 Options for ESTIMATION Record for alternative MAP eta optimization methods and evaluating individual variances by numerical derivative methods for FOCE Laplace NM73 Conditional Individual Weighted Residual CIWRES added to residual variance diagnostics While CIWRES for uncorrelated data is readily evaluated as DV iPRED W CIWRES provides a proper individual weighted residual for L2 correlated data as well which requires more extensive linear algebraic calculation Furthermore individual predicted and individual residual values what are typically designated as IPRED and IRES and has often been inserted by hand into the control stream by users is now assessed by NONMEM called CIPRED and CIRES respectively and can be requested in the STABLES record See section 1 13 TABLE Additional Statistical Diagnostics Associated Parameters and Output Format A range of Etas may be requested to be outputted Instead of requesting for each eta to be outputted in a STABLE record as ETA1 ETA2 ETA3 etc a range of etas using the format of ETAS x y may be requested See 1 13 TABLE Additional Statistical Diagnostics Associated Parameters and Output Format Boot strap simulations to be performed in NONMEM See section 1 21 Bootstrap Select
150. arameter and thetas 6 12 are to be MU modeled MUMEN 3 5 8 X 2 M 6 12 Thetas not specified are given a default D designation GRD GNGNNND By default if a theta parameter has a Mu associated with it and its relationship to its Mu is sufficiently linear the program tests this by evaluating the partial second derivative of MU with respect to theta then the program will use Gibbs sampling for that parameter However for Mu modeled parameters the user can over ride these decisions made by the program and force a given parameter to be Gibbs sampled G or Metropolis Hastings sampled N In the above example thetas 1 and 3 are to be Gibbs sampled and the other thetas are M H sampled If the parameter is not Mu modeled or its Mu modeling is turned off by an MUM option setting the program performs an M H sampling D for default specifies you want the program to decide whether to use Gibbs sampling For SIGMA parameters if a particular SIGMA is associated with only one data point type and conversely the data point type has only that one SIGMA parameter defining its residual error and that data point type is not linked by an L2 item with any other data point types then that SIGMA will by default be Gibbs sampled with a chi square distribution Otherwise that SIGMA parameter will be sampled by Metropolis Hastings You can force Meroplis Hastings by specifying an N The first m letters of GRD refer to the m THETA s Then the m 1th
151. ariance assessment performed and a final FOCE analysis performed but did not want the program to spend time on standard error assessments for FOCE which can take a long time relative to the other methods RESUME NM73 If an MSFOzmsffile specification was made in the SEST step and analysis was interrupted during the COV step for the FO FOCE Laplace method then the COV step may be resumed where it was interrupted by executing another control stream file that uses the MSFI record specifying the MSFO file of the interrupted analysis and the RESUME option is entered at the COV record MSFl msffile COV RESUME 1 42 A Note on Covariance Diagnostics There are several conditions that can occur in assessing the variance covariance matrix of the estimates which are best defined according to eigenvalues that it detects in them 1 Positive definite means there are only positive eigenvalues NONMEM outputs proper variance variance matrices 2 Non positive definite means there is at least one eigenvalue that is less than or equal to zero 3 Positive semidefinite means there are no negative eigenvalues but at least one zero valued eigenvalue singular 4 Non positive semidefinite means there is at least one negative eigenvalue 5 Non positive semidefinite and singular means there is at least one negative eigenvalue and at least one zero valued eigenvalue Non inverted matrices may be outputted by NONMEM nm730 doc 106 of 210 NON
152. ata records for which DV j 0 or nearly so for which are modeled IPREDjj theta 3 TYPE 1 j i for i 1 to 16 SID values and j l to 3 TYPE values NONMEM thus adds for each TYPE gt 0 data record objective function value terms DV IPRED X DV IPRED that evaluates to 6 0 and the control stream file places a dependency of the last 0 0f each element that is each of the three Nsip TYPE s such that gt 0 0 The L2 data item allows NONMEM to assess correlation hence i l off diagonal elements to the SIGMA block between the three TYPEs within a given SID Thus for the added data portion NONMEM sees 16 subjects one for each of the SID values each of which have 3 data points one for each PK parameter TYPE The above problem can alternatively be coded more easily using the LEVELS mapping of etas as follows example superid3_1 without needing to add pseudo data to the data file SPROB RUN SINPUT C ID TIME DV AMT RATE EVID MDV CMT ROWNUM SID SDATA superid3 csv IGNORE C SSUBROUTINES ADVAN2 TRANS2 SPK MU_1 THETA 1 MU 2 THETA 2 MU 3 THETA 3 KA DEXP MU_1 ETA 1 ETA 4 CL DEXP MU_2 ETA 2 ETA 5 V DEXP MU_3 ETA 3 ETA 6 S2 V SERROR IPRE F PRE IPRE EPS 1 H Initial values of THETA STHETA 0 2 4 2 ITIAL values of OMEGA SOMEGA BLOCK 3 QT 0 001 0 0 001 0 001 0 1 SOMEGA BLOCK 3 Inter SID variance 0 3 0 001 0 3 0 001 0 001 0 3 In
153. ate EM search methods NM73 Syntax ANNEAL number list1 valuel number list2 value2 etc for as many lists that are needed Example SANNEAL 1 3 5 0 3 6 7 1 0 Sets starting diagonal Omega values for purposes of simulated annealing Thus initial values of OMEGA 1 1 OMEGA 2 2 OMEGA 3 3 and OMEGA 5 5 are set to 0 3 while initial OMEGA 6 6 and OMEGA 7 7 are set to 1 0 When SEST CONSTRAIN gt 4 an algorithm in constraint f90 will initially set the omegas to these values and then shrink these OMEGA values more and more with each iteration and eventually shrinks the OMEGA s to 0 the intended target value for that Omega This is a technique that may be used especially with SAEM to provide an annealing method for moving thetas that have 0 omega values associated with them The default is the use of gradient methods which are good for problems starting near the solution whereas the annealing method is more suitable for problems starting far from the solution An example is anneal ctl an EMAX model in which the Hill coefficient does not have inter subject variance that is its omega variance is set to 0 SPROB Emax model with hill 3 SINPUT ID DOSE DV SDATA anneal dat IGNORE SPRED MU 1 THETA 1 EMAX EXP MU_1 ETA 1 MU 2 THETA 2 ED50 EXP MU 2 ETA 2 MU 3 THETA 4 EO EXP MU_3 ETA 3 MU_4 THETA 3 HILL EXP MU_4 ETA 4 nm730 doc 103 of 210 NONMEM Users Guide Introduction
154. ations Project Name nm7examples Project ID NO PROJECT DESCRIPTION PROB RUN example3 from adltrim2s SINPUT C SET ID JID TIME CONC DV DOSE AMT RATE EVID MDV CMT VC1 K101 VC2 K102 SIGZ PROB DATA example3 csv IGNORE C S SUBROUTINES ADVAN1 TRANS1 The mixture model uses THETA 5 as the mixture proportion parameter defining the proportion of subjects in sub population 1 P 1 and in sub population 2 P 2 MIX P 1 THETA 5 P 2 1 0 THETA 5 NSPOP 2 Prior information setup for OMEGAS only SPRIOR NWPRI NTHETA 5 NETA 4 NTHP 0 NETP 4 NPEXP 1 PK The MUs should always be unconditionally defined that is they should never be defined in IF THEN blocks THETA 1 models the Volume of sub population 1 MU 1 THETA 1 THETA 2 models the clearance of sub population 1 MU 2 THETA 2 THETA 3 models the Volume of sub population 2 MU 3 THETA 3 THETA 4 models the clearance of sub population 2 MU 4 THETA 4 VCM DEXP MU_1 ETA 1 K10M DEXP MU_2 ETA 2 VCF DEXP MU_3 ETA 3 K10F DEXP MU_4 ETA 4 Q 1 IF MIXNUM EQ 2 Q 0 V Q VCM 1 0 Q VCF K Q K10M 1 0 Q K10F S1 V ERROR Y F F EPS 1 Initial THETAs STHETA 1000 0 4 3 1000 0 MU 1 1000 0 2 9 1000 0 MU 2 1000 0 4 3 1000 0 MU 3 1000 0 0 67 1000 0 MU 4 0 0001 0 667 0 9999 P 1 Initial OMEGA block 1 for sub population 1 OMEGA BLOCK 2 04 p l 01 f 027 pl
155. cated if real If not a number then column name in the data set containing NEVAL value If NEVAL 1 then you wish to interpolate covariate values in the original data set but not add any additional records TDELTA Alternative to entering NEVAL the increment in time may be entered If not a number then the column name in the original data set containing the TDELTA is used TSTOP stop time real number or integer for creating incremental time records IF TSTOP is not specified then default is LAST and the last record of the subject or occasion or time section is used If TSTOP is not a number and is not LAST then it is assumed be the column name in the original data set containing the stop time FILE output data file name to contain original data records interspersed with incremental time records AXIS Name of column containing times usually TIME Optionally designate LIN or LOG in parenthesis to indicate linear or geometric time incrementing If LIN additive time increment tstop tstart neval 1 If LOG multiplicative time increment tstop tstart 1 neval 1 DELIM delimiter of output data file if it is to be different from the input data file DLEIM S is space DELIM t is tab ITEM number list of values for data item ITEM for which there is to be a record at each time increment This can be done for a series of data items For example if you enter SFINEDATA CMT 1 3 EVID 2 2 then two records per time point are inse
156. ch begins with a short iterative two stage run to provide good initial eta estimates for each subject followed by the SAEM analysis which uses these initial eta estimates as a starting point for its Markov Chain Monte Carlo scan of each subject s conditional posterior density followed by objective function evaluation EST METHOD ITS INTERACTION NITER 5 EST METHOD SAEM NBURN 1000 ISAMPLE 2 NITER 1000 EST METHOD IMP EONLY 1 ISAMPLE 1000 NITER 5 MAPITER 0 Values of NBURN NITER and ISAMPLE may be changed as needed If you want conditional mean values values listed in root phi evaluated by MCMC sampling used in the SAEM method but at a constant set of the final fixed parameters then you could invoke EONLY 1 with the SAEM method as well EST METHOD ITS INTERACTION NITER 5 EST METHOD SAEM NBURN 1000 ISAMPLE 2 NITER 1000 EST METHOD SAEM EONLY 1 NBURN 200 ISAMPLE 2 NITER 1000 EST METHOD IMP EONLY 1 ISAMPLE 1000 NITER 5 MAPITER 0 1 28 Full Markov Chain Monte Carlo MCMC Bayesian Analysis Method The goal of the MCMC Bayesian analysis 11 12 is not to obtain the most likely thetas sigmas and omegas but to obtain a large sample set of probable population parameters usually 10000 30000 The samples are not statistically independent but when analysis is properly performed they are uncorrelated overall Various summary statistics of the population parameters may then be obtained such as means standard deviations and even confid
157. ch the NONMEM executable will be constructed To anticipate large sizes without needing to specify values in SIZES then set LTH LVR PD PC DIMTMP MMX DIMCNS and or PDT in sizes f90 to the maximum you think you will ever need NTMRAN will still create a NONMEN executable that is sized to fit the problem Be aware however that if parameter values are set too large NMTRAN may not run as it uses sizes 90 to set its array sizes at the beginning before it knows the actual size of the problem As of NM73 as an alternative to modifying sizes f90 to very large maximum sizes you can tell NMTRAN the maximum size that may be needed by specifying a SIZES parameter as a negative value Thus a user can give NMTRAN permission to deal with all problems that have data input files that have up to 1000 data items and up to 150 omegas and up to 200 thetas by the following SIZES PD 1000 LVR 150 LTH 200 but the size of these parameters when the NONMEM executable is constructed will be only what is needed for the particular problem In contrast SIZES PD 1000 LVR 150 LTH 200 will result in sizing the NONMEM executable with these values and won t make a tailor fit This would result in a very large executable regardless of the model size Thus SIZES PD 1000 tells NMTRAN that you may need as many as 1000 data items in a data file whereas SIZES PD 1000 tells NMTRAN that you need exactly that size With nonmem 7 1 2 and earlier releases o
158. d NOTITLE options in the SEST command the same as for the raw output file root xml NM72 An XML markup version of the contents of the NONMEM report file is produced automatically The rules schema document type definition by which it is constructed are given in output xsd and output dtd in the NONMEM util or vun directory In NM73 termination textmsgs catalogs termination text messages by number which can be mapped to source textmsgs f90 In nm73 termination status catalogs the error status For traditional analyses an error number is listed If negative the analysis was user interrupted For EM Bayes analysis error numbers map as follows 0 4 optimization was completed 1 5 optimization not completed ran out of iterations 2 6 optimization was not tested for convergence 3 7 optimization was not tested for convergence and was user interrupted 8 12 objective function is infinite problem ended 4 5 6 7 12 reduced stochastic sationary portion was not completed prior to user interrupt root cnv NM72 This file contains convergence information for the Monte Carlo EM methods if CTYPE gt 0 2000000000 mean of last CITER values 2000000001 standard deviation of last CITER values for objective function STD of second to last CITER values 2000000002 linear regression p value of last CITER values against iteration number 2000000003 Alpha used to assess statistical significance p value lt alpha Please note the
159. d for expectation step default 300 Usually 300 is sufficient but may require 1000 3000 for very sparse data and when desiring objective function evaluation with low Monte Carlo noise ISAMPEND n STDOBJ d NM73 For importance sampling and direct sampling only if ISAMPEND is specified as an integer value greater than ISAMPLE and STDOBJ is set to a real value greater than 0 then NONMEM will vary the number of Monte Carlo samples under each subject between ISAMPLE and nm730 doc 66 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 ISAMPEND until the stochastic standard deviation of the objective function falls below STDOBJ IACCEPT 0 4 Expand proposal sampling density variance relative to conditional density so that on average conditional density proposal densityzIACCEPT default 0 4 For very sparse data or highly non linear posterior densities such as with categorical data you may want to decrease to 0 1 to 0 3 IACCEPT 0 0 NM7 3 For importance sampling only you may set IACCEPT 0 0 and NONMEM will determine the most appropriate ACCEPT level for each subject and if necessary will use a t distribution by altering the DF for each subject as well If IACCEPT 0 the individual ACCEPT values and DF values will be listed in root imp where root is the name of the control stream file ISCALE_MIN 0 1 defaults for IMP NM72 ISCALE_MAX 10 0 NM72 In importance sampling the scale factor used to vary the size
160. de and have manager complete it paraprint 1 print to console the parallel computing process Can be modified at run time with ctrl B toggle Regardless of paraprint setting control stream log always records parallelization progress transfer type 0 for file transfer unloading and reloading workers with each estimation transfer type 1 for mpi transfer type 2 for file transfer maintaining a single loaded process throughout the run EH HE EXCLUDE INCLUDE may be used to selectively use certain nodes out of a large list SEXCLUDE 5 7 exclude nodes 5 7 O I E XCLUDE ALL 1 4 6 S NCLUDE I SNAMES Give a label to each node for convenience 1 MANAGER 2 WORKER1 3 WORKER2 4 WORKER3 COMMANDS each node gets a command line used to launch the node session Command lines must be on one line for each process The following commands nm730 doc 137 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 are for FPI method on Windows First node is manager so it does not get a command line when using FPI 1 NONE load on a core of the same computer as manager For psexec notice that the worker directories are named as the worker sees them not as the manager sees them Very important distinction for remote worker computers w refers to working directory for particular process 2 psexec d w workerl c
161. des If loading time too costly will eventually revert to single CPU mode timeouti seconds to wait for node to start if not started in time deassign node and give its load to next worker until next iteration timeout minutes to wait for node to compelte if not completed by then deassign node and have manager complete it paraprint 1 print to console the parallel computing process Can be modified at run time with ctrl B toggle Regardless of paraprint setting control stream log always records parallelization progress transfer type 0 for file transfer unloading and reloading workers with each estimation transfer type 1 for mpi transfer type 2 for file transfer maintaining a single loaded process throughout the run THE EXCLUDE INCLUDE may be used to selectively use certain nodes out of a large list SEXCLUDE 5 7 exclude nodes 5 7 Or SEXCLUDE ALL SINCLUDE 1 4 6 SNAMES Give a name to each node which is displayed 1 MANAGER 2 WORKER1 3 WORKER2 COMMANDS each node gets a command line used to launch the node session The first one launches the manager s NONMEM wdir refers to working directory for particular process mean to transfer all options from command line to manager process s nonmem exe 1 mpiexec wdir cd hosts 1 localhost 1 noprompt nonmem exe the next one launches a worker process on the
162. dicates this line contains the standard errors to the OMEGA and SIGMA elements in standard deviation correlation format 9 Iteration 100000006 indicates 1 if parameter was fixed in estimation 0 otherwise 10 Additional special iteration number lines may be added in future versions of NONMEM The raw output file is provided automatically independent of the formatted files that may be requested by the user using the STABLE command For the output files generated during the SEST step the following parameters may be specified FILE my_example ext Parameters objective function printed to this raw output file every PRINT iterations Default is control ext where control is name of control stream file DELIM s or FORMAT t or FORMATS Delimiter to be used in raw output file FILE S indicates space delimited T indicates tabs not case sensitive Default is spaces DELIM s1PE15 8 or FORMAT s1PG15 8 or FORMAT tF8 3 In addition to the delimiter a format FORTRAN style may be defined for the presentation of numbers in the raw OUTPUT file Default format is s PEI2 5 The variables DELIM and FORMAT are interchangeable The lines produced in the ext file may be very long You may optionally provide a line length followed by a continuation marker to be tagged at the end of each line e and or a continuation marker to be tagged at the beginning of the continuing line FORMAT s1PE15 8 160 amp will print lines of at most 160 characters f
163. different cores then they can communicate on an agreed upon directory on a local drive Both manager and worker must have read and write privileges To obtain the greatest efficiency in parallel computing make sure the LIM values to buffers 1 3 4 13 and 15 are set to the largest needed for ensuring the buffers can be loaded all into memory and no file reading and writing is required See the section 1 7 Changing the Size of NONMEM Buffers on how to do this File Passing Interface FPI Method Two information passing methods between manager and worker processes are available file passing interface FPD and message passing interface MPI The FPI method requires no additional software installation other than what is normally required to run a single process NONMEM run that is it needs only NONMEM plus compiler All transfer of information between a manager NONMEM process and its worker processes is done by writing files to a directory throughout the analysis Message Passing Interface MPI method The message passing interface MPI allows exchange of data much more rapidly than the FPI MPI requires installation of free but ubiquitous use third party software and we recommend you set this up for your cluster Fortunately MPI is free and available for most platforms and Fortran compilers The MPI s speed is particularly notable over FPI when FOCE Laplace SAEM and BAYES are done For ITS and IMP IMPMAP the speed difference is less
164. dom Now to analyze the data we may first analyze it by assuming a normal distribution as in this control stream file examples tdist6 ctl SPROB RUN Example 1 from samp51 SINPUT ID TIME DV CONC AMT DOSE RATE EVID MDV CMT SDATA tdist6 csv IGNORE C SSUBROUTINES ADVAN3 TRANS4 PK MU_1 THETA MU 2 THETA MU 3 THETA MU 4 THETA NU 4 0 CL EXP MU_1 ETA 1 V1 EXP MU_2 ETA 2 Q EXP MU_3 ETA 3 V2 EXP MU_4 ETA 4 S1 V1 SERROR Y F F EPS 1 STHETA 1 68338E 00 1 58811E 00 8 12694E 01 2 37435E 00 STHETA 2 2 2 2 SOMEGA BLOCK 4 0 3 0 001 0 3 0 001 0 001 0 3 0 001 0 001 0 001 0 3 SIGMA 03 SEST METHOD ITS LAPLACE INTERACTION MAXEVAL 9999 PRINT 5 NOHABORT SIGL 8 CTYPE 3 NITER 200 SEST METHOD IMP INTERACTION MAXEVAL 9999 PRINT 1 NOABORT ISAMPLE 3000 NITER 200 SIGL 8 DF 1 SEST METHOD 1 LAPLACE INTERACTION MAXEVAL 9999 PRINT 1 NOHABORT SCOV MATRIX R UNCONDITIONAL Note that Laplace is used for conditional estimation since the posterior density will by quite a bit not normally distributed For importance sampling a t distribution proposal density is used to approximately match the posterior density shape The result will be thetas and sigmas that approximate the simulation values used whereas the OMEGAS will be increased by a factor of about NU NU 2 see 11 bottom of page 341 When estimating in the manner in which it was simulated the thetas sigmas and omegas will more closely match t
165. e end function END MODULE MYFUNCS Nonmem reserved myfunc is the include file that declares its use USE myfuncs only mymin mymax and the following control stream file uses the function nm730 doc 27 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SPROB THEOPHYLLINE POPULATION DATA SINPUT ID DOSE AMT TIME CP DV WT SDATA THEOPP SSUBROUTINES ADVAN2 OTHER myfuncmodule SPK THETA 1 MEAN ABSORPTION RATE CONSTANT 1 HR THETA 2 MEAN ELIMINATION RATE CONSTANT 1 HR THETA 3 SLOPE OF CLEARANCE VS WEIGHT RELATIONSHIP LITERS HR KG SCALING PARAMETER VOLUME WT SINCE DOSE IS WEIGHT ADJUSTED include nonmem reserved myfunc CALLFL 1 KA THETA 1 ETA 1 K THETA 2 ETA 2 CL THETA 3 WT ETA 3 SC CL K WT I mymin 1 2 3 4 5 0 Print Ahi CET STHETA 1 3 5 008 08 5 004 04 9 SOMEGA BLOCK 3 6 005 0002 3 006 4 SERROR Y F EPS 1 SSIGMA 4 If you use the wrong argument type real instead of integer or perhaps use the wrong number of arguments the compiler will readily flag this Numerical Equality Comparison for IGNORE option in DATA Record NM73 When the IGNORE option is used to filter records from the input file the EQ NE and symbols perform literal string comparisons To provide a numerical equality comparison use EQN for numer
166. e S is Sobol sequence and m is the Sobol scrambler See the description of RANMETHOD under 1 25 Monte Carlo Importance Sampling EM NONMEM s default random number generator for the SIM step is 4 in contrast default random number generator for EST and STABLE is 3 Number 4 is NONMEM s classic nm730 doc 62 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 random number generator Whatever random number generator is selected it affects all seedl sources and all source seed2 The Sobol method is used only to generate normally distributed random vectors of etas and epsilons when the S descriptor is selected and SEEDI source 1 is used to set the seed Among the Sobol sequence methods the S2 method appears to provide the least biased random samples that is nearly uniform distribution with good mixing in multi dimensional spaces 1 22 Some Improvements in Nonparametric Methods NM73 EXPAND NM73 NONP EXPAND After the parametric estimation is performed the final eta MAP or empirical Bayes estimates EBE estimates based on the final SIGMAS OMEGAS and THETAS are normally used as support points If the natural distribution of etas among subjects is highly non normal with large tails or there are several outlier subjects the final Omega values may constrain the EBE s of these outliers so they do not fit these subjects well When EXPAND is selected an alternative set of EBE s are evaluated using the initial OMEGA
167. e continues to the next problem This is corrected in NONMEM 7 2 As a work around with earlier releases do not use DROP in control streams with more than one problem unless the same items are dropped in all problems 1 3 Introduction to NONMEM 7 and higher Many changes and enhancements have been made from NONMEM VI release 2 0 to NONMEM 7 In addition to code modification and centralization of common variables for easier access and revision the program has been expanded to allow a larger range of inputs for data items initial model parameters and formatting of outputs The choice of estimation methods has been expanded to include iterative two stage Monte Carlo expectation maximization EM and Monte nm730 doc 18 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Carlo Bayesian methods greater control of performance for the classical NONMEM methods such as FOCE and Laplace and additional post analysis diagnostic statistics Attention NONMEM 7 and higher produces a series of additional output files which may interfere with files specified by the user in legacy control stream files The additional files are as follows root ext root cov root coi root cor root phi root phm root shk root shm root xml root smt root rmt root agh root fgh Where root is the root name not including extension of the control stream file given at the NONMEM command line or root nmbayes if the control stream file name is not given
168. e FOCE method and the problem may not even optimize successfully If you choose one of the new methods and you do not incorporate MU referencing into your model you are likely to be disappointed in its performance For simple two compartment models the new EM methods are slower than FOCE even with the MU references But for 3 compartment models or numerical integration problems the improvement in speed by the EM methods properly MU modeled can be 5 10 fold faster than with FOCE Example 6 described at the end of the SIGL section is one example where importance sampling solves this problem in 30 minutes with R matrix standard error versus FOCE which takes 2 10 hours or longer and without even requesting the COV step So for complex PK PD problems that take a very long time in FOCE it is well worth putting in MU references and using one of the EM methods even if you may need to rephrase some of the fixed random theta eta effects relationships In addition FOCE is a linearized optimization method and is less accurate than the EM and Bayesian methods when data are sparse or when the posterior density for each individual is highly non normal It cannot be stressed too much that MU referencing and using the new EM methods will take some time to learn how to use properly It is best to begin with fairly simple problems to understand how a particular method behaves and determine the best option settings When nm730 doc 90 of 210 NONMEM
169. e a Monte Carlo based set of EPRED ERES ECWRES NPDE and EWRES ESAMPLE should be specified only on the first STABLE command By default ESAMPLE 300 WRESCHOL NM73 Normally population and individual weighted residuals are evaluated by square root of the eigenvalues of the population or individual residual variance However an alternative method is to Cholesky decompose the residual variance suggested by France Mentre personal nm730 doc 48 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 communication by entering the WRESCHOL option This should be specified only on the first TABLE command The Cholesky form has the property of sequentially decorrelating each additional data point in the order of the data set SEED Specify starting seed for Monte Carlo evaluations of EPRED ERES EWRES ECWRES and NPDE The default seed is 11456 SEED should be specified only on the first STABLE command RANMETHOD nISImIP NM72 default n 3 By default the random number generator used for Monte Carlo simulations of weighted residual items is ran3 of reference 5 We feel this is the best random number generator for many purposes However you may choose alternative random number generators as follows 0 ranO of reference 5 minimal standard generator 1 ranl of reference 5 Bays and Durham 2 ran2 of reference 5 3 ran3 of reference 5 Knuth 4 NONMEM s traditional random number generator used in SSIMULATION
170. e correlated normalized prediction distribution error reference 3 does not take into account within subject correlations also a Monte Carlo assessed diagnostic item For each simulated vector of data yx IWRES V n Yu f my These are then averaged over all the random samples 1 K pd G IIWRES Ki Then an inverse normal distribution transformation is performed npd pd The default PRED RES and WRES will be given the same values as PREDI RESI and WRESI when INTERACTION in EST is specified or NPRED NRES and NWRES when INTERACTION in SEST is not specified As the PRED RES and WRES may be referenced in a user supplied INFN routine or in PK or PRED when ICALL 3 as PRED RES WRES so the additional parameters may be referenced by their names followed by for example EWRES CIWRES CIPRED CIRES CIWRESI NM73 The CIWRES is the conditional individual weighted residual as evaluated during the estimation equivalent to DV F F SQRT SIGMA 1 1 for simple problems with proportional residual error With L2 data or CORRL2 data the individual weighted residuals are in their decorrelated forms CIWRES V Y 8 when INTERACTION in the previous SEST record is set and a conditional analysis non FO was performed For individual i where individual residual variance matrix V and individual predicted vectorf f are evaluated at the conditional mode or mean eta designated as eta
171. e following example OPTMAP 1 is chosen to provide forward finite difference eta derivatives for the search and ETADER 2 is chosen to provide numerically assessed central finite difference derivatives to the Hessian matrix of the posterior density sde12 ctl allowing ITS and FOCE to obtain results similar to Importance sampling SEST METHOD ITS INTERACTION NOABORT PRINT 1 CTYPE 3 OPTMAP 1 ETADER 2 SIGLO 6 SIGL 6 MCETA 1 SEST METHOD IMP INTERACTION NOABORT PRINT 1 IACCEPT 1 0 CTYPE 3 OPTMAP 0 ETADER 0 SIGLO 6 SIGL 6 MCETA 1 MAPITER 0 SEST MAXEVAL 9999 METHOD 1 INTER NOABORT NSIG 1 PRINT 1 MSFO sdel2 msf OPTMAP 1 ETADER 2 SIGLO 6 SIGL 6 MCETA 1 SLOW COV MATRIX R UNCONDITIONAL TOL 9 SIGL 8 SIGLO 8 STABLE ID TIME FLAG AMT CMT IPRED IRES IWRES nm730 doc 165 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 ONEHEADER NOPRINT FILE sde9 fit 1 56 Turning on First Derivative Assessments for EM Bayes Analysis NM72 NONMEM 7 2 0 normally calculates first derivatives in the FSUBS file for classical NONMEM methods and does not evaluate them for IMP SAEM and BAYES methods This improves the speed at which the problem is evaluated However on occasion such derivatives are needed for example when steady state values are to be calculated or when stochastic differential equations are to be evaluated In such cases insert as the first line in a control stream section such as PK SERROR DES etc FIRSTEM 1 Then incidental derivativ
172. e for a given problem Even the classical NONMEM methods can be facilitated using an EM method by first having a rapid EM method such as iterative two stage be performed first with the resulting parameters being passed on to the FOCE method to speed up the analysis SEST METHOD ITS INTERACTION SEST METHOD CONDITIONAL INTERACTION More information on this is described in the Composite Methods section 1 11 Interactive Control of a NONMEM batch Program A NONMEM run can now be controlled to some extent from the console by issuing certain control characters Console iteration printing on off during any Estimation analysis ctrl J from console NONMEM Iterations button from PDx POP Exit analysis at any time which completes its output and goes on to next mode or estimation method ctrl K from console or Next button in PDx POP Exit program gracefully at any time ctrl E or Stop button nm730 doc 40 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Monitor the progress of each individual during an estimation by toggling ctrl T Wait 15 seconds or more to observe a subject s ID and individual objective function value It is also good to test that the problem did not hang if a console output had not been observed for a long while If you run NONMEM from PDx POP you can get graphical view of objective function or any model parameter progress during the run The parameter and objective function progress is written in
173. e model records For each subject record NONMEM generates information about the component models of a mixture model this information constitutes the mixture model record The size of buffer 5 is related to the number LIMS of mixture model records stored in memory at any one time The default size of buffer 5 has been set to allow LIM5 200 mixture model records Each mixture model record consists of five 8 byte single precision computer words The allocation of memory for buffer 5 is MMX 1 LIM5 3 8 bytes Buffer 6 holds a number of contiguous PRED defined records For each data record of a given subject record NONMEM stores the values found in module NMPRD4 these values comprise the NMPRD4 record The size of buffer 6 is related to the number LIM6 of PRED defined records stored in memory at any one time The size of buffer 6 has been set to allow LIM6 400 PRED defined records The least number of PRED defined records allowable must exceed the largest number of data records used with any one subject which rarely will be as large as 400 Each PRED defined record consists of PDT 8 byte double precision computer words The allocation of memory for buffer 6 is PDT LIM6 3 8 bytes Buffer 7 holds a number of contiguous NMPRD4 records for a single individual only For each problem in a NONMEM run NONMEM generates information about the problem this constitutes the problem header record The size of buffer 7 is related to the number LIM7 of NMPR
174. e modeled via THETA a sigma like Theta parameters is set up in example 2 For a thorough reference to the options in the PRIOR record see the html Help manual The following describes the setup for most Bayesian analysis purposes To set up the PRIOR NWPRI statement keep in mind the following NTHETA number of Thetas to be estimated NETA number of Etas Omegas to be estimated and is to be described by an NETAxNETA OMEGA matrix NEPS number of epsilons Sigmas to be estimated and is to be described by an NEPSXNEPS SIGMA matrix NTHPznumber of thetas which have a prior NETP number of Omegas with prior NEPP Number of Sigmas with prior NM73 Before NM73 the NEPP option was ignored as supplying priors for Sigma s was not activated For example PRIOR NWPRI NTHETA 4 NETA 4 NEPS 1 NTHP 4 NETP 4 NEPP 1 Then the STHETA records list the parameters in order the following NTHETA of initial thetas NTHP of Priors to THETAS Degrees of freedom to each OMEGA block Prior Degrees of freedom to each SIGMA block Prior The 0MEGA records list the variances in order the following NETAxNETA of initial OMEGAS NTHPxNTHP of variances of Priors to THETAS NETPxNETP of priors to OMEGAS matching the block pattern of the initial OMEGAS The SIGMA records list the variances in order the following NEPSxNEPS of initial SIGMAS NEPPxNEPP of priors to SIGMAS matching the block pattern of the initial SIGMAS NM73 So we may have the
175. e of these augmented records would be taken from example6b ctl in the util directory SPROB RUN example6 from r2compl SINPUT C SET ID JID TIME DV CONC DOSE AMT RATE EVID MDV CMT SDATA example6b csv IGNORE C SSUBROUTINES ADVAN13 TRANS1 TOL 4 SMODEL NCOMPARTMENTS 3 SPK SDES SERROR CALLFL 0 ETYPE 1 nm730 doc 175 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 F CMT NE 1 ETYPE 0 IPRED F Y F F ETYPE EPS 1 F 1 0 ETYPE EPS 2 SEST METHOD ITS INTERACTION SIGL 4 NITER 25 PRINT 1 FILE example6 ext NOABORT STABLE ID TIME CONC IPRED CMT MDV EVID NOAPPEND NOPRINT FILE example b fin FORMAT 1PE12 5 ONEHEADER Of importance here is the TABLE record The file example6b fin is generated by NONMEM providing individual predicted values for each incremental time because of their presence in the input data file example6b csv Because incremental time records have MDV 1 there will be no impact on the estimation results The table structure and contents of example6b fin is suitable for importing into plotting programs which can present smooth prediction curves choose connect line and no symbol superimposed on observed data choose with symbol and no connect line Although the added MDVz1 fine date lines do not impact the estimation results except where NONMEM may utilize time changing covariates and pick up a covariate value from these new records they can increase estimation time It may therefore be of advant
176. e theta pertaining to the highest SID value NSID to be the negative sum of the thetas to the other SID values 1 through NSID 1 using a DOWHILE loop For this method some pseudo data must be added to the data file Original data portion TYPE 0 ID TIME DV AMT RATE EVID MDV CMT ROWNUM BID WYPE he 0OE 00 1 00E 00 0 00E 00 0 00E 00 1 00E 00 0 00E 00 1 00E 00 1 00E 00 1 00E 00 1 00E 00 1 00E 00 0 00E 00 1 00E 00 0OE 00 1 00E 00 1 00E 01 2 44E 00 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 2 00E 00 1 00E 00 0 00E 00 2 00E 00 0OE 00 1 00E 00 2 00E 01 4 45E 00 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 3 00E 00 1 00E 00 0 00E 00 3 00E 00 0OE 00 1 00E 00 5 00E 01 9 93E 00 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 4 00E 00 1 00E 00 0 00E 00 4 00E 00 0OE 00 1 00E 00 1 00E 00 1 65E 01 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 5 00E 00 1 00E 00 0 00E 00 5 00E 00 00E 00 1 00E 00 2 00E 00 2 05E 01 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 6 00E 00 1 00E 00 0 00E 00 6 00E 00 0OE 00 1 00E 00 5 00E 00 1 82E 01 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 7 00E 00 1 00E 00 0 00E 00 7 00E 00 0OE 00 1 00E 00 1 00E 01 7 20E 00 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 8 00E 00 1 00E 00 0 00E 00 8 00E 00 0OE 00 1 00E 00 2 00E 01 1 29E 00 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 9 00E 00 1 00E 00 0 00E 00 9 00E 00 0OE 00 1 00E 00 5 00E 01 6 80E 03 0 00E 00 0 00E 00 0 00E 00 0 00E 00 2 00E 00 1 00E 01 1 00E 00 0 00E 00 1 00
177. eam File input limits 39 1 10 Issuing Multiple Estimations within a Single Problem 39 1 11 Interactive Control of a NONMEM batch Program sese 40 1 12 COV Unconditional Evaluation 2 2 6 2e icut ue ct det ret reae ce tt e ocn e te canere cs 42 1 13 STABLE Additional Statistical Diagnostics Associated Parameters and Qu utpu t Format e 42 Requesting a Range of Etas to be Outputted Etas x y NM73 eee 42 OIA i PAE rtt Ete ER oa dee ERN EEEE EAME RU RR Uu Eo Va EE RR UO DON oS a las OUR S ONU 43 NPRED NRES NWRES iii teovicetesa tob een FRI Rea INE P VEEXE SER e EE EE PERENNEM eR Tbe sosro VE a E ERE REPE ERR XAR 43 PREDI RESI WRESL iue anos GNO Cni PME ei IEEE MR MEE 43 CPRED CRES CWRES sccsstescaseicevecsscdsdenssiddveecisseasotstdovsccscdecesssehcssedetsatsetnitovs sabs sses ese sisses 43 CPREDI CRESI CWRESI eossecssescossocssessossocsossooesosssecocescossccosescosoocsscesossossosesossossssessesosss 44 EPRED ERES EW RES isc scivesecscksSuukscinceceasvecusses uss been pro Mae E Ire soe eral ebore sotsi ssi osre is soseo 44 HCW RES m n 45 NPDE 2 bb pq ep a ge E RUN b Ie a c PR DUK SNR rr DP OR REI IN OPEN DOE 45 IND D EE AAE TTE NCRS ORNARE ENT RR NEN NU RETRO CURR RUE NI STAR DAR 46 CIWRES CIPRED CIR
178. eam file the attempt will be to match ID numbers rather than subject numbers if an ID column in the file exists nm730 doc 130 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 which it will if you are using a phi or phm file generated from a previous nonmem analysis The phc etc variances will also be brought in The etas inputted by ETAS PHIS can be used in several ways In BAYES SAEM and IMP MAPITER 0 they are used as the starting etas in the first iteration In MAP estimation matters such as METHOD 1 or ITS or IMP MAPITER gt 0 or IMPMAP and if MCETA gt 0 then these etas are one of the initial eta vector positions tested during the first iteration and the one giving the lowest OBJ is then selected In cases where FNLETA 2 the estimation step is skipped and etas inputted from ETAS are passed directly to the Final processing steps That is these etas are treated as if they were the final result of an estimation The final processing steps use routines such as FNLETA FNLMOD PRRES NP4F that contribute to generating STABLE SCATTER outputs including the various WRES diagnostics where applicable When METHOD O these initial etas are not used as this method does not require initial etas One purpose to bringing initial eta phi and etc phc values is you can readily resume an analysis if an MSF file was not set up in the previous analysis the MSF file system is still the most complete information transfer for res
179. ection 1 27 Stochastic Approximation Expectation Maximization SAEM Method If an interruption occurred during FOCEI Laplace FO during the COV step covariance analysis may be resumed where it left off See RESUME NM73 in section 1 41 COV Additional Parameters and Behavior In addition the following bugs have been fixed that were in NONMEM 7 2 0 1 Some operating systems do not like the word nul for a file name for FNULL Work around for earlier versions of NONMEM change nul to JUNK in resource nmdata f90 rebuild NONMEM by running SETUP72 or SETUP72 bat in the installed NONMEM directory For example for Windows gfortran if c nm72g is your installed NONMEM directory then from c nm72g execute the following command in the command window setup72 c nm72g c nm72g gfortran y ar same rec n 2 In parallelization Windows 64 gfortran compiled using population mixture model a variable is not initialized and causes parallelization failure Work around for earlier versions of NONMEM is to add the gfortran compiler switch finit integer 0 To do this edit setup72 bat line 247 or setup72 362 adding finit integer 0 just before ffast math do not place it as the last optimizing option Then rebuild NONMEM For example if c nm72g is your installed NONMEM directory then from c nm72g execute the following command in the command window setup72 c nm72g c nm72g gfortran y ar same rec n 3 BY USER INTERUPT
180. ed All other control stream records are ignored Thus a way to create a control stream is to copy the first records describing the data layout from an existing NONMEM control stream file and then adding the FINEDATA record The options to FINEDATA are as follows nm730 doc 172 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 TSTART start time real number or integer for creating incremental time records If you specify FIRST or do not specify a value for TSTART then the time of the first record of the subject or occasion see OCC below is used or when the time is less than that of the previous record or when EVID 3 or EVID 4 If TSTART is not a number and is not FIRST then it is interpreted as the column name in the original data set containing the start time In such cases the TSTART value of the first data record of the subject is used or of the first data record or upon occasion change if OCC was given or if EVID 3 or 4 or after a re initialization of time indicated by the time in the data record being less than that of the previous record Thus TSTART could differ according to instance The same holds true for TSTOP TDELTA or NEVAL see below if they are obtained from the data file OCC name of occasion column This is optional and will restart the time incrementing when the occasion changes in addition to the other conditions listed above NEVAL number of incremental time records per subject integer or trun
181. eeeeeeeeeeeeeeeeeenn nnn 106 1 43 Adding Nested Random Levels Above Subject ID NM73 107 1 44 Model parameters as log t Distributed in the Population NM73 112 1 45 Format of NONMEM Report File eeeeeeeeeeeeeeeeeeeeeee nennen 115 PARA CININET 2 iss uvice cua encdan ino ec eel i X ke dra esu fei ipei b FR URP EP Pede Er RE aen 115 ZFIBLN NM72 uiteeixtoxisiviseonm mdbicterid ket eese oen ti cesi orson as oeiee 115 SiAn 0 DA N a EEES E EEE E EA A EAEE 115 cia M DA td PR EEA EEEE AES A EAEE E E SEE 115 L 280i 10 D POTIUS 116 CBB y p Pera rcp TRE 116 81 3 reb T E RATE 116 nm730 doc 5 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 FOB oc Seri eeoa 116 FOYE o TINT TS eror str 116 CPUT M73 sdscsicansasicnankspeatandetasandagdencaeiabiceshiesaueaedstincedeposacassascooehh stutanestoacednscsocssenbsseuta DUE 116 Shrinkage and ETASTY PE NMT3 scsoicsoviesssvnissvcooscsvonssucccvsrsonsesensssdbecesevonsesonssnsensvobosssnace 116 1 46 EST Format of Raw Output File cccccsseeceeeeeeeeeeeeeeeeeeeeeseeeeeeeeeneeeeeeeeees 118 BEERS my Ox Ar ple extis oes ciscavivesdivs scasccusscvedavaddadsotenisttuvs dsatdedcutitveraddesdeddsintes sieeatdesdesn esasdess 119 DELIMEs or FORMAT t or FORMATS cssssccssssccssssccssscccsssccsssscscssssscssssssssssssessec
182. efault For many problems the default LIM values are high enough that all of the data may reside in memory without resorting to the buffer files For large nm730 doc 35 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 data sets buffer files are likely to be used The user may however select a LIM value that is different from that specified in sizes f90 via the SIZES record in the control stream file e g SIZES LIM1 20000 It is not necessary to recompile NONMEM just rerun the nmfe73 script and the appropriate arrays will be allocated according to the user specified LIM value It is most desirable to set the LIM value that is the proper size for the run so that the buffer file does not have to be used With today s very large memory computers this should usually be alright to do without running out of memory Below is a table describing the minimal allowable value for each LIM and the value needed to prevent using the buffer file for a particular problem LIM Minimum Maximum Value needed to prevent Buffer files used Value buffer file usage FILExx 1 MAXDREC TOTDREC 10 13 20 33 2 MAXDREC TOTDREC 39 14 3 2 MAXIDS 12 4 2 MAXIDS 15 16 5 2 MAXIDS 17 18 6 MAXDREC TOTDREC 7 19 7 2 MAXDREC 21 22 8 2 MAXIDS 23 24 9 NOT USED 10 NOT USED 11 2 NPROB 31 32 12 NOT USED 13 2 MAXIDS 11 14 NOT USED 15 2 MAXIDS 26 27 16 MAXDREC TOTDREC 26 27
183. efined then nmfe73 will behave as in earlier versions and rely on the presently existing PATH for finding the compiler and MPI system The nmfe73 script will display a statement as to what path it will use 1 6 Dynamic Memory Allocation NM72 With NONMEM 7 2 0 and higher versions the user need no longer specify big or reg when using SETUP72 or SETUP73 to install NONMEM The reg big same choice is ignored It is in effect always same and is shown as same in all examples However some constants in SIZES are not dynamically allocated for example LSTEXT or PNM MAXNODES See help entry for sizes or see comments regarding the various parameters in resource VS IZES f90 NMTRAN sizes each NONMEM executable only as large as it needs to be for the specific control stream run NONMEM 7 2 0 has the ability to dynamically size the main arrays in NONMEM according to the number of subjects and number of parameters described in the nm730 doc 30 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 control stream file etc To do this NMTRAN determines the appropriate sizes for arrays and puts this information in a subroutine called FSIZESR in the FSUBS file NONMEM dynamically allocates the sizes of arrays at run time based on the values in FSIZESR Although unnecessary for most problems the user may over ride the size that NMTRAN assesses for a select number of arrays by including a SIZES statement as the first non co
184. ell as to individual cores on those computers However depending on your intranet connection between computers the process will be a little slower across computers than among cores on the manager computer alone Eight to 16 cores per computer with about 2 GB RAM per core should be sufficient for almost any problem in NONMEM Alternatively 0 4 GB per core is more than enough for many NONMEM problems If there is insufficient RAM many operating systems utilize virtual memory usually mapped to hard drives but this may slow down execution The manager process is the user s process that runs the nmfe73 script reads the control stream file executes NMTRAN and runs the main NONMEM process The worker process is NONMEM in worker mode not taking any input from the user only from the manager NONMEM process If the manager process is on one computer and the worker process is on a second computer then a network communication must be possible between these computers and the manager computer must be able to have access to a network drive and directory that is mapped to a drive and directory that is locally accessible by the worker directory It is possible for this directory to also be accessible from the worker computer as a network drive but this can slow down the data transfer If the manager process and the worker process are on the same computer but are nm730 doc 135 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 simply running on
185. ence or credible ranges The mean population parameter estimates and their variances are evaluated with considerable stability Maximum likelihood parameters are not obtained but with problems of sufficient data these sample mean parameters are similar to maximum likelihood values and the standard deviations of the samples are similar to standard errors obtained with maximum likelihood methods A maximum likelihood objective function is also not obtained but a distribution of joint probability densities is obtained from which 95 confidence bounds assuming a type I nm730 doc 75 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 error of 0 05 is desired can be constructed and tested for overlap with those of alternative models As with the SAEM method there are two phases to the BAYES analysis The first phase is the burn in mode during which population parameters and likelihood may change in a very directional manner with each iteration and which should not be used for obtaining statistical summaries The second phase is the stationary distribution phase during which the likelihood and parameters tend to vary randomly with each iteration without changing on average lt is these samples that are used to obtain summary statistics The Bayesian method is specified by EST METHOD BAYES INTERACTION Followed by one or more of the following parameter options NBURN 4000 Maximum number of iterations in which to perform the bur
186. ence of commands EST METHOD SAEM INTERACTION NBURN 2000 NITER 1000 EST METHOD IMP EONLY 1 ISAMPLE 1000 NITER 5 Here after SAEM is performed importance sampling with MAP estimation done on its first iteration is performed but without updating the main population parameters Sometimes the MAP estimation is problematic and or the user wishes to use the SAEM s last conditional mean and variances as the parameters to the importance sampler s sampling density for the first iteration so one may try EST METHOD SAEM INTERACTION NBURN 2000 NITER 1000 EST METHOD IMP EONLY 1 ISAMPLE 1000 NITER 5 MAPITER 0 For very large dimensioned problems many Omegas the IMP evaluated objective function can have a lot of stochastic variability more than plus or minus 10 units or continually increase nm730 doc 74 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 with each iteration even though the population parameters are kept fixed One way to reduce this volatility is to use IMPMAP instead of IMP if the MAP estimation is not an issue EST METHOD IMPMAP EONLY 1 ISAMPLE 1000 NITER 5 MAPITER 0 Another way is to increase the ISAMPLE to 3000 EST METHOD IMP EONLY 1 ISAMPLE 3000 NITER 5 MAPITER 0 and sometimes using the combination of IMPMAP with ISAMPLE 3000 is needed Using IMPMAP or increasing ISAMPLE do increase computation time and it is a choice of which is more efficient Another set of commands for SAEM is the following whi
187. encivnisstediedeatdaemanieinuiaaanwas 132 1 51 Imposing Thetas Omegas and Sigmas by Algebraic Relationships Simulated Annealing Example cccccccssssesseeeeeeeeeeeeeeeeeeeeeeeeeeeeeeseeeeeeeeeeeeeeeeeeeeenees 133 1 52 Stable Model Development for Monte Carlo Methods 133 1 53 Parallel Computing NM72 ccccssscsseeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeseeeeeeeeeeeeeeeeseeeeeeeeeees 135 File Passing Interface FPI Method ce eee ee ee ee ee eee eee eese eese e eese e eese ess 136 Message Passing Interface MPI method 4 eee e eee ee eee eee ee ee ee ee eee seen a setas 136 The PARAFEIDE recita ii UE CR S ERRTE HL ER IOVIS EARN DERE D ERAT APA E REPERTUS 136 nm730 doc 6 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Substitution Variables in the parafile ccce e eee eee eee eene ee eee eee en eee ta seen seen uu 139 Easy to Use icc eec 142 Setting up a network drive on Windows for multiple Computers 143 Setting up FPI on Windows wsvsivssesisasssdeseceessaceassviocsvavocsousdeveevensovecossdeecesexonsosscnssndenevabesssnane 143 Installing MPI on Wind WS iiitiliees toi seine ka seva De bae Febo oa GR sa SPESE FAVERE RR RR EROR URN M e Eo EM Ko ERREUR RR 146 Setting up share directory and ssh on a Linux System
188. enerated samples but will just be what was listed in THETA and OMEGA of the control stream file If NSAMPLE 0 but ISAMPLE some number then it is expected that FILE already exists and its iteration number specified by ISAMPLE is to be read in for setting initial values SEST METHOD CHAIN FILE examplel chn NSAMPLE 0 ISAMPLE 3 One could create a control stream file that first creates a random set of population parameters and then sequentially uses them as initial values for several trial estimation steps SPROBLEM 1 SINPUT C SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT SDATA wexamplell csv IGNORE SSUBROUTINES ADVAN3 TRANS4 SPK SERROR STHETA 2 0 2 0 4 0 4 0 Initial Thetas SOMEGA BLOCK 4 Initial Parameters for OMEGA 2 0 01 2 0 01 0 01 2 0 01 0 01 0 01 2 SSIGMA 0 5 First problem creates NSAMPLE 5 random sets of initial parameters stores them in examplell chn Then selects the first sample ISAMPLE 1 nm730 doc 126 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 for estimation SEST METHOD CHAIN FILE wexamplell chn NSAMPLE 5 CTYPE 2 ISAMPLE 1 DF 4 SEED 122234 IACCEPT 0 8 SEST METHOD COND INTERACTION MAXEVAL 9999 NSIG 2 SIGL 10 PRINT 5 NOABORT FILE wexamplell l ext T SPROBLEM 2 INPUT C SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT SDATA wexamplell csv IGNORE
189. eneuseesesessasstanvaensvaons 61 NOREPLEACE NM373 scient dksupn estet a tod is bo ene sisese sesane seision NES SEM UO 61 STRAT4NMT73 5 osihenratiuttvte tenete aces ine ote bed es CU ases uunc Ree NC NR RIEN Qus 62 STRATF NM cese iiit te D INR pei eEH SEEN ael pe ie tigres EN EROR ease RUNI aasenso iais 62 RANMETHODS ilSImlP NM73 ssessssosessesososesesessososossososeseseseosososcesososesesessososossesososesessesese 62 1 22 Some Improvements in Nonparametric Methods NM793 63 EXPAND NM793 5 isassntiesieteee a od EU PS RDUM IIS et aroas oro EET LEER 63 NPSUPP NMT3 ient nt Oto EU Odette NE DR Hai Qt ten i Qr RN UE 63 NPSUPPE NIVITS T H 63 BOOTSTRAP NMT3J ccn uso Dite nti MRSOK MP ERN MD er IU RHONE 63 SERAT STRATE NMT3 45 eto rrictieasetu etes rotae epe ide bede eer de sosiaa tesisse ossosa dee SRL LN PA Sen 64 1 23 Introduction to EM and Monte Carlo Methods 65 1 24 Iterative Two Stage ITS Method eeeeeeeeeeeeeee 65 EST METHOD ITS INTERACTION NITER 50 sscssssssssesscsessccesecsssesscsesecsesecseseees 65 1 25 Monte Carlo Importance Sampling EM eeeeeeeeeeeeeeeeee 66 SEST METHOD IMP INTERACTION 4 eirs risecortieie sentis erriskaede ciet shentectsccssetovaisasins 66 NITER NSAMPLE 50 svssescsonsisctsnecconsspses enia anon di raa ie eri o PAN
190. er modified to provide any constraining pattern on any population parameters RANMETHOD RANMETHOD SEED SEED 1 32 MU Referencing The new methods in NONMEM are most efficiently implemented if the user supplies information on how the THETA parameters are associated arithmetically with the etas and individual parameters wherever such a relationship holds Calling the individual parameters phi the relationship should be phi_i mu_i theta eta 1i For each parameter i that has an eta associated with it and mu iis a function of THETA The association of one or more THETA s with ETA 1 must be identified by a variable called MU 1 Similarly the association with ETA 2 is MU 2 that of ETA 5 is MU 5 etcetera Providing this information is as straight forward as introducing the MU variables into the PRED or PK code by expansion of the code For a very simple example the original code may have the lines CL THETA 4 ETA 2 nm730 doc 85 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 This may be rephrased as MU 2 THETA 4 CL MU_2 ETA 2 Another example would be CL THETA 1 AGE THETA 2 EXP ETA 5 V THETA 3 EXP ETA 3 which would now be broken down into two additional lines inserting the definition of a MU as follows MU 5 LOG THETA 1 THETA 2 LOG AGE MU_3 LOG THETA 3 CL EXP MU_5 ETA 5 V EXP MU 3 ETA 3 wa WH Note the arithmetic relationship ident
191. errors VAR R 2 A OMEGACSE O Omega correlation off diagonal standard errors COV R 2 A SIGMACSE D Sigma correlation diagonal standard errors VAR R 2 A SIGMACSE O Sigma correlation off diagonal standard errors COV R 2 A THETANP Nonparametric Thetas GENERAL EXNPETA EX non paramatric etas GENERAL COVNPETA D Covariance of nonparametric etas diagonals DIAG COVNPETA O Covariance of nonparametric etas off diagonals OFFDIAG OMEGANP D Omega of nonparametric analysis diagonals DIAG OMEGANP O Omega of nonparametric analysis off diagonals OFFDIAG COVNPETAC D Correlation of nonparametric etas diagonals DIAG R 2 A COVNPETAC O Correlation of nonparametric etas off diagonals COR OMEGANFC D Omega correlation of nonparametric analysis diagonals DIAG R 2 A OMEGANFC O Omega correlation of nonparametric analysis off diagonals COR COVARIANCE D Diagonals of variance covariance of estimates VAR COVARIANCE O Off Diagonals of variance covariance of estimates COV CORRELATION D Diagonals of correlation of variance covariance of estimates VAR R 2 A CORRELATION_O Off Diagonals of correlation of variance covariance of estimates COR INVCOVARIANCE_D Diagonals of inverse of variance covariance of estimates VAR INVCOVARIANCE_O Off Diagonals of inverse of variance covariance of estimates COV SMATRIX_D Diagonals of S MATRIX VAR SMATRIX_O Off diagonals of S MATRIX COV RMATRIX_D Diagonals of RRMATRIX VAR RMATRIX_O Off diagonals of R MATRIX COV Because of the versatil
192. ers Guide Introduction to NONMEM 7 3 0 There is additional increased efficiency in the evaluation of the problem if the MU models are linear functions with respect to THETA As mentioned in the previous examples above we could re parameterize such that MU 5 THETA 1 THETA 2 LOG AGE CL EXP MU_5 ETA 5 MU 3 THETA 3 V EXP MU_3 ETA 3 This changes the values of THETA 1 and THETA 3 such that the re parameterized THETA 1 and THETA 3 are the logarithm of the original parameterization of THETA 1 and THETA 3 The models are identical however in that the same maximum likelihood value will be achieved The only inconvenience is having to anti log these THETA s during post processing The added efficiency obtained by maintaining linear relationships between the MU s and THETA s is greatest when using the SAEM method and the MCMC Bayesian method In the Bayesian method THETA s that are linearly modeled with the MU variables have linear relationships with respect to the inter subject variability and this allows the Gibbs sampling method to be used which is much more efficient than the Metropolis Hastings M H method By default NONMEM tests MU THETA linearity by determining if the second derivative of MU with respect to THETA is nearly or equal to 0 Those THETA parameters with 0 valued second derivatives are Gibbs sampled while all other THETAS are M H sampled In the Gibbs sampling method THETA values are sa
193. es 119 DELIM s1PE15 8 or FORMAT s1PG15 8 or FORMA T tF8 3 cccssccsssscssssccssssceees 119 INO MM ISTA over 120 NOEABEDS 0 L oder bU ee DrbR IRE Eee pr Ea EE EPI Roe Qa e QURE PA T APR FUE VAR RERE EE IURE MURR 120 ORDER NNI72J ioeeiict yeQrrcese2nE vex eniqbvddeeuso ad eden deP ekat co Phi on dso codesto oaste M presa ae Pre oes PER RI oER 120 1 47 EST Additional Output Files Produced esee 121 RU PR IL 121 iiim Meme rrr HE E 121 FOOLCOL abbr co dieb iti cab ai Eb are E ATA 121 FOOLDDE oe rats chsh is PES hs cvinnn dishes shaun Do ed E LM ease ase ERAS M EE a vd sauces Lb iu P ARE RIEHE SQUE CR EROR Leod 121 TOOt pba E Un did 2 RD 121 root shk NM72 ente lop e Toe T cH 122 nsn FA P 122 POOH OTANI 2 PDT 123 root xnil CIN IW 2 sesicesevccsntetensccecetesistscsiceeaSeacededatonccoaceccoecescndsieaseokededasonscoesedeedocsesicceaseocedeuavenocs 123 root cnv CNIM7 2 eos eessececonoscepeeaosososeusavast esso se tonos ca pasas sodes oues vases e Gao do konek ee pe eaaseseseUssea se Nasa 123 root smt NIT 2 cocsccisvecsiececbectissisecceeceecscsisceceveccedeatoabeoccdoccdsdecossescodecseSectosbscecdoecdccecolseocedecoedes 124 root TIME du rp 124 root mp CN MIT S cassis catia cocaatwis ska rede Gensiutid weedaahaadtvidctivanndetensasdidetatucaboasannssbunnisedeadinaideessiasboass 124 POGUE
194. es will be evaluated for the new methods as well NMTRAN has been modified such that it collects all first derivative computations together and performs them only if FIRSTEM 1 For example in the PK subroutine generated for examples example1 ctl IF FIRSTEM 1 THEN A00033 DERIVATIVE OF CL W R T ETA 01 A00033 B00002 l A00038 DERIVATIVE OF V1 W R T ETA 02 A00038 B00004 M A00043 DERIVATIVE OF Q W R T ETA 03 A00043 B00006 i A00048 DERIVATIVE OF V2 W R T ETA 04 A00048 B00008 1 A00051 DERIVATIVE OF S1 W R T ETA 02 A00051 A00038 GG 01 1 1 CL GG 01 02 1 2A00033 GG 02 1 1 V1 GG 02 03 1 2400038 GG 03 1 1 0 GG 03 04 1 A00043 GG 04 1 1 V2 GG 04 05 1 2A00048 GG 05 1 1 S1 GG 05 03 1 A00051 ELSE GG 01 1 1 CL GG 02 1 1 V1 GG 03 1 1 0 GG 04 1 1 v2 GG 05 1 1 81 ENDIF Every effort has been made to assure that this new process by NMTRAN works for every type of model However it may occur that NMTRAN arranges the equations in the wrong order and your problem may not work correctly whereas it may have worked correctly in NONMEM 7 1 2 or earlier Should this occur the re arrangement of equations by NMTRAN can be turned off by inserting SABBREVIATED NOFASTDER in the control stream file If the problem is resolved using this setting please send your example control stream file to nmconsult and we will fix the error for the next version nm
195. eta positions are rejected by the Monte Carlo algorithm because of the poor resulting objective function But occasionally floating point overflows divide by zero or domain errors may occur which can result in failure of the analysis This may occur especially when beginning an analysis at poor initial parameter values In NM72 NONMEM can recover from many of these errors but there may be still occasion where such domain errors can terminate the analysis Here are some suggestions to provide a more robust user model that protects against domain errors or floating point overflows or allows NONMEM to reject these positions of eta that cause them and continue the analysis If it is impossible to calculate the prediction due to the values of parameters thetas or etas from NONMEM then the EXIT statement should be used to tell NONMEM that the parameters are inappropriate The EXIT statement allows NONMEM to reject the present set of etas by setting an error condition index which is in turn detected by classical NONMEM algorithms as well as the Monte Carlo algorithms With the NOABORT switch of the SEST statement set NONMEM may then recover and continue the analysis For example if you have an expression that uses LOG X You may wish to flag all non positive values and let NONMEM know when the present eta values are unacceptable by inserting IF X 0 0 EXIT LOG X On some occasions you may need to have the calculations complete then
196. ethod that was used classical or EM Monte Carlo to obtain the thetas omegas and sigmas in the last SEST step STABLE parameters are estimated based on a post hoc evaluation of the etas at the mode of the posterior density position eta hat These eta hat values are identical to those evaluated during the estimation for ITS FOCE Laplace methods but differ from the conditional mean values estimated during an IMP SAEM analysis Setting FNLETA 0 prevents the post hoc analysis so that STABLE parameters are evaluated based on the eta values generated by the last iteration of the last EST method implemented which are mode of nm730 doc 60 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 posterior values for ITS FOCE Laplace and conditional means for IMP SAEM The etas after a BAYES analysis yields single sample position values of the very last iteration and have limited use Regardless of the FNLETA setting the phi and phm tables see 1 47 EST Additional Output Files Produced always output the phi eta values used for the particular method mode of posterior and approximate Fisher information based variances for ITS FOCE Laplace methods Monte Carlo assessed conditional means and conditional variances for SAEM IMP methods If you set FNLETA 2 NM73 then the estimation step is not done and whatever etas are stored in memory at the time are used in any subsequent TABLE s This has value if you loaded the individual etas f
197. f data points to that subject is then EPRED q p 9 0 2 dq Then the corresponding residual vector for observed values y is ERES y EPRED The residual epsilon variance matrix using the nomenclature in Guide I Sections E 2 may be V m diag h q Zh y or it may be the more complicated form described in section of E 4 in the case of L2 data items Then the expected residual epsilon variance assessed by Monte Carlo sampling is EV V pm 0 Q an The full variance covariance matrix of size n xn that includes residual error epsilon and inter subject eta variance contributions is EC EV 1 m EPRED f q EPRED pl 0 Q dn And is the expected population variance Monte Carlo averaged over all possible eta Then following the Guide I section E nomenclature the population weighted residual vector for subject i is EWRES EC ERES where the square root of a matrix is defined here by default as evaluated by diagonalizing the matrix and multiplying its eigenvector matrices by the square roots of the eigenvalues Selecting the WRESCHOL option obtains the square root of the matrix by Cholesky decomposition nm730 doc 44 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 ECWRES ECWRES is a Monte Carlo assessed expected weighted residual evaluated with only the predicted function evaluated over a Monte Carlo sampled range of etas with population variance Omega while residual
198. f the file per NONMEM run Each line contains information of table number problem number sub problem number super problem iteration nm730 doc 64 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 number subject number and ID This is followed by the individual probabilities at each support point of which there are NSUPP NSUPPE or NIND of them whichever is greater The line with Subject number 0 contains the joint probability of each support point the same as listed in root npd under the column PROBABILITY For each support point K the joint probability is equal to the sum of the individual probabilities over all subject numbers I Thus row of subject number I column of support K contains the individual probability IPROB LK The sum of the individual probabilities over all support points for any given line subject is equal to 1 NIND The format of the file is fixed at 1PE22 15 and cannot be changed It is intended for use in further analysis by analytical software and is designed to report the full double precision information of each probability 1 23 Introduction to EM and Monte Carlo Methods Expectation maximization methods use a two step process to obtain parameters at the maximum of the likelihood In the expectation step the thetas omegas and sigmas are fixed while for each individual expected values conditional means of the eta s and their variances are evaluated If necessary expected values of gradients
199. fault variable substitution value is available to the parafile interpreter by the time it needs to use it in the rest of the parafile In addition if a file called defaults pnm exists in the run directory it may list alternative defaults that over ride those in the parafile such as SDEFAULTS nodes 2 The defaults pnm file is expected to have only entries for SDEFAULTS and no other parafile records The order of over ride is Command line on nmfe73 script over rides defaults pnm which over rides defaults defined in parafile The advantage to this ordering is that a generic parafile file can be created for most environments A user may then over ride defaults specified in this generic parafile with his own in defaults pnm that may be more suitable to his environment Finally a user can temporarily over ride his own defaults by giving an alternative value as an nmfe73 script command option For example the 8 pnm files listed in the NONMEM run directory serve as generic parafiles that can be run for up to 8 nodes on a multi core single computer system Also in the NONMEM run directory there is an example defaults pnm file that has nodes 2 defined as a default If this file were placed in the user s run directory and the user used fpiliwini8 pnm as a parafile nmfe73 mycontrol ctl mresults res parafile fpiwini8 pnm then the number of nodes would be that given in defaults pnm nodes 2 The user may over ride this by specifying
200. fferential equation solver ADVAN s MXSTEP NM73 Additional control may be obtained by setting the maximum number of integration steps default is 10000 SPK MXSTEP 5000 ADVANO s maximum integration steps can also be controlled by this variable nm730 doc 53 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 15 EST Improvement in Estimation of Classical NONMEM Methods In pre NM7 NONMEM installations the classical first order conditional estimation methods tended to be particularly sensitive to the formation of a non positive definite Hessian matrix during the estimate of etas In NONMEM 7 if the user selects NOABORT as a SEST option most Hessian matrices will be forced to be positive definite if not already allowing the program to continue and abnormal termination of an estimation will occur less often The occasional occurrence and correction of non positive definite Hessian matrices during the intermediate steps does not typically result in erroneous results Even with the NOABORT option there is one remaining component in the NONMEM algorithm for which positive definite correction is not performed which can still cause problems at the beginning of an estimation It remains so the user may diagnose a serious problem in the setup of the estimation Should this still be a nuisance in NONMEM 7 2 0 the user may select the NOHABORT option which will perform positive definite correction at all levels of the estimation but
201. fluential etas reported as a non zero value if it is correlated with influential etas From a pure statistical stand point this is the true EBE although intuitively it may be puzzling for some users Whether NONINFETA 1 or 0 the individual s objective function will change very little if at all because NONMEM provides a corrective algorithm to assess the correct objective function But for purposes of post hoc evaluated etas one may wish to set NONINFETA depending on the desired interpretation The NONINFETA option applies only to FO FOCE Laplace The Monte Carlo and EM methods have always used even with earlier versions of NONMEM 7 the pure statistical option NONINFETA 1 FNLETA 1 default NM72 Set FNLETA to 0 if you do not want it to spend time performing the end FNLMOD which evaluates final mixture proportions for each subject in mixture models and FNLETA which evaluates final etas routines using the original algorithm after the estimation and covariance steps are completed You may want to turn this off if each objective function call takes a long time with very complex problems or large data sets NONMEM will use instead a more efficient means which has not been thoroughly vetted Be aware that certain STABLE outputs such as the traditional WRES RESI and PRED may or may not be properly evaluated if the FNLMOD and FNLETA steps are omitted Normally when you do not set FNLETA or when you set FNLETA to 1 regardless of the m
202. for SAEM and BAYES whereas S2 and S3 perform better As of NM73 if you add a P descriptor to RANMETHOD such as RANMETHOD P RANMETHOD 3P RANMETHOD 3S2P then each subject will receive its own seed path that will stay with that subject regardless of whether the job is run as a single process or parallel process This assures that stochastically similar answers will be obtained for Monte Carlo estimation methods regardless of the number of processes or different kinds of parallelization setups used to solve the problem There is additional memory cost in using this option because the seed and seed status additional internal variables of the random number algorithm that establish the seed path must be stored for each subject and for SOBOL QR sampling there may even be a reduction in speed because the random sampling algorithm has to be re set for each subject To reiterate a single job run nm730 doc 69 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 without the P descriptor will not be stochastically similar to a single job run with the P descriptor although they will be statistically similar or to any parallel job run But a single job run using the P descriptor will be stochastically similar to any parallel job run also using the P descriptor If maintaining stochastic similarity regardless of how the job is run single or any parallel profile is important to you then always set the P descriptor so RANMETHOD P at least
203. from other subjects as possible values by default this is not used ISAMPLE_M1A 0 followed by ISAMPLE M2 mode 2 iterations using the present parameter vector position as mean and a scaled variance of OMEGA as variance 10 Next ISAMPLE M3 mode 3 iterations are nm730 doc 71 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 performed in which samples are generated for each parameter separately The scaling is adjusted so that samples are accepted IACCEPT fraction of the time The final sample for a given chain is then kept The average of the isample parameter vectors and their variances are used in updating the population means and variances Usually these options need not be changed The ISAMPLE MIA method of sampling has limited use to assist certain subjects to find good parameter values by borrowing from their neighbors in case the neighbors had obtained good values while the present subject has difficulty finding good samples This mode should generally not be used and can be inaccurate if not all subjects share the same pand Q such as in covariate modeling Alternatively use mode 1A sampling at the beginning of an SAEM analysis for a few burn in iterations then continue with a complete SAEM analysis with mode 1A sampling turned off with more burn in and accumulated sampling iterations for example EST METHOD SAEM INTERACTION NBURN S500 NITER 0 ISAMPLE_M1A 2 EST METHOD SAEM INTERACTION NBURN 500 NITER 1000 ISAMPLE_M1A 0
204. g In the FPI method the manager NONMEM process has total control of loading followed by implementing all the workers and is in fact loaded before the pnm file is interpreted and acted upon With MPI the mpi system has control and the manager NONMEM program is just the first of a set of processes The mpi system is first loaded using a DOS batch file called nmmpi bat constructed by the nmfe73 script by a call to nonmem_mpi and with commands constructed from the COMMANDS entries in the pnm file The mpi program loads all the processes including the manager Therefore the manager s SCOMMANDS entry has to have all of the parameters passed to it that was entered at the nmfe73 command line by the user as shown in the example above by using For the Windows version of MPI sometimes you have to specify the full file path of the nonmem exe program when launching on a remote computer LINUX Setting up share directory and ssh on a Linux System The ssh system and share directory used to pass files between worker and manager must be set up for FPI and MPI methods if the worker computer differs from the manager computer The following instructions serve only as a guide as to how to set up the ssh system You may need to vary some of the commands to suit your environment Consult your Linux user manual as well The network files system NFS is used for the manager computer to access a network drive that points to a worker computer s local
205. gnored and this option may be used when analytic derivatives are difficult to compute e g user supplied code such as SDE 3 2 derivative method of evaluating V using numerical second derivatives of log L with respect to etas This is equivalent to using the Laplace NUMERICAL method even though FOCE may be selected When relying on numerical derivatives by using OPTMAP gt 0 or ETADER O you may need to set the SLOW option for proper estimation of FOCE or Laplace SLOW is not utilized by EM BAYES methods Note also that non Monte Carlo weighted residual diagnostics such as NWRES NWRESI CWRES CWRESI use first derivatives of F with respect to eta and the appropriate numerical derivatives will be used to assess them if ETADER gt 1 NUMDER 0 default NM73 The file root fgh is produced if the user selects SEST NUMDER 1 The file lists the numerically evaluated derivatives of Y or F with respect to eta where G I 1 partial F with respect to eta 1 G I J 1 Second derivatives of F with respect to eta i eta j H L1 partial Y with respect to eps i H i j l partial Y with respect to eps i eta j This option is useful for comparing with and checking analytic derivatives values The analytical derivatives values are stored in root agh if NUMDER 2 is selected If you want both set NUMDER 3 MCETA 0 Default NM73 0 Eta O is initial setting for MAP estimation eta optimization during FOCE LAPLACE ITS IMPMAP a
206. h obtains a uniform random variate between al and a2 If a seed a3 is given that is not O it means to initialize the seed The initialization should be done once in a series For example The following line sets the seed nmtemplate wexamplel2 nmt dummy ctl SAMPLE R 1 10000 113345 with a throw away result file dummy ctl Then one could perform a for loop in a DOS batch file to generate a series of control stream files with different starting seeds for l n in 1 1 9 do nmtemplate wexamplel2 nmt wexamplel2 n ctl SAMPLE R n000 n999 0 where for i ssn in 1 1 9 is a DOS command generating n starting at 1 incrementing by 1 and ending at 9 When n 3 for example R n000 n999 0 Will be 2 3000 3999 0 generating nm730 doc 179 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 a random number between 3000 and 3999 to be substituted wherever SAMPLE shows up in the template file wexample12 nmt The template file wexample12 nmt may contain SEST METHOD CHAIN FILE wexamplel2 txt NSAMPLE 0 ISAMPLE lt SAMPLE gt and the resulting files wexamplel2 l ctl through wexamplel2 9 ctl will contain random ISAMPLE values such as wexample12 1 ctl SEST METHOD CHAIN FILE wexamplel2 txt NSAMPLE 0 ISAMPLE 1345 wexamplel12 2 ctl SEST METHOD CHAIN FILE wexamplel2 txt NSAMPLE 0 ISAMPLE 2456 wexample12_3 ctl SEST METHOD CHAIN FILE wexamplel2 txt NSAMPLE 0 ISAMPLE 3089
207. han LT lt Less than or equal to LE lt In FORTRAN 95 the continuation marker amp must be on the line to be continued rather than at the sixth position of the continued line Fortran 77 CL THETA 6 GENDER XTHETA 7 AGE Fortran 95 CL THETA 6 GENDER amp THETA 7 AGE This affects verbatim code and user written subroutines For example an NMVI version of CCONTR would be written as follows SUBROUTINE CCONTR I CNT P1 P2 IER1 IER2 PARAMETER LTH 40 LVR 30 NO 50 COMMON ROCMO THETA LTH COMMON ROCM4 Y DOUBLE PRECISION CNT P1 P2 THETA Y W ONE TWO DIMENSION P1 P2 LVR DATA ONE TWO 1 0D 00 2 D 00 IF I LE 1 RETURN W Y Y Y THETA 3 ONE THETA 3 CALL CELS CNT P1 P2 IER1 IER2 Y W CNT CNT TWO THETA 3 ONE LOG Y RETURN END nm730 doc 20 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Whereas in NM7 it would be written as SUBROUTINE CCONTR I CNT P1 P2 IER1 IER2 USE SIZES ONLY ISIZE DPSIZE USE ROCM REAL ONLY THETA gt THETAC Y gt DV_ITM2 USE NM INTERFACE ONLY CELS IMPLICIT NONE INTEGER KIND ISIZE INTENT IN OUT I IER1 IER2 REAL KIND DPSIZE INTENT IN OU
208. he covariance step can be performed unconditionally even when an estimation terminates abnormally by specifying SCOV UNCONDITIONAL 1 13 STABLE Additional Statistical Diagnostics Associated Parameters and Output Format Requesting a Range of Etas to be Outputted Etas x y NM73 Instead of requesting each ETA specifically in a TABLE item list a range of etas may be requested ETAS 2 4 is equivalent to requesting ETA2 ETA3 and ETA4 ETAS 5 or ETAS 5 LAST is equivalent to requesting ETA 5 ETA 6 to ETA NETAS The SCAT will also interpret this syntax for example SCAT ETAS 1 2 VS ETA3 is equivalent to SCAT ETAI ETA2 VS ETA3 However unlike STABLE SCAT will ignore implied endings such as SCAT ETAS I LAST VS ETA3 And just interpret it as SCAT ETAI VS ETA3 New diagnostic items Additional types of pred res and wres values may be requested than the usual set available in NONMEM VI They may be specified at any STABLE command or SCATTER command as one would request PRED RES or WRES items If STABLE statements succeed multiple SEST nm730 doc 42 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 statements within a run the table results as well as scatter plots if requested via SSCATTER will pertain to the last analysis OBJI These are objective function values for each individual The sum of the individual objective function values is equal to the total objective function
209. he simulated values examples tdist7 ctl SPROB RUN Example 1 from samp51 SINPUT ID TIME DV CONC AMT DOSE RATE EVID MDV CMT SDATA tdist6 csv IGNORE C nm730 doc 113 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 M M M M N C v Q V AGO SGQAHHHH GO SGCQHHHHUSGOKSAKSO GASY v in ial 2 K 0 0 0 0 0 Ome SUBROUTINES ADVAN3 PK U_1 THET U 2 THET U 3 THET U 4 THET U 4 0 LA E1 QA E LA ETA 1 0 173 1A ETA 2 0 173 QA ETA 3 0 173 2A ETA 4 0 173 B ETA 5 B ETA 6 B ETA 7 B ETA 8 R CLA CLA CLB CL R V1A V1A V1B V1 R QQA QQA QQ0B OO R V2A V2A V2B V2 L 1 0E 08 CLR GT 40 0 CL VIR GT 40 0 V1 QOR GT 40 0 QQ V2R GT 40 0 v2 LRQ 1 RQ 1 RQ 1 RQ 1 0 CLR GT DEL CLRQ F VIR GT DEL V1RQ F QOR GT DEL QORQ YHA N IO Er nMNIO al J NO ooo F V2R GT DEL V2RQ L EXP EXP EXP EXP MU_4 ETA ROR F F EPS 1 STHETA 1 68338E 00 THETA 2222 OMEGA BLOCK 4 v 01 0 1 01 0 01 0 1 01 0 01 0 01 0 1 OMEGA 1 0 FIXED SIGMA pal EST TA 1 SQRT OMEGA 1A ETA 2 SQRT OMEGA TA 3 SQRT OMEGA 2A ETA 4 SQRT OMEGA SQRT SQRT SQRT SQRT MU_1 ETA 1 CLRQ MU_2 ETA 2 V1RQ MU_3 ETA 3 OORO 4 V2RQ TRANS4 1 1 2 2 3 3 4 4 B NU B NU B NU B NU R 40 R 40 R 40 R 40 Key xc EXP EXP EX
210. her items Two items of identical element and attributes are compared between the two files where the equation for comparison is between value X of xml file 1 and value Y of xml file 2 ABS X Y gt R MAX ABS X ABS Y A The OBJ_BAYES is given a special test as it has a standard deviation with it STD X Y SQRT STD X STD Y 7 ABS X Y gt R STD X Y A In the above example OBJ_BAYES 2 0 means that if the Bayes objective functions in the two files differ by more than 2 standard deviations then the difference is noted Please note that while the above test is suitable for tolerance comparison in an installation qualification setting this is not an appropriate statistical test for model comparisons To ignore an item for comparison specify 1 To specify an exact comparison use 0 0 To refer to a particular optimization method then enter METHOD SAEM for example and thereafter all entries of items pertain to that estimation method until METHOD is changed The METHOD attribute may have one of the following settings FOCE ITS IMP SAEM DIRECT BAYES The total list of items and their scope are as follows R 2 1 2 of relative error NAME DESCRIPTION DEFAULT R A GENERAL Default to most non matrix items 0 2 0 2 DIAG Diagonal elements of OMEGA SIGMA estimates 0 1 0 OFFDIAG Off diagonal elements of OMEGA SIGMA estimates 0 0 0 2 VAR Diago
211. his condition is satisfied the estimation will terminate successfully 1 19 Additional Control for MSFI record NM73 Sometimes the MSFI error check is too strict and prevents an MSF file from being utilized in a subsequent control stream file or problem This occurs particularly when using classical NONMEM methods To turn off MSFI error checking set NOMSFTEST default is MSFTEST SMSFI myfilename NOMSFTEST 1 20 Options for ESTIMATION Record for alternative MAP eta optimization methods and evaluating individual variances by numerical derivative methods for FOCE Laplace NM73 OPTMAP 0 default NM73 0 Standard variable metric Broyden Fletcher Goldfarb and Shanno BFGS optimization method used by NONMEM to find optimal eta values aka EBE CPE MAP or conditional mode estimates referred to symbolically 7 or eta hat for each subject at the mode of their posterior densities using analytical derivatives of F with respect to etas and analytical derivatives of H with respect to etas that were supplied by NMTRAN or by the user 1 Variable metric method using numerical finite difference methods for first derivatives of F with respect to etas Necessary when not all code used in evaluating F G and H for observation event records is abbreviated code some may be in verbatim code and or some portions of the computation of F G and H are evaluated in a hidden subroutine specified by SUBROUTINES OTHER and the user writ
212. ho are male under 30 2 for subjects that are female under 30 3 for subjects who are male over 30 4 for subjects who are female over 30 Any discrete numerical values will do as long as the stratifier is not a continuous variable and the subjects need not be sorted according to the stratification data item STRATF NM73 SIML BOOTSTRAP 50 SUBP 100 NOREPLACE STRAT CAT STRATF FCAT The option STRATF points to a data item that contains the fraction that should represent a category in the bootstrapped data set Without STRATF the number of subjects to be taken from a given category is proportional to the number of subjects in the base data set If you want the category to be represented at a different proportion then specify a STRATF data item in this example FCAT Suppose FCAT 0 5 for CAT 1 and 0 5 for CAT 2 as well Even though only 33 of subjects in the base data set belong to category 1 exactly 50 of subjects from group 1 will be randomly selected out of 50 total 25 and exactly 5096 of subjects will be randomly selected from group 2 25 in the formation of each bootstrap data set This allows you to alter the proportions in each category from what is in the original data det RANMETHOD nISImIP NM73 As of NM73 the RANMETHOD option is available for the SIM record to use alternative random numbers generators default is NONMEM s traditional one number 4 SIML RANMETHOD nISImIP Where n is the random number generator typ
213. ial values of THETA STHETA 1 68693E 00 1 61129E 00 8 19604E 01 2 39161E 00 INITIAL values of OMEGA SOMEGA BLOCK 4 1 65062kE 01 7 41489E 04 1 31429E 01 1 24115E 02 1 59565E 02 1 87547E 01 1 27356E 02 1 39056E 02 3 32699E 02 1 49906E 01 Initial value of SIGMA SSIGMA 5 71632E 02 7 P SETAS FILE etafile phi phi FORMAT S1PE15 7 TBLN 6 SEST METHOD 1 INTERACTION NSIG 3 PRINT 1 FNLETA 2 STABLE ID CL V1 Q V2 FIRSTONLY NOAPPEND NOPRINT FILE etafile par FORMAT 1PE13 6 STABLE ID ETA1 ETA2 ETA3 ETA4 LCL LV1 LO LV2 FIRSTONLY NOAPPEND NOPRINT FILE etafile eta STABLE ID TIME IPRED DV CPRED CWRES NOAPPEND ONEHEADER FILE etafile tab NOPRINT 1 50 Obtaining individual predicted values and individual parameters during MCMC Bayesian Analysis Usually it is enough to obtain the population parameters thetas omegas and sigmas for each accepted sample which is listed in the raw output file specified by FILE of the SEST command Occasionally one wishes to obtain a distribution of individual parameters or even predicted values This is done be incorporating additional verbatim code This is best shown by example 8 The BAYES EXTRA REQUEST is set to 1 informing NONMEM that PRED PK ERROR are to be called after an example has been accepted The sample is indicated as accepted when NONMEM sets BAYES EXTRA to 1 An IF block can be written by the user to for example write the individual parameters in a separate file as shown in example 8 or the
214. iances phc For parameters not MU referenced phi i eta i When a classical method is performed FOCE Laplace then mode of posterior eta i are printed out along with their Fisher information first order expected value for FOCE second order for Laplace assessed variances etc For ITS these parameters are the modes of the posterior density with first order approximated expected variances or second order variances if SEST METHOD ITS LAPLCE is used For IMP IMPMAP SAEM methods they are the Monte Carlo evaluated conditional means and variances of the posterior density For MCMC Bayesian they are random single samples of phi as of the last position Their variances are zero Individual objective function values obji are also produced root phm NM72 Individual phi eta obji parameters per sub population This file is only produced in MIXTURE problems The conditional variances in the root phi and root phm files can represent the information content provided by a subject for a given eta or phi For example if data supplied by the subject is rich then the variance tends to be smaller If little data is supplied by the subject for that eta then the conditional variance will approach its omega In fact a subject s shrinkage can be evaluated as follows ETAshrinkage 100 1 1 etc j j Omega j j nm730 doc 121 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Or ETAshrinkage 100 1
215. iances to parameters were removed because the FOCE had difficulty estimating a large parametered problem and so it was an artificial constraint to begin with EM methods are much more robust and are adept at handling large full block OMEGA s so you may want to incorporate as many etas as possible when using the EM methods You should Mu reference as many of the THETA s as possible except those pertaining to residual variance which should be modeled through SIGMA whenever possible If you can afford to slightly change the theta eta relationship a little to make it MU referenced without unduly influencing the model specification or the physiological meaning then it should be done When the arithmetic mean of an ETA is associated with one or more THETA s in this way EM methods can more efficiently analyze the problem by requiring in certain calculations only the evaluation of the MU s to determine new estimates of THETAs for the next iteration without having to re evaluate the predicted value for each observation which can be computationally expensive particularly when differential equations are used in the model For those THETA s that do not have a relationship with any ETA s and therefore cannot be MU referenced including THETA s associated with ETAS whose OMEGA value is fixed to 0 computationally expensive gradient evaluations must be made to provide new estimates of them for the next iteration nm730 doc 88 of 210 NONMEM Us
216. ical equals and NEN for numerical not equals For example DATA FILE2myfile txt IGNORE OCC EQN 1 Will filter out all records for which the data item OCC is equal numerically to 1 even if it is stored as 1 0 or 1 00e 00 etc DATA FILE myfile txt IGNORE OCC EQ 1 only filters out records for which OCC is literally 1 1 5 Invoking NONMEM NONMEM 7 3 can be invoked using one of the supplied scripts nmfe73 bat for Windows nmfe73 for Linux Unix nm730 doc 28 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 These script files take at least two arguments the control stream file name and the main report file name such as Windows nmfe73 mycontrol ctl myresults res Unix nmfe73 mycontrol ctl myresults res The control stream file name is passed to NONMEM as its first argument Write and print statements supplied by the user in verbatim code will be routed as follows Unit prints to console Unit 6 prints to report file WRITE or PRINT to console WRITE 6 to report file If you wish to reroute all console output to a file the execution statement could have a redirection added to it Windows nmfe73 mycontrol ctl myresults res console txt Linux nmfe73 mycontrol ctl myresults res console txt To prevent NONMEM from polling the standard input for ctrl key characters a new feature described later Windows nmfe73 mycontrol ctl myresults res background console txt Lin
217. ific parameters so their Mu referencing is not used M indicates that the parameter should be Mu modeled assuming there is an association of a Mu for that theta which the program will verify and N indicates it should not be Mu modeled In the above example thetas 1 2 5 6 are MU modeled and 3 4 are not to be Mu modeled D for default indicates you want the program to decide whether to MU model useful for specifying back to a default option in a future SEST statement if the present setting is N The MUM parameter can also be used to specify which THETAS are used in a mixture problem by marking the position with an X For example MUM DDDDX Where THETA 5 is involved in mixture modeling in a MIX statement This is only necessary for covariate dependent mixture models such as SMIX IF KNOWGENDER 1 THEN IF GENDER 1 THEN P 1 1 0 P 2 0 0 ELSE P 1 0 0 P 2 1 0 nm730 doc 91 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 ENDIF ELSE P 1 THETA 5 P 2 1 THETA 5 ENDIF and it guarantees that the new estimation methods are aware of the proper parameters An alternative method for specifying MU modeled parameters is by using the following syntax MUMczv n v2 n2 vs n3 Where v refers to a letter N M D or X and n refers to a number list For example to specify thetas 3 5 through 8 to not be MU modeled theta 2 is a population mixture p
218. ified by the last two lines where MU_5 ETA 5 and MU_3 ETA 3 are expressed This action does not change the model in any way It is better to have a linear relationship between all thetas and MU s as we shall see below MU 5 THETA 1 THETA 2 LOG AGE MU 3 THETA 3 CL EXP MU_5 ETA 5 V EXP MU_3 ETA 3 The above parameterization would also entail log transforming initial values of THETA 1 and THETA 3 If the model is formulated by the traditional typical value TV mean followed by individual value then it is straight forward to add the MU_ references as follows TVCL THETA 1 AGE THETA 2 CL TVCL EXP ETA 5 TVV THETA 3 V TVV EXP ETA 3 MU_3 LOG TVV MU 5 LOG TVCL This also will work because only the MU x equations are required in order to take advantage of EM efficiency It is not required to use the MU variables in the expression EXP MU_5 ETA 5 since the following are equivalent CL TVCL EXP ETA 5 EXP LOG TVCL ETA 5 EXP MU_5 ETA 5 but it helps as an exercise to determine that the MU_ reference was properly transformed in this case log transformed so that it represents an arithmetic association with the eta Again it is preferable to re parameterize so that the MU s are linear functions of all thetas LTVCL THETA 1 THETA 2 LOG AGE CL EXP LTVCL ETA 5 nm730 doc 86 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 LTVV THETA 3
219. il none of the parameters change by more than NSIG significant digits For the COV step the step size for evaluating the R matrix central difference second derivative is set to SIGL 4 which according to numerical analysis yields the optimal precision of SIGL 2 for the second derivative terms If only the S matrix is evaluated central difference first derivative then the step size for it is set to SIGL 3 If the user does not specify SIGL or sets SIGL 100 then the optimization algorithm will perform the traditional NONMEM VI optimization which as discussed above may not be ideal For forward finite difference h is set to NSIG precision For central finite difference h is set to NSIG precision For forward second order difference h is set to NSIG precision The individual fits for evaluating optimal eta values will be maximized to a precision of SIGL 10 nm730 doc 56 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Optimization of population parameters occurs until none of the parameters change by more than NSIG significant digits For the COV step the step size for evaluating the R and S matrix is set to NSIG as is done in NONMEM VI This is far from optimal particularly for analyses requiring numerical integration and is often the cause of the inability to evaluate the R matrix Command syntax Example SEST METHOD 1 INTERACTION SIGL 9 NSIG 3 To see the advantage of properly setting NSIG TOL and SIGL con
220. ines NM73 in section 1 4 Expansions on Abbreviated and Verbatim Code NM72 NM73 More user functions for use in abbreviated code may be defined using FUNCA through FUNCI See Guide VIII Additional functions MIN MAX MOD and GAMLN may be used in abbreviated code See MIN MAX Functions NM73 MOD Function NM73 and GAMLN Function NM73 in section I 4 Expansions on Abbreviated and Verbatim Code NM72 NM73 ATOL now also acts on ADVANO s differential equation solver where by default absolute significant digits accuracy absolute tolerance is 12 Enhanced selection methods from CHAIN records for use in multiple sub problems For each sub problem population parameters may be randomly with or without replacement or sequentially selected from a chain file See SELECT option in 1 48 Method for creating several instances for a problem starting at different randomized initial positions EST METHOD CHAIN and CHAIN Records Total CPU time is reported in the NONMEM report file Tag CPUT and in the root cpu file See ZCPUT nm73 in section 1 45 Format of NONMEM Report File and root cpu NM73 in section 1 47 EST Additional Output Files Produced Analytical and numerical derivatives of predicted and residual variance values with respect to eta may be outputted See NUMDER 0 default NM73 in 1 20 Options for ESTIMATION Record for alternative MAP eta optimization methods and evaluating individual variances by numerical derivative me
221. ing a Random Method and Other Options for Simulation NM73 Example control stream files demonstrating how to model population densities of individual parameters that are t distributed See section 44 Model parameters as log t Distributed in the Population NM73 Option to use Nelder Mead optimization for obtaining best fit individual etas particularly useful to improve robustness for importance sampling See OPTMAP in section 1 20 Options for ESTIMATION Record for alternative MAP eta optimization methods and evaluating individual variances by numerical derivative methods for FOCE Laplace NM73 Option to use either eigenvalue square root or Cholesky square root algorithms for assessing weighted residual diagnostics See WRESCHOL in section 1 13 TABLE Additional Statistical Diagnostics Associated Parameters and Output Format Option to have etabar and eta shrinkage information include only subjects which influence the etas Furthermore you may specify certain etas of particular subjects to be excluded or specify certain etas of certain subjects to be included from the average eta shrinkage assessment by using a reserved variable ETASXI in the PK or PRED section An alternative eta shrinkage evaluation using empirical Bayes variances EBVs or conditional mean variances are now also reported See information on shrinkage in section 1 45 Format of NONMEM Report File and information on the shk and shm files in 1 47 EST Additio
222. ion as well as Table outputs without significantly slowing down the estimation See L61 finedata Utility Program NM73 See also the examples section of on line help and guide VIII on using the INFN routine to create interpolated values The infnl example has been completely rewritten The infn2 and finel examples are new A utility program to fill in substitution variables in template control stream files See 1 62 nmtemplate Utility Program NM73 New command line options tprdefault and maxlim are provided for more dynamic assessment of needed memory allocation Furthermore the dynamic memory allocation has been made even more efficient in assessing memory requirements See 1 6 Dynamic Memory Allocation NM72 and I 7 Changing the Size of NONMEM Buffers The various random number generating techniques including Sobol quasi random sampling with scrambling have been expanded for use with SAEM BAYES simulations and Monte Carlo assessed population diagnostics See the descriptions on RANMETHOD in 1 13 TABLE Additional Statistical Diagnostics Associated Parameters and Output Format 1 25 Monte Carlo Importance Sampling EM and Error Reference source not found In addition an option to have each subject retain their own seed path is available so that near identical estimation results are obtained for Monte Carlo methods in single process or parallelized process problems See the RANMETHOD item and the P descriptor in 1 25 Monte Carlo Imp
223. iq bat repeatedly calls dif bat Remember to modify the dir option in iq bat to point to the actual NONMEM installed directory Also modify dif bat and iq bat as needed for your particular environment The iq bat script will return a total differences count among all the example files This is a convenient way of automating an installation qualification 1 61 finedata Utility Program NM73 The utility program finedata in the util directory will augment an NM TRAN data file to incorporate additional non observation time values spaced at regular increments so that when a table is generated NONMEM can fill these records with predicted values from which smooth prediction curves may be plotted The syntax is as follows finedata fineplot ctl where util fineplot ctl is an example control stream file with special commands for the finedata program The fineplot ctl example is extracted from part of example6 ctl SPROB RUN example6 from r2compl SINPUT C SET ID JID TIME DV CONC DOSE AMT RATE EVID MDV CMT SDATA example6 csv IGNORE C SFINEDATA TSTART 0 TSTOP 50 NEVAL 100 AXIS TIME LIN CMT 1 3 FILE example6b csv The only records that finedata pays attention to is INPUT from which it obtains the column names DATA from which it obtains the input data file SFINEDATA which contains instructions of how to fill in with additional fine increment time records and PROB by which problems are separat
224. ision xtl mydifferences txt where myprecision xtl is a precision specification or control file Default delimiter is space and default control file is xml compare xtl It is useful to redirect difference results to a file in this example mydifferences txt The control file can be quite elaborate but it allows specification of various precision values for the many different types of values in the NONMEM report XML file and to ignore certain entries as well An example xml compare xtl file is in the util directory and has the following contents SIGNORE monitor elapsed time datetime covariance status termination status nonmem version parallel est parallel cov PRECISION GENERAL 0 2 0 2 OBJ BAYES 2 0 0 0 OBJ SAEM 0 100 0 OBJ ITS 0 5 0 OBJ IMP 0 10 0 OBJ F 0 5 0 DIAG 0 3 0 OFFDIAG 0 0 5 COR 0 0 0 3 VAR 0 3 0 1 COV 1 0 EIGENVALUES 2 0 0 OBJ_DIRECT 0 100 0 correlation o 1 0 INVCOVARIANCE O 1 INVCOVARIANCE D 1 etashrink 0 20 epsshrink 0 10 METHOD DIRECT ALL 1 METHOD SAEM epsshrink 0 20 The SIGNORE record will ignore all elements with the substrings that are listed or just a specific attribute of an element such as nonmem version nm730 doc 169 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Under the PRECISION record a GENERAL R A can be given for most items where R is the relative tolerance and A is the absolute tolerance Following the GENERAL specification tolerances may be specified for ot
225. itial value of SIGMA SIGMA Qo PLP SLEVEL SID 4 1 5 2 6 3 SEST METHOD ITS INTERACTION PRINT 1 NSIG 2 NITER 500 SIGL 8 FNLETA 0 NOABORT CTYPE 3 MCETA 0 SEST METHOD IMP INTERACTION PRINT 1 NSIG 2 NITER 500 SIGL 8 FNLETA 0 NOABORT CTYPE 3 MCETA 0 nm730 doc 111 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 ISAMPLE 300 MAPITER 0 SEST METHOD SAEM INTERACTION PRINT 10 NSIG 2 NITER 100 SIGL 8 FNLETA 0 NOABORT CTYPE 3 MCETA 0 ISAMPLE 2 CONSTRAIN 0 SEST METHOD IMP EONLY 1 INTERACTION PRINT 1 NSIG 2 NITER 5 SIGL 8 FNLETA 0 NOABORT CTYPE 3 MCETA 0 ISAMPLE 300 MAPITER 0 SEST METHOD BAYES INTERACTION PRINT 10 NSIG 2 NBURN 1000 NITER 500 SIGL 8 FNLETA 0 NOABORT CTYPE 3 SEST METHOD 1 INTERACTION PRINT 5 NSIG 2 NBURN 1000 NITER 500 SIGL 10 FNLETA 0 NOHABORT SLOW NONINFETA 1 MCETA 20 SCOV MATRIX R UNCONDITIONAL SIGL 10 Notice in all of the above examples FNLETA 0 is set so that the etas reflect what were used in the estimation If FNELTA 0 is not set super ID eta values outputted using TABLE will incorrectly differ with each subject rather than averaged for each LEVEL item value 1 44 Model parameters as log t Distributed in the Population NM73 Sometimes one may suspect that PK PD model parameters are actually log t distributed among the population with degrees of freedom NU instead of the usual log normal distributed To simulate such data for a two compartment model as an example consider the following control strea
226. ity modeling NM T73 ecce eee eee eee eee eee eee enean nu 24 DO WHILE enhancement NM7T3 ee eeeeeeee ecce ee eese een se sese e eese eaeasa sese see tese sess sese see ee eeeao 24 Subscripted Variables Enhancement NM73 eee ecce e eee esee ee ee ee seen seen see ea seen sueco 25 Autocorrelation CORRL2 NM73 e eeeeeeeeee ee eese seta sane sesto tete tasa sese eee tese ease sese eee tetas 25 MOD Function INMT73 iis ceuab ac ERE Corea beH QM REE PV Le ee phone Een a Ee n be PEE ERREUR FERRO soas PME EU CAE RE ELE FERRE 25 MIN MAX Functions NM T3 ais festes tbt potens eae Eo EYENE REN RR REPETI VENERE CERE HE LEN EXE RR EE REY EE OR EORR XR D 26 GANIEN Function NM 73 eiiis bbcive svkva eon Poker khe FERE EERGEPP CM F SERRE REC PEREERRR AMOR EXER OPE VER naR arp 26 Declaring Reserved Variables NM773 eee e ecce eee ee eee eee ee ee ee eene eese totns esee ease sesto au 26 Numerical Equality Comparison for IGNORE option in DATA Record NM73 28 1 5 Invoking NONMEM 28 1 6 Dynamic Memory Allocation NM72 ceeeeeeeeeeeeeeeeenennnnnnnn 30 I 7 Changing the Size of NONMEM Buffers eeeeeeeeeeeeeeeeeenn ene 35 1 8 Multiple RUINS asc cra ebex oa Ege rb Ga Ex ra Ea ELM Fu Y as dau Fe XY Gute fea e e du E Dai Fi dd 39 1 9 Improvements in Control Str
227. ity of selecting which items are to be compared and with what precision the xml_compare program can be used for batch processing installation qualification procedures in comparing NONMEM results of a test run against a reference run All results given in the standard NONMEM output file are also reported in the XML file For example you may wish to compare your results for example against the results given in the examples directory of your NONMEM installation run from your run directory or a special installation qualification directory you may have set up Nmfe73 examplel ctl examplel res xml compare nonmem7 2 0 examples examplesl xml examplel xml examplel xtl examplel dif nm730 doc 171 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 examplel xtl would be a file you may have modified from xml compare xtl to suit your installation qualification needs These xtl files are listed in the examples directory and are simply replicates of xml compare xtl You may change these for each example problem as needed The file examplel dif will contain a list of differences if any Available in the util directory are some example batch processing installation files that will execute examplel through example101 then perform an installation qualification on these results files against the ones in NONMEM s examples directory Call example bat this will take many hours Call iq bat this will take 10 minutes The
228. iven in the root phi file NPD is the correlated or non decorrelated NPDE value Also whole record format options are now available LFORMAT and RFORMAT See section 1 13 TABLE Additional Statistical Diagnostics Associated Parameters and Output Format Native parameters are intermediately printed to the console during classical estimation along with scaled parameters and gradients nm730 doc 17 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Alternative convergence criterion for FO FOCE Laplace See Section 1 18 Alternative convergence criterion for FO FOCE Laplace NM72 S Matrix evaluation of Variance covariance Allowed when NOPRIOR 1 If SEST NOPRIOR 1 is set and COV MATRIX S is set NONMEM will evaluate the variance covariance matrix unlike in earlier versions of NONMEM 7 Three digit limitation indexed Variables The limitation of number of digits expressing the index to thetas etas Omegas Mus and Sigmas has been increased from 2 1 99 to 3 1 999 In addition the following bugs have been fixed that were in NONMEM 7 1 2 1 With very large problems of more than to 180 estimated parameters thetas omegas and sigmas the eigenvalues list with two sets of column labels 2 When the number of records in a subject exceeds 250 a stack overflow in the Intel version of NONMEM may occur 3 On occasion after an analysis with SAEM with a very complex problem estimation of objective function with IMP or IMPMAP
229. lar SID value This is suitable if there are not too many distinct values of the super ID data item otherwise the number of thetas can become very large and the analysis may take a considerable amount of time This analysis method could be performed in earlier versions of NONMEM but the many thetas that needed to be mapped with the different levels could make NMTRAN the code quite large and tedious to write Fortunately NM73 comes with a series of substitution variable techniques and short hand entries for initial values and this method is now easier to program in NMTRAN Here is an example to code using separate thetas pertaining to each value of the SID data item example superid3 6 SSIZES LTH 60 nm730 doc 108 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SPROB RUN SINPUT C ID TIME DV AMT RATE EVID MDV CMT ROWNUM SID TYPE L2 SDATA superid3 6 csv IGNORE C SSUBROUTINES ADVAN2 TRANS2 SABBR REPLACE THETA SID KA THETA 4 to 19 SABBR REPLACE THETA SID CL THETA 20 to 35 SABBR REPLACE THETA SID V THETA 36 to 51 SABBR DECLARE DOWHILE I SABBR DECLARE INTEGER NSID PK MU 1 THETA 1 MU 2 THETA 2 MU 3 THETA 3 NSID 16 THSUM KA 0 0 THSUM CL 0 0 THSUM V 0 0 1 DO WHILE I lt NSID THSUM_KA THSUM_KA THETA I 3 THSUM CL THSUM CL THETA I 19 THSUM V THSUM V THETA I 35 I 1 ENDDO F SID lt NSID THEN KA DEXP MU_1 ETA 1 THETA SID_ KA CL DEXP MU_2 ETA 2 THETA SID CL V DEXP
230. letter refers to SIGMA 1 1 m 2 refers to SIGMA 2 2 etc going along the diagonal of SIGMA Not all thetas and sigmas need to be designated If just the Thetas are designated for example then the designations for SIGMA are assumed to be D For example for Y IPRED CMT 1 IPRED GAMMA EPS 1 2 CMT IPRED EPS 2 And with no correlation set between SIGMA 1 1 and SIGMA 2 2 then both SIGMA 1 1 and SIGMA 2 2 will be Gibbs sampled Mixed homoscedastic heretoscedastic residual errors are not Gibbs sampled nm730 doc 92 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Y IPRED IPRED EPS 1 EPS 2 GRD DDDDDDSSN The S and D specification are used only for Monte Carlo EM methods The S specification is optional and can improve the speed of IMP IMPMAP and SAEM methods Sometimes users model parameters that could have been a Sigma parameter but model them as Theta parameters instead such as Y IPRED THETA 7 IPRED EPS 1 THETA 8 EPS 2 These theta parameters are therefore Sigma like and are typically not MU referenced To have the S designation these thetas are not allowed to be involved in evaluating the predicted function IPRED Specifying theta parameters 7 and 8 as sigma like in this example note 7 and 8 position of S in the GRD option setting indicates to the program that when it evaluates forward difference partial derivatives to these thetas which it must when etas are not associated with
231. lied STHETAR THETAR EXP THETA Or STHETAR THETAR 1 NTHETA THETAPR 1 NTHP EXP THETA 1 NTHETA EXP THETAP 1 NTHP le e The code in THETAI and THETAR is verbatim code and is transferred to the FORTRAN compiler without interpretation An example is shown with thetair ctl SPROB RUN From Example 1 SINPUT C SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT SDATA examplel csv IGNORE C SSUBROUTINES ADVAN3 TRANS4 STHI THETA 1 NTHETA DLOG THETAI 1 NTHETA THETAP 1 NTHP DLOG THETAPI 1 NTHP STHR nm730 doc 100 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 H I DEXP TH ETAR 1 NTHETA ETAPR 1 NTHP H I Uo U gom om on oam E Hh eh Du H Di NQGOKSQOEBE EEN PNM _ PE T F I Ur w ROR Y F F EPS 1 Initial value STHETA 7 38905 INITIAL values SOMEGA BLOCK 4 Initial value SSIGMA 0 6 7 LP SPRIOR NWPRI DEXP THETA s of THETA 6099 X4 of OMEGA VALUES 0 2 of SIGMA ETA 1 NTHETA P 1 NTHP 0 001 prior information on thetas STHETAP variance to th STHETAPV BLOCK SOMEGAP BLOCK 4 0 2 FIX EGAPD EST METHOD ITS eta priors 7 389056099 FIX X4 4 FIX VALU INTERACTION NOABORT
232. list SEXCLUDE 5 7 exclude nodes 5 7 ror SEXCLUDE ALL SINCLUDE 1 4 6 SNAMES Give a name to each node which is displayed 1 MANAGER 2 WORKER1 3 WORKER2 COMMANDS each node gets a command line used to launch the node session first one launches manager version l mpirun PWD n 1 nonmem This launches a worker process on the manager s computer 2 wdir PWD nonmem wrk mpi n 1 nonmem This launches a worker process on a separate computer 3 wdir home myself share workerl n 1 host any worker nonmem SDIRECTORIES 1 NONE FIRST DIRECTORY IS TH 2 nonmem wrk mpi NEXT SET A 3 mnt share workerl COMMON DIRECTORY iR directories D El H pa s o Jj A SIDRANGES USED IF PARSE TYPE 3 1 1 50 2 51 100 You will want to modify the pnm file for your particular environment and use some of the other options available in setting up the mpiexec mpirun command line Unlike FPI the MPI system can only use the starting PARFILE specified at the command line and it may not be easily switched later in the control stream All processes remain resident throughout the entire job although it will honor requests of parafile off or parafile on individual SEST records which allows you to have control of which estimation method will use parallel processing nm730 doc 157 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Earlier
233. lte if not completed by then deassign node and have manager complete it paraprint 1 print to console the parallel computing process Can be modified at run time with ctrl B toggle Regardless of paraprint setting control stream log always records parallelization progress transfer type 0 for file transfer unloading and reloading workers with each estimation transfer type 1 for mpi transfer type 2 for file transfer maintaining a single loaded process throughout the run HE EXCLUDE INCLUDE may be used to selectively use certain nodes out of a large list EXCLUDE 5 7 exclude nodes 5 7 or EXCLUDE ALL INCLUDE 1 4 6 SNAMES Give a label to each node for convenience 1 2 3 MANAGER WORKER1 WORKER2 COMMANDS each node gets a command line used to launch the node session 1 r 2 3 D 1 2 3 Command lines must be on one line for each process command not needed for node 1 manager NONE following is a launch on a core of the manager computer Beolaunch sh is a simple script available from the NONMEM run directory beolaunch sh wrk ftif nonmem workerl out following is a launch on a remote worker computer ssh n any computer cd home myself share workerl nonmem workerl out amp IRECTORIES NONE FIRST DIRECTORY IS THE COMMON DIRECTORY wrk ftif NEXT SET ARE THE WORKER directories mnt share workerl
234. lti core computer try the following Copy foce parallel ctl and examplel csv from the NONMEM examples directory mpiwini8 pnm from the NONMEM run directory and mpiexec exe from the NONMEM run directory into your standard run directory Then execute the following from your standard run directory Nmfe73 foce parallel ctl foce parallel res parafile mpiwini8 pnm nodes 4 where the values of nodes should be no greater than the number of cores available on your computer For instructional purposes a typical structure of a PARAFILE is listed below that would be used for NONMEM on Windows using MPI note the setting of TRANSFER TYPE 1 GENERAL NODES 2 PARSE TYPE 3 PARSE NUM 200 TIMEOUTI 60 TIMEOUT 10 PARAPRINT 0 TRANSFER TYPE 1 COMPUTERS 2 nm730 doc 147 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 NODES number of nodes that is process whether cores or computers SINGLE node NODES 1 MULTI node node means process whether cores or computers NODES gt 1 WORKER node NODES 0 parse num number of subjects to give to each node parse type 0 give each node parse num subjects parse type 1 evenly distribute numbers of subjects among available nodes parse type 2 load balance among nodes parse type 3 assign subjects to nodes based on idranges parse type 4 load balance among nodes taking into account loading time This setting of parse type will assess ideal number of no
235. m file examples tdist6_sim ctl SPROB RUN Example 1 from samp51 SINPUT C SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT SID SDATA tdist sim csv IGNORE C SSUBROUTINES ADVAN3 TRANS4 SPK MU_1 THET MU 2 THET MU 3 THET MU 4 THET NU 4 0 CLA ET V1A ET QQA ET V2A ET CLB ET V1B ET SQRT OMEGA SQRT OMEGA SQRT OMEGA SQRT OMEGA 1 1 2 2 3 3 4 4 Sp DD SS V2B ETA 8 CLR CLA CLA CLB CLB NU V1IR V1A V1A V1B V1B NU QOR QQA QQA QQB QOB NU V2R V2A V2A V2B V2B NU CL EXP MU_1 ETA 1 SQRT EXP CLR 1 0 CLR V1 EXP MU_2 ETA 2 SORT EXP VLR 1 0 V1R Q EXP MU_3 ETA 3 SQRT EXP QOR 1 0 QQR V2 EXP MU_4 ETA 4 SQRT EXP V2R 1 0 V2R S1 V1 ERROR Y F F EPS 1 Initial values of THETA STHETA 1 68338E 00 1 58811E 00 8 12694E 01 2 37435E 00 INITIAL values of OMEGA SOMEGA BLOCK 4 0 03 0 01 0 03 0 006 0 01 0 03 0 01 0 006 0 01 0 03 SOMEGA 1 0 FIXED 1 0 FIXED 1 0 FIXED 1 0 FIXED nm730 doc 112 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SIGMA 0 01 SSIMULATION 567811 NORMAL 2933012 UNIFORM ONLYSIMULATION SUBPROBLEMS 1 STABLE ID TIME CONC DOSE RATE EVID MDV CMT ETA1 ETA2 ETA3 ETA4 CL V1 Q V2 NOAPPEND ONEHEADER FILE tdist6 csv NOPRINT The data file produced tdist6 csv will have CL V1 Q and V2 t distributed among the 100 subjects with NU degrees of free
236. m or folder will be mounted as Volumes mydir E g in a terminal window ls Volumes mydir Enabling ssh with no password on MAC OS X Select the Apple menu System Preferences Sharing Remote Login The instructions for Linux using ssh keygen should work on Mac OS X There may be an interaction with keychain and this may be problematic If ssh n cannot be made to work you can use the workaround for mpdboot described in the MPICH2 Installer s Guide See start the daemons by hand on page 7 of mpich2 1 2 1 installguide pdf Disabling Open MPI commands on MAC OS X The Open MPI commands that are supplied with Mac OS X must be disabled The following is suggested sudo s cd usr bin mkdir default mpi mv mpi default mpi exit If this is not done this message may appear Unfortunately this installation of Open MPI was not compiled with Fortran 90 support As such the mpif90 compiler is non functional Installing MPICH2 on MAC OS X MPICH2 must be compiled and installed for Mac OS X Please look at mpich2Z README vin mht and the other documents First see what kind of binaries have been installed e g cd opt nm72 mpi mpi ling or mpi lini with ifort file mpi o You will see either of the following mpi o Mach O 64 bit object x86 64 mpi o Mach O object 1386 1386 indicates 32 bit binaries nm730 doc 160 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Suggested
237. m two sub populations Project Name nm7examples Project ID NO PROJECT DESCRIPTION PROB RUN example4 from adltrim2t SINPUT C SET ID JID TIME CONC DV DOSE AMT RATE EVID MDV CMT VC1 K101 VC2 K102 SIGZ PROB DATA example4 csv IGNORE C S SUBROUTINES ADVAN1 TRANS1 MIX P 1 THETA 4 P 2 1 0 THETA 4 NSPOP 2 Prior information setup for OMEGAS only PRIOR NWPRI NTHETA 4 NETA 3 NTHP 0 NETP 3 NPEXP 1 PK MU_1 THETA 1 MU_2 THETA 2 MU_3 THETA 3 V DEXP MU_1 ETA 1 K10M DEXP MU_2 ETA 2 K10F DEXP MU_3 ETA 3 Q 1 IF MIXNUM EQ 2 Q 0 K Q K10M 1 0 Q K10F S1 V ERROR Y F F EPS 1 STHETA 1000 0 4 3 1000 0 MU 1 1000 0 2 9 1000 0 MU 2 1000 0 0 67 1000 0 MU 3 0 0001 0 667 0 9999 P 1 OMEGA BLOCK 3 04 p 0 01 f 027 p 0 01 f 0 001 f 0 06 p Degrees of Freedom defined for Priors STHETA 3 FIX Prior OMEGA OMEGA BLOCK 3 0 05 FIX 0 0 0 05 0 0 0 0 0 05 SIGMA 0 01 p nm730 doc 193 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 EST METHOD ITS INTERACTION NITER 30 PRINT 5 NOABORT SIGL 6 FILE example4 ext NOPRIOR 1 CTYPE 3 CITER 10 CALPHA 0 05 EST METHOD IMP INTERACTION NITER 20 ISAMPLE 300 PRINT 1 NOABORT SIGL 6 NOPRIOR 1 SEST NBURN 500 NITER 500 METHOD SAEM INTERACTION PRINT 1 SIGL 6 ISAMPLE 2 NOPRIOR 1 EST METHOD IMP INTERACTION EONLY 1 MAPITER 0 NITER 20 ISAMPLE 3000 PRINT 1 NOABORT SIGL 6 NOPRIO
238. matching the block pattern of the initial SIGMAS Additional examples of setting up prior information for various problems are shown in the example problems listed at the end of this document As of NM73 you can use more informative names as follows THETAP for theta priors THETAPV for variance to theta priors OMEGAP for omega priors OMEGAPD for degrees of freedom or dispersion factor for omega priors SIGMAP for SIGMA priors SIGMAPD for degrees of freedom or dispersion factor for SIGMA priors This allows you to intersperse these records at will in the control stream files but it also gives NMTRAN an alternative source for values to NTHETA NETA NTHT NETP NEPS and NEPP that is typically given in the PRIOR NWPRIOR record However if these values are also listed in PRIOR NWPRI then these values are chosen over what is surmised from the informatively labeled theta omega sigma records Thus the above control stream file could be structured as follows with the various records in any order and a shortened PRIOR record PRIOR NWPRI Prior information of THETAS NTHP 4 of them STHETAP 2 0 FIX 2 0 FIX 2 0 FIX 2 0 FIX STHETA 2 0 2 0 4 0 4 0 Initial Thetas SOMEGA BLOCK 4 Inital Parameters for OMEGA 0 4 0 01 0 4 0 01 0 01 0 4 0 01 0 01 0 01 0 4 Set degrees of freedom of SIGMA Prior one value per SIGMA block nm730 doc 81 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SIGMAP
239. md exe C nonmem exe load on a core of the same computer as manager Wr psexec d w worker2 cmd exe C nonmem exe load on a core of a different computer than manager 4 psexec any computer d w c share worker3 cmd exe C nonmem exe SDIRECTORIES Names of directories as a manager sees them NONE FIRST DIRECTORY IS THE COMMON DIRECTORY Make it NONE if no common directory is to be used This is the best option 2 workerl NEXT SET ARE THE WORKER directories 3 worker2 4 w share worker3 This directory is on a different computer from manager n SIDRANGES USED IF PARSE TYPE 3 1 1 50 2 51 100 You may load the problem as follows nmfe73 mycontrol ctl mycontrol res parafile fpiwini8 pnm Strictly speaking drive letter mapping on the manager side is not necessary One could refer to the network drive as any_computer share worker3 instead of w share worker3 in the pnm file The most versatile PARSE TYPE selections are 2 and 4 If you select PARSE_TYPE 0 make sure that PARSE_NUM gt no of subjects no of nodes otherwise the problem may not run properly If you select PARES TYPE 3 make sure all subjects are accounted for in the IDRANGES listings The NAMES record is optional If left out or if a name is not defined for a process the default name is MANAGER for position 1 WORKERI for position 2 WORKER for position 3 etc The structure of the COMMANDS
240. meters to be used in the pnm file as long as the substitution variable begins with a or lt For example you may enter at the command line of nmfe73 the following variable wd for a worker directory definition Nmfe73 mycontrol ctl mycontrol res parafile mypara pnm wd c myworker and your pnm file may contain the following loading SCOMMANDS 2 psexec d w wd ql cmd exe C nonmem exe 3 psexec d w wd q2 cmd exe C nonmem exe and DIRECTORIES 2 wd Nq1 3 wd q2 For user defined variables the value of the variable is substituted into the placeholder rather than the entire var value Then c myworker will be substituted in place of wd in the COMMANDS and DIRECTORIES entries Add as many substitution variables as you need to create a generalized pnm file To make the user substitution process even more flexible default values for these variables may be defined in case the user does not specify a value for it on the command line For example in run fpiwini8 pnm There is a section called DEFAULTS where a default value for nodes is given SDEFAULTS nodes 8 and in GENERAL nodes is used as the number of nodes SGENERAL nodes is a User defined variable NODES nodes PARSE TYPE 2 PARSE NUM 50 TIMEOUTI 500 TIMEOUT 2000 PARAPRINT 0 TRANSFER TYPE 0 nm730 doc 140 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Make sure that SDEFAULTS is placed at the head of the file so the de
241. mment line of the control stream file For example SIZES MAXIDS 230 NO 300 LTH 50 LVR 30 The following is an example of FSIZESR information from a run with CONTROLS All parameters can be changed with SIZES see resource sizes f90 for descriptions and default values except NTT NOMEG NSIGM PPDT which are always evaluated properly by NMTRAN and should not be over ridden LTH LVR LVR2 LPAR 10 LPAR3 NO MMX LNP4 LSUPP LIM7 LWS3 MAXIDS oO 8 W E cm Toes i 1 oO 1 S M2 IM3 M4 M5 IM6 M8 M M M M16 MAXRECID EG PCT PIR PD PAL MAXFCN MAXIC PG NPOPMIXMAX U Uw P EXEC m EE i oaa ER cae Et IE E 1 Ss ONO OO C50 OOO Qoo Ims COCO CNHI Ft nm730 doc 31 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 MAXOMEG MAXPTHETA MAXITER ISAMPLEMAX DIMTMP DIMCNS DIMNEW PDT LADD MAX MAXSIDL NTT NOMEG NSIGM PPDT O w w w O CO i OO CO D RN 4 W The file FSIZES is also produced that contains the same contents as the FSIZESR routine in FSUBS The FSIZES file is produced for easy reading for the user and is not used by the NONMEM system Those parameters with a O cannot be determined or are not given by NMTRAN and will default to the values hard coded in resource SIZES f90 See the file SIZES f90 itself or on line help entry for sizes for these values On occasion NMTRAN mis interprets the true scope of the r
242. mpled from a multi variate normal conditional density given the latest PHIZMU ETA values for each subject and the samples are always accepted In M H sampling the sampling density used is only an approximation so the sampled THETA values must be tested by evaluating the likelihood to determine if they are statistically probable requiring much more computation time As much as possible define the MU s in the first few lines of PK or PRED Do not define MU values in SERROR Have all the MU s particularly defined before any additional verbatim code such as write statements NMTRAN produces a MUMODEL2 subroutine based on the PRED or PK subroutine in FSUBS and this MUMODEL2 subroutine is frequently called with the ICALL 2 settings more often than PRED or PK The fewer code lines that MUMODEL2 has to go through to evaluate all the MU s the more efficient Whenever possible have the MU variables defined unconditionally outside IF THEN blocks Time dependent covariates or covariates changing with each record within an individual cannot be part of the MU equation For example MU_3 THETA 1 TIME THETA 2 should not be done Or consider MU_3 THETA 2 WT Where WT is not constant within an individual but varies with observation record time This would also not be suitable However we could phrase as MU_3 THETA 2 CL WT MU_3 ETA 3 nm730 doc 89 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 where MU 3 represents
243. n a machine with infinite precision then the code should be written so as to produce the correct result To protect against this SORT X could be converted to SORT ABS X Or SQRT SQRT X X The EXIT statement should not be used in such near zero cases It could lead to a failure in NONMEM with a message containing text such as DUE TO PROXIMITY OF NEXT ITERATION EST TO A VALUE AT WHICH THE OBJ FUNC IS INFINITE An EXIT may still be issued for values of X that are clearly negative because of erroneous inputs and you may wish to flag this calculation so that the estimation algorithm rejects this position IF X 1 0E 06 EXIT SORT ABS X Such protection codes described above need not be inserted for every LOG EXP or SORT but only if your analysis fails frequently or tends to be sensitive to initial values nm730 doc 134 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 53 Parallel Computing NM72 General Concepts of Parallel Computing If you have a run that takes a long time to estimate you may submit it for parallel computing This is the process of splitting the objective function evaluations of individual subjects among a set of computers or CPUs to speed up analysis of a particular run Only estimations SEST and covariance assessments COV are parallel processed From our tests we have found that the optimal number of processes needed depends o
244. n in phase of the MCMC Bayesian method default 4000 During this time the advance of the parameters may be monitored by observing the results in file specified by the FILE parameter and or the objective function displayed at the console The objective function progress is also written in OFV TXT and the report file Full sets of population parameters and likelihood functions are also written in the file specified with the FILE option When all parameters and objective function do not appear to drift in a specific direction but appear to bounce around in a stationary region then it has sufficiently burned in A termination test may be implemented to perform a statistical assessment of stationarity for the objective function and parameters As mentioned earlier the objective function MCMCOBJ that is displayed during BAYES analysis is not valid for assessing minimization or for hypothesis testing in the usual manner It does not represent a likelihood that is integrated over all possible eta marginal density but the likelihood at a given set of etas NSAMPLE NITER 10000 Sets number of iterations in which to perform the stationary distribution for the BAYES analysis default 10000 ISAMPLE M1 2 defaults listed ISAMPLE M1A 0 NM72 ISAMPLE M22 ISAMPLE M32 IACCEPT 0 4 These are options for the MCMC Bayesian Metroplis Hastings algorithm for individual parameters ETAS used by the SAEM and BAYES methods For Bayesian a
245. n test default Process goes through the full set of NBURN SAEM or BAYES or NITER IMP IMPMAP or ITS iterations CTYPE 1 Test for termination on objective function thetas and sigmas but not on omegas CTYPE 2 Test for termination on objective function thetas sigmas and diagonals of omegas CTYPE 3 Test for termination on objective function thetas sigmas and all omega elements nm730 doc 93 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 CTYPE 4 As of NONMEM 7 2 0 there is an alternative test for FO FOCE Laplace NONMEM will test if the objective function has not changed by more then NSIG digits beyond the decimal point over 10 iterations If this condition is satisfied the estimation will terminate successfully The traditional criterion for successful termination of a classical NONMEM method is that if all of the parameters change by no more than NSIG significant digits then successful termination results CINTERVAL Every CINTERVAL iterations is submitted to the convergence test system If CINTERVAL is not specified then the PRINT option is used as CINTERVAL If neither PRINT nor CINTERVAL are specified then default CINTERVAL is listed as 9999 which is interpreted as CINTERVAL 1 If CINTERVAL 0 NM73 then a best CINTERVAL will be found then used CITER or CNSAMP Number of latest PRINT or CINTERVAL iterations on which to perform a linear regression test where independent variable is iteration number depe
246. n the problem On one extreme if the problem contains many subjects and each subject takes a long time to evaluate because of a large number of differential equations and or a large number of dose events so that one subject takes a minute to evaluate on each function evaluation then as many cores as there are subjects would still be efficient Our parallelization algorithm does not split up the problem beyond one subject per process On the other hand if the problem takes just 0 01 second to evaluate all subjects for a function evaluation then it may not be worth using parallel processing For each function call the manager process packages a subset of subjects and sends the data to a worker process then the worker process returns its results to the manager and the manager summarizes the information from all of the workers For the next function call the procedure begins again The length of time to perform one subject s evaluation in a function call varies with the estimation method as well In importance sampling there is one function call per iteration and if you have high ISAMPLE then it can take some time to evaluate each subject Such a problem is very efficiently parallelized On the other hand BAYES analysis performs only one sample per subject per function call so it may perform a function evaluation very quickly on a single process and parallelization may not improve computation time NONMEM can parallelize across computers as w
247. nal Output Files Produced nm730 doc 10 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Subscripted variables may be used in abbreviated code with fewer restrictions on DOWHILE See section 1 4 Expansions on Abbreviated and Verbatim Code NM72 NM73 for and example on residual variance correlation and see section 1 43 Adding Nested Random Levels Above Subject ID NM73 for another use Additional reserved variables may be declared in the control stream file not natively recognized by NMTRAN Some useful but not often needed global variables may be accessed by listing them in an NMTRAN include file referenced in a control stream file which can also be used in abbreviated code See section 4 Expansions on Abbreviated and Verbatim Code NM72 NM73 Enhanced non parametric analysis methods such as extended grid of support points use of an outsize inter subject variance to obtain support points that fit outlier subjects better and built in bootstrap analysis methods for obtaining empirical confidence ranges to non parametric probability parameters See 1 22 Some Improvements in Nonparametric Methods NM73 The TRANSLATE option of the DATA record has been expanded Now any value may be given for dividing time and II values and any precision may be requested Examples are TIME 1 0000 or TIME 1 4 for formatting times in FDATA with 4 digits to the right of the decimal Or I1 0 01 6 which divides II values by 0 01
248. nals of variance of estimates 0 2 0 COV Off diagonals of covariance of estimates 0 0 2 COR Correlations 0 0 2 TABLE Table items listed in NONMEM report file GENERAL OBJ_BAYES BAYES objective function 1 0 OBJ_SAEM SAEM objective function 0 100 OBJ_ITS ITS objective function 0 2 OBJ_IMP IMP IMPMAP objective function 0 5 OBJ_DIRECT Direct sampling objective function 0 100 OBJ_F FO FOCE Laplace objective function 0 0 5 EIGENVALUES Eigenvalues 2 2 ETABAR Etabar GENERAL ETABARSE Etabar Se GENERAL ETABARPVAL Etabar Pval GENERAL ETASHRINK Eta shrinkage GENERAL EPSSHRINK EPS shrinkage GENERAL nm730 doc 170 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 NAME DESCRIPTION DEFAULT R A THETA Thetas GENERAL OMEGA D Omega diagonals DIAG OMEGA O Omega off diagonals OFFDIAG SIGMA D Sigma diagonals DIAG SIGMA O Sigma off diagonals OFFDIAG OMEGAC D Omega correlation diagonals DIAG R 2 A OMEGAC O Omega correlation off diagonals COR SIGMAC D Sigma corrlation diagonals DIAG R 2 A SIGMAC O Sigma correlation off diagonals COR THETASE Theta standard errors VAR R 2 A OMEGASE D Omega diagonal standard errors VAR R 2 A OMEGASE O Omega off diagonal standard errors COV R 2 A SIGMASE D Sigma digaonl standard errors VAR R 2 A SIGMASE O Sigma off diagonals standard errors COV R 2 A OMEGACSE D Omega correlation diagonal standard
249. nalysis the nm730 doc 76 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 MCMC algorithm performs SAMPLE Ml mode 1 iterations using the population means and variances as proposal density followed by ISAMPLE MIA mode IA iterations testing model parameters from other subjects as possible values by default this is not used ISAMPLE M1A O followed by SAMPLE M2 mode 2 iterations using the present parameter vector position as mean and a scaled variance of OMEGA as variance 10 Next ISAMPLE M3 mode 3 iterations are performed in which samples are generated for each parameter separately The scaling is adjusted so that samples are accepted IACCEPT fraction of the time The final sample is then kept Usually these options need not be changed There is only one chain of samples produced for a given NONMEM run ISAMPLE is not used for MCMC only for SAEM If you would like additional chains then create separate control stream files with different starting seed numbers ISCALE_MIN 1 0E 06 defaults for SAEM BAYES NM72 ISCALE_MAX 1 0E 06 NM72 In MCMC sampling the scale factor used to vary the size of the variance of the proposal density in order to meet the IACCEPT condition is by default bounded by ISCALE MIN of 1 0E 06 and ISCALE_MAX 1 0E 06 This should left alone for MCMC sampling but on occasion there may be a reason to reduce the boundaries perhaps to ISCALE MIN O0 001 ISAMPLE MAX 1000 After the SAEM estimation meth
250. nce ine sacer e iane lcu Sane oun etnucun uS touc one ibSesupudb s 57 1 18 Alternative convergence criterion for FO FOCE Laplace NM72 58 1 19 Additional Control for MSFI record NM73 cecus 58 1 20 Options for ESTIMATION Record for alternative MAP eta optimization methods and evaluating individual variances by numerical derivative methods for FOCE Laplace NMZS 2 1 aaea aatia aaa a aaa nbus ubzemadu uiui de usui add aaa iiti Drag 58 OPTMAP 0 default NMT73 4 ascsisetbre Inset kein ierat eipe rini ERe ebR iiri CIR Lor Ee t Nba iin ien ipa ia cius 58 EFADER O0 default NM73 siccisssthncnssshecetonsssncchpsetncasasesnanspsunsneonesenoctosbsaususeniacnssoantensecnedinns 59 NUMDER 0 defarullt INIM73 wssicesi ccciasessveceasscdsvesetapscsobscasdesccaspedsssedsuresajeadetadesteceuasacbossecences 59 MCETA 0 Default NM73 sccessxcastecdseusucesscenvedencusoaoscesrastodennensXeneaecnesoanscapnpeasanssuaasauorsatoans 59 NONINFETA 0 default NIMITS sinisscecssccssncunsaschsetascseerehossnstavasebcnshcecasasnccbneentacosemansenvesbhonse 60 ENLETA 1 default NM72 ssssansicecasssuihoceccsnsheastascopnansnessasentsdousessatnasevesen eaetseosasssansnneteasverites 60 1 21 Bootstrap Selecting a Random Method and Other Options for Simulation rem M M HR 61 BOOTSTRAP NMT73 oecesshoekenvereosveita ala even tavit snnsdvaeseveisassseevanenestnsan
251. nd sometimes IMP nm730 doc 59 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 ETA values of previous iteration is initial setting for MAP estimation or ETA 0 whichever gives lower objective function gt MCETA 1 Random samples of ETA using normal random distribution with variance OMEGA are tested Plus previous ETA is tested and ETA 0 is tested The test is whichever supplies the lowest objective function is the eta set used as initial parameters for the MAP optimization NONINFETA 0 default NM73 NONMEM has traditionally not assessed post hoc eta hat also known as empirical Bayes Estimates EBE s conditional mode etas or conditional parametric etas CPE if the derivative of the data likelihood with respect to that eta is zero for a given subject and simply specified that eta as zero This eta is called a non influential eta The true EBE is zero anyway if this eta is not correlated by an off diagonal omega element with an eta that is influential If the non influential eta Is correlated with an influential eta then the true EBE of the non influential eta will in general not be 0 When NONINFETA O the default then this traditional algorithm is in effect so that all non influential etas even those correlated with influential etas will be reported as 0 when outputted with STABLE However if NONINFETA 1 then all etas are involved in the MAP estimation regardless of their influence This will result in non in
252. nd will act on the prsame option You may skip the NMTRAN step using the trskip switch nmfe73 mycontrol ctl myresults res background trskip The trskip option is useful if you wish to modify FSUBS created by a previous run and insert extra debug lines into FSUBS and prevent your modified FSUBS from being over written by NMTRAN it will still be compiled The trskip and any one of prsame prcompile or prdefault switches may be used together I 7 Changing the Size of NONMEM Buffers The entire data set is not necessarily stored in memory at one time It may be stored in a temporary disk file and parts of it are brought into a memory buffer as needed Some other large arrays are also stored on disk files Of course memory file swapping of data set information leads to increased computer run time So the bigger the buffer size the shorter may be the run time The sizes of the NONMEM buffers are set by constants LIMI to LIMI6 The default settings of these constants are set in SIZES f90 If these constants are not adequate NONMEM will produce error messages such as the following TOT NO OF DATA RECS IN BUFFER 1 IS LESS THAN NO OF DATA RECS IN INDIVIDUAL REC NO 1 IN INDIVIDUAL REC ORDERING Unlike most of the other dynamically changeable parameters NMTRAN does not determine the most appropriate LIM value for the problem but instructs NONMEM to use the default value specified in resource SIZES f90 by d
253. ndent variable is parameter value If CITER 10 then 10 of the most recent PRINTed or CINTERVAL iterations are used for the linear regression test CITER 10 is the default CALPHA CALPHA 0 01 0 05 Alpha error rate to use on linear regression test to assess statistical significance The default value is 0 05 At each iteration the program performs a linear regression on each parameter which parameters depends on the CTYPE option if CTYPE 3 then all parameters If the slope of the linear regression is not statistically different from 0 for all parameters tested then convergence is achieved and the program stops the estimation If you complete NBURN for SAEM or BAYES methods or NITER for IMP IMPMAP or ITS methods iterations and convergence has not occurred the optimization stops or goes to the next mode anyway So if you want the termination test to properly take effect give a rather high value to NBURN 1000 10000 for SAEM BAYES or NITER 200 1000 for ITS MAP IMPMAP so you don t run out of iterations Typically consecutive importance sampling iterations tend to be nearly statistically uncorrelated and so it is reasonable to have CITER 10 consecutive iterations CINTERV AL 1 tested at the alpha 0 05 level For MCMC methods SAEM and BAYES consecutive iterations can be highly correlated so to properly detect a lack of change in parameters you may want to test every 10 to 100 iteration CINTERVAL 10 to 100 so that the li
254. ndomly Even though samples in physically close proximity in the file may have some correlation selecting randomly among the entire set assures de correlation while assuring the samples taken represent the empirical distribution of uncertainty of the parameters In general sampling is performed between the larger of ISAMPLE and the lowest iteration sample number of a raw output file and the smaller of ISAMPEND and the largest iteration number in the file So it is safe to make ISAMPEND 1000000 for example to cover most Bayesian sample set sizes If ISAMPEND is specified in the SCHAIN record then SSIML s TRUE PRIOR will be ignored nm730 doc 129 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SELECT 0 DEFAULT NM73 When SELECT 0 and ISAMPEND gt ISAMPLE then the default action for selecting between ISAMPLE and ISAMPEND is taken which for EST METHOD CHAIN is to find the one giving the best OBJ at the initial values and for CHAIN is to randomly select a sample with replacement as described above Alternative actions may be obtained which apply to both record types SELECT 1 the sample is selected sequentially from ISAMPLE to ISAMPEND with each new use of CHAIN SIML with multiple sub problems for the given problem and with each new EST METHOD CHAIN with multiple sub problems and across problems When ISAMPEND is reached the sample selection begins at SAMPLE again SELECT 2 uniform random selection of sample withou
255. near regression on parameter change is spread out over a larger segment of iterations An alternative method to convergence testing is to set NBURN to a very high number 10000 monitor the change in MCMCOBJ or SAEMOBJ and enter ctrl K see section l 11 Interactive Control of a NONMEM batch Program when you feel that the variations are stationary which will end the burn in mode and continue on to the statistical accumulation mode It is better to provide a large NBURN number and end it at will with ctrl K or allow the nm730 doc 94 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 convergence tester to end it rather than to have a small NBURN number and have the burn in phase end prematurely The termination test for the Monte Carlo methods can often be very conservative and may result in very long run times even when the objective or likelihood function as well as the parameters appear randomly stationary by eye To make the termination test more liberal use one of the lower level CTYPE s CTYPE 1 or CTYPE 2 to test the more important parameters or reduce CALPHA to 0 01 or 0 001 Once the objective function is randomly stationary then often the analysis has converged statistically so CTYPE 1 is often enough Remaining parameters that appear to continue to change in a directional manner may often not have much impact on the fit This can be particularly true of covariances of OMEGAs 1 34 Use of SIGL and NSIG with the new meth
256. nly FSUBS is compiled at run time With nmfe72 NONMEM 7 2 0 or nmfe73 NONMEM 7 3 0 certain of the PREDPP files in the pr directory are also compiled at run time with the sizes and values given in prsizes f90 Thus arrays internal to PREDPP are statically allocated In contrast the NONMEM source code in Mm are precompiled and the main NONMEM arrays are allocated dynamically PREDPP source code is not pre compiled and dynamically allocated due to significant increase in run times Many compilers produce a much more elaborate binary code in order to deal with variables that are dynamically shaped which occurs with dynamically sized variables that have nm730 doc 33 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 more than one dimension to them and this slows down execution considerably with routines that are accessed very frequently such as PREDPP routines The nmfe73 script file copies the required PREDPP routines from the nonmem pr directory into a temporary folder called temp dir under the user s run directory and compiles the routines there The resulting object files are then linked with NONMEM and the nonmem executable is created The compilation of the PREDPP routines may take some time about 10 to 50 seconds If you are repeatedly running the same problem by default the nmfe73 script will skip the PREDPP recompilation It does this by testing that all of the PREDPP files listed in the file LINK LNK from the previo
257. non linear hierarchical models In Markov Chain Monte Carlo in Practice W R Gilks et al Chapman amp Hall 1996 chapter 19 pp 341 342 12 Gilks Richardson and Spiegelhalter Introducing Markov chain Monte Carlo In Markov Chain Monte Carlo in Practice W R Gilks et al Chapman amp Hall 1996 chapter 1 pp 5 8 13 Karlsson MO and Savic RM Diagnosing Model Diagnostics Clinical Pharmacology and Therapeutics 2007 82 1 17 20 14 Overgarrd RV Jonsson N Tornoe CW and Madsen H Non Linear Mixed Effects Models with Stochastic Differential Equations Implementation of an Estimation Algorithm J Pharmacokinetics and Pharmacodynamics 2005 32 1 85 107 15 Predictive Performance for Population Models Using Stochastic Differential Equations Applied on Data From an Oral Glucose Tolerance Test Moller JB Overgaard RV Madsen H Hansen T Pedersen O and Ingwersen SH J Pharmacokinetics and Pharmacodynamics 2010 37 85 08 16 Tornoe CW Overgaard RV Agerso H Nielsen H Madsen H and Jonsson EN Stochastic Differential Equations in NONMEM Implementation Application and Comparison with Ordinary Differential Equations Pharmaceutical Research 2005 22 8 1247 1258 nm730 doc 184 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 17 Bauer RJ Guzy S Ng CM A survey of population analysis methods and software for complex pharmacokinetic and pharmacodynamic models with examples AAPS Journal 2007 9 1 E60 83
258. noticeable There is some initial file copying required between manager and worker directories or computers but after the initial loading of the NONMEM processes all information transfer is via the message passing interface without requiring file transfer The PARAFILE Parallel computing with NONMEM 7 2 0 uses a parallel file or parafile that controls the parallelization process implemented by NONMEM and is written by the user The NONMEM installed run directory has sample pnm files that can be used as a template The name of the parallel file may be given at the command line as Nmfe73 myexample ctl myexample res parafile myparallel pnm quotes of some kind may be needed for Windows otherwise the parameters are improperly parsed This parallel file will remain in effect throughout the control stream file to be used in all SEST methods If no parafile switch was given then the default name parallel pnm is assumed The reserved default name of parallel pnm should not be used as it is only for the worker process Make sure no file called parallel pnm exists in your manager s run directory The PARAFILE option may be alternatively set to the keywords ON or OFF If a PARAFILE parameter is set to OFF in a EST command then parallelization does not occur for that EST command If a subsequent PARAFILE is set to ON the parallelization occurs using the most nm730 doc 136 of 210 NONMEM Users Guide Introduction to NONMEM
259. ns process whether cores or computers NODES gt 1 WORKER node NODES 0 parse num number of subjects to give to each node parse type 0 give each node parse num subjects parse type 1 evenly distribute numbers of subjects among available nodes parse type 2 load balance among nodes parse type 3 assign subjects to nodes based on idranges parse type 4 load balance among nodes taking into account loading time This setting of parse type will assess ideal number of nodes If loading time too costly will eventually revert to single CPU mode nm730 doc 156 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 timeouti seconds to wait for node to start if not started in time deassign node and give its load to next worker until next iteration timeout minutes to wait for node to compelte if not completed by then deassign node and have manager complete it paraprint 1 print to console the parallel computing process Can be modified at run time with ctrl B toggle Regardless of paraprint setting control stream log always records parallelization progress transfer type 0 for file transfer unloading and reloading workers with each estimation transfer type 1 for mpi transfer type 2 for file transfer maintaining a single loaded process throughout the run THE EXCLUDE INCLUDE may be used to selectively use certain nodes out of a large
260. nstrained etas as is done in indestm ctl a single residual variance as SIGMA 1 1 is estimated across subjects for indestms For this analysis a re iterative analysis to improve SIGMA must be performed so MAXEVAL 0 must be set Non zero THETAS may also be introduced to provide additional shared parameters as is done in standard population analysis Please note that when using this POPULATION WITH UNCONSTRAINED ETAS analysis NM TRAN still sees the data as population and will declare it as such in its warning statements NMTRAN NONMEM process the problem as population while the statistical algorithms treat the data as single subject at least concerning unconstrained etas offering the best of both worlds Thus NONMEM is capable of parallelizing these problems The traditional single subject analysis however cannot be parallelized because NONMEM processes each subject in sequence nm730 doc 183 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 64 References 1 Hooker AC Staatz CE Karlsson MO Conditional weighted residuals CWRES a model diagnostic for the FOCE method Pharmaceutical research 2007 24 2187 97 2 Comets E Brendel K Mentre F Computing normalized prediction distribution errors to evaluate nonlinear mixed effects models the npde add on package for R Computer Methods and Programs in Biometrics 2008 90 154 166 3 Brendel K Comets E Laffont C Laveille C Mentre F Metrics for External Model Eval
261. o N SIGL where N is the number of subjects for a maximum of 15 the limiting precision of double precision Thus with 100 subjects the actual precision that the total objective function is evaluated could be 12 One should not necessarily rely on this so it is safest to suppose the more conservative precision of 10 for which a suitable NSIG would be 3 For analytical problems those which do not utilize DES one can usually expect a reasonably efficient convergence to the minimum of the objective function with NSIG 3 However with differential equation problems those used for ADVAN 6 8 9 or the new ADVAN method 13 the limiting precision that objective function values may be evaluated is not based on the internal SIGL of 10 but rather on the TOL level set by the user where TOL represents the relative significant digits precision to which differential equations are to be integrated so the precision is 10795 which is used by PREDPP when differential equations are integrated The relationship between the predicted value and the individual subject s maximized objective function is complex but one can use the rule of thumb that the individual s objective function is evaluated to a precision of the smaller of TOL and the internal SIGL Thus when a user specifies a TOL 4 then it may well be that the sum objective function has no greater precision than 4 If the user then specifies NSIG 3 then the main search algorithm evaluates finite gradien
262. o a value greater than or equal to 100 it is converted to that value minus 100 upon input but will also not be used at all during estimation only for table outputting This option allows you to use the same enhanced data file for estimation and Table outputs without significantly slowing down the estimation So the finedata control stream file would be SPROB RUN example6 from r2compl SINPUT C SET ID JID TIME DV CONC DOSE AMT RATE EVID MDV CMT SDATA example6 csv IGNORE C SFINEDATA TSTART 0 TSTOP 50 NEVAL 100 AXIS TIME LIN CMT 1 3 MDV 101 101 FILE example6b csv In the following example TSTART TSTOP and NEVAL are obtained from columns TIMESTART TIMESTOP and NEVAL respectively SPROB RUN example6 from r2compl SINPUT C SET ID JID TIME DV CONC DOSE AMT RATE EVID MDV CMT TIMESTART TIMESTOP NEVAL SDATA example6c csv IGNORE C SFINEDATA TSTART TIMESTART TSTOP TIMESTOP NEVAL NEVAL AXIS TIME LIN CMT 1 3 FILE example6 d csv Multiple data sets may be processed by one finedata control stream file by using PROB records to separate the problems SPROB SINPUT C DROP ID TIME CMT OBSV DV COHT EVID AMT DOSE MDV SDATA mydata csv IGNORE C SFINEDATA tstart 0 TSTOP 700 NEVAL 500 AXIS TIME LIN CMT 1 4 file mydata fine csv SPROB SINPUT C DROP ID TIME CMT OBSV DV COHT EVID AMT DOSE MDV SDATA mydatab csv IGNORE
263. oading and reloading workers with each estimation transfer type 1 for mpi transfer type 2 for file transfer maintaining a single loaded process throughout the run HE EXCLUDE INCLUDE may be used to selectively use certain nodes out of a large list SEXCLUDE 5 7 exclude nodes 5 7 O I OXCLUDE ALL 1 4 6 NCLUDE TI SNAMES Give a label to each node for convenience 1 2 3 4 C W 4 D 1 2 MANAGER WORKER1 WORKER2 WORKER3 OMMANDS each node gets a command line used to launch the node session Command lines must be on one line for each process The following commands are for FPI method on Windows First node is manager so it does not get a command line when using FPI NONE load on a core of the same computer as manager Note that worker does not really need a control stream file but something must be there as a place holder Also for psexec notice that the worker directories are named as the worker sees them not as the manager sees them Very important distinction for remote worker computers wdir refers to working directory for particular process do not user cd with psexec Just user relative directory notation psexec d w workerl cmd exe C nonmem exe load on a core of the same computer as manager psexec d w worker2 cmd exe C nonmem exe load on a core of a different compu
264. od remember to revert these parameters back to default operation on the next SEST step ISCALE MIN 100 ISCALE MAX 100 The default operation is that NONMEM sets ISCALE MIN ISCALE MAX to 0 01 100 for importance sampling as described earlier and to 1 0E 06 1 0E 06 for MCMC sampling PSAMPLE MI 1 defaults listed PSAMPLE M2 1 PSAMPLE M3 1 PACCEPT 0 5 These are the options for the MCMC Metropolis Hastings algorithm These options only have meaning for population parameters theta sigma that are not Gibbs sampled Normally NONMEM determines whether THETA and SIGMA parameters are Gibbs sampled or not based on the model setup see MU_ Referencing section below For each iteration a vector of thetas sigmas are generated using a multivariate normal proposal density that has mean variances based on the previous samples done PSAMPLE M1 times Next a vector of parameters are generated using a multivariate normal proposal density with mean at the present parameter position and variance scaled to have samples accepted with PACCEPT frequency This is done PSAMPLE M2 times if PSAMPLE M2 0 then program performs this as many times as there are M H parameters Finally each parameter is individually sampled PSAMPLE MG times The final accepted parameter vector is kept Usually these options do not need to be changed from their default values listed above nm730 doc 77 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 PSCALE_MI
265. ods For the new analysis methods SIGL is also used to set up forward difference or central difference gradients as needed Such finite difference gradients need to be set up for sigma parameters and thetas not MU modeled to etas or where OMEGA values of etas to which the thetas are MU associated are set to 0 NSIG is used only with the iterative two stage method among the new methods The iterative two stage is not Monte Carlo and has a more deterministic smooth trajectory for its parameter movements with each iteration In this case NSIG is used as follows The average of the last CITER 2 parameters are evaluated and compared with the average of the next to last CITER 2 parameters If CITER is odd valued CITER 1 2 will be used For example for CITER 5 at iteration 102 iterations 97 99 are compared with iterations 100 102 If they differ by no more than NSIG significant digits then this parameter is considered to have converged When this is true for all parameters tested optimization is completed 1 35 List of EST Options and Their Relevance to Various Methods Option Classical 2LL X ATOL ADVAN9 13 X AUTO x l un gt gt lt gt lt gt lt gt gt lt gt lt gt lt gt gt lt gt lt gt lt gt gt lt gt lt gt lt gt gt lt gt lt gt lt CALPHA CENTERING X CINTERVAL CITER CNSAMP CONDITIONAL X CONSTRAIN X PX X PX
266. of the likelihood with respect to the thetas and sigmas are also evaluated integrated over all possible values of the etas From these constructs the thetas and sigmas are updated during the maximization step using these conditional means of the etas and or the gradients The omegas are updated as the sample variance of the individual conditional means of the etas plus the average conditional variances of the etas The maximization step 1s therefore typically a single iteration process requiring very little computation time The more accurately these constructs are evaluated during the expectation step the more accurately the total likelihood will be maximized 1 24 Iterative Two Stage ITS Method Iterative two stage evaluates the conditional mode not the mean and first order expected or second order Laplace approximation of the conditional variance of parameters of individuals by maximizing the posterior density This integration step is the same as is used in FOCE or Laplace Population parameters are updated from subjects conditional mode parameters and their approximate variances by single iteration maximization steps that are very stable usually converging in 50 100 iterations Because of approximations used population parameters almost but not quite converge towards the linearized objective function of FOCE Iterative two stage method is about as fast as FOCE with simple one or two compartment models and when set up with MU refe
267. ollowed by a amp for each line that needs to be continued if using an ampersand and it is at the end of the line in the control stream file place a after it so it is not interpreted as a continuation indicator by the NMTRAN control stream file reader nm730 doc 119 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 FORMAT s1PE15 8 160 amp c Will print lines of at most 160 characters with amp tagged at the end of the line to be continued and a c at the beginning of the continued line FORMAT s1PE15 8 160sc Will print lines of at most 160 characters with no character at the end of each line to be continued and a c at the beginning of the continued line S represents space and a space may not serve as a continuation marker because of its ambiguity so it serves here as a place holder in the FORMAT definition These line continuation formats are ignored in TABLE records but are used in the SEST record for all additional file formats and can are used in EST CHAIN METHOD and CHAIN records NOTITLE 0 1 If NOTITLE 1 default 0 then the Table header line will not be written to the raw output file specified by FILE NOLABEL 0 1 If NOLABEL 1 default 0 then the column label line will not be written to the raw output file specified by FILE ORDER NM72 The order in which the thetas omegas and sigmas are listed in the output file is by default as follows Thetas T SIGMAS S OMEGAS O The SIG
268. ollowing it on the same line will be found a text that describes the method for example First Order Conditional Estimation Method with Interaction TERM This tag indicates that beginning on the next lines text describes the termination status of the analysis Included in the results are average of the individual etas ETABAR its standard error SE P value on the null hypothesis that ETABAR is not statistically different from 0 and eta and epsilon shrinkage Shrinkage is not reported after a BAYES or FO analysis See below for more information on shrinkage nm730 doc 115 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 The individual etas used to assess ETABAR SE p value Shrinkage are modes of the posterior density for ITS FOCE Laplace for each individual or conditional mean etas for IMP SAEM for each individual as of the last iteration ETABAR SE P Value and Shrinkage are not always accurately calculated after an SAEM analysis as these are averaged over the entire set of iterations of the reduced stochastic mode assuming NITER gt 0 during which the estimates of thetas omegas and sigmas are also averaged After an SAEM analysis run a EST METHOD IMP EONLY 1 to obtain good post analysis estimates of shrinkage standard errors and objective function as described earlier TERE This tag indicates the end of the lines describing the termination status of the analysis Thus a software program may transfer all line
269. on SAEM Method An AUTO option to allow NONMEM to determine the best options for Monte Carlo Expectation Maximization EM and Bayesian Markov Chain Monte Carlo methods instead of the user having to determine these settings for each problem See section 1 31 Some General Options and Notes Regarding EM and Monte Carlo Methods Perform a Monte Carlo search or select from a pre existing list of initial thetas omegas and sigmas that provide the lowest starting objective function for estimation See section 1 48 Method for creating several instances for a problem starting at different randomized initial positions EST METHOD CHAIN and CHAIN Records Perform a Monte Carlo search for initial best estimates of etas for each subject Together with a Monte Carlo search of best initial thetas omegas and sigmas this provides a global search technique for the traditional deterministic estimation methods with less reliance on starting position for incidence of success See MCETA in section 1 20 Options for ESTIMATION Record for alternative MAP eta optimization methods and evaluating individual variances by numerical derivative methods for FOCE Laplace NM73 nm730 doc 9 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 FOCE Laplace and ITS to be assessed using only numerical eta derivatives for search of best etas and or eta Hessian matrix assessment This feature relaxes the requirement that analytic derivatives be computed for FOCE and L
270. ontrol stream file are completed then select TRANSFER TYPE 2 In these cases new PARAFILE settings at EST steps within the control stream file will be ignored except for PARAFILE ON or PARAFILE OFF Installing MPI on Windows Go to the web site http phase hpcc jp mirrors mpi mpich2 and select the suitable Windows version with extension msi Or select the mpich2 1 2 1pl win ia32 msi file listed in the MPI directory of the NONMEM installation disk Install the full version on the manager computer by double clicking on the msi file or running it from START gt run Follow the instructions in section 7 of mpich2 1 2 1 windevguide pdf and verify that the MPI system is working Copy the program mpiexec exe from the bin directory of the MPICH2 directory to your manager NONMEM run directory NONMEM comes with the MPI library files they are located in mpi MPI_WINI for Intel Fortran and mpi MPI_WING for gfortran For communication across computers make sure you also have a network file allocated as described above If the MPI library files do not match the version which you downloaded or there are linking difficulties when you run nmfe73 bat then copy the appropriate lib file from the MPICH2 installed directory mpich2 lib to mpi MPI_WINI directory Keep in mind that we have supplied 32 bit versions of libraries Environments with 64 bit processing may require libraries from the mpich2 web site nm730 doc 146 of 210 NONMEM Use
271. oooooN ooooo ooo0oN oooo COON ooo COON oo ON SIGMA 0 1 p 0 1 p Starting with a short iterative two stage analysis brings the results closer SO less time needs to be spent during the burn in of the BAYES analysis SEST METHOD ITS INTERACTION SIGL 4 NITER 15 PRINT 1 FILE example6 ext NOABORT NOPRIOR 1 SEST METHOD BAYES INTERACTION NBURN 4000 SIGL 4 NITER 30000 PRINT 10 CTYPE 3 FILE example6 txt NOABORT NOPRIOR 0 By default ISAMPLE M are 2 Since there are many data points per subject setting these to 1 is enough and it reduces the time of the analysis ISAMPLE M1 1 ISAMPLE M2 1 ISAMPLE M3 1 IACCEPT 0 4 COV MATRIX R UNCONDITIONAL r r nm730 doc 197 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 71 Example 7 Inter occasion Variability Model Desc Interoccasion Variability Project Name nm7examples Project ID NO PROJECT DESCRIPTION PROB runi example7 from adltr2 occ SINPUT C SET ID TIME AMT RATE EVID MDV CMT DV DATA example7 csv IGNORE C SSUBROUTINES ADVAN1 TRANS2 PRIOR NWPRI NTHETA 2 NETA 5 NTHP 0 NETP 5 NPEXP 1 PK MU 1 THETA 1 MU 2 THETA 2 V DEXP MU_1 ETA 1 CLB DEXP MU_2 ETA 2 DCL1 DEXP ETA 3 DCL2 DEXP ETA 4 DCL3 DEXP ETA 5 Si v DCL DCL1 IF TIME GE 5 0 DCL DCL2 IF TIME GE 10 0 DCL DCL3 CL CLB DCL vVc V SERROR IPRED F Y F F EPS 1 Initial Thetas STHETA 2 0 MU 1 2 0 MU 2 Initial omegas OMEGA BL
272. or MCMC Bayesian analysis not necessary for maximization methods In this example only the OMEGAs have a prior distribution the THETAS do not For Bayesian methods it is most important for at least the OMEGAs to have a prior even an uninformative one to stabilize the analysis Only if the number of subjects exceeds the OMEGA dimension number by at least 100 then you may get away without priors on OMEGA for BAYES analysis PRIOR NWPRI NTHETA 11 NETA 4 NTHP 0 NETP 4 NPEXP 1 PK LCLM log transformed clearance male LCLM THETA 1 LCLF log transformed clearance female LCLF THETA 2 CLAM CL age slope male CLAM THETA 3 CLAF CL age slope female CLAF THETA 4 LV1M 1og transformed V1 male LV1M THETA 5 LV1F log transformed V1 female LV1F THETA 6 VIAM V1 age slope male V1AM THETA 7 VlAF V1 age slope female V1AF THETA 8 LAGE log transformed age LAGE DLOG AGE Mean of ETA1 the inter subject deviation of Clearance is ultimately modeled as linear function of THETA 1 to THETA 4 Relating thetas to Mus by linear functions is not essential for ITS IMP or IMPMAP methods but is very helpful for MCMC methods such as SAEM and BAYES MU 1 1 0 GNDR LCLM LAGE CLAM GNDR LCLF LAGE CLAF Mean of ETA2 the inter subject deviation of V1 is ultimately modeled as linear function of THETA 5 to THETA 8 MU 2 1 0 GNDR LV1M LAGE V1AM GNDR LV1F LAGE V1AF MU
273. orm Examples SOMEGA BLOCK 2 or OMEGA VARIANCE COVARIANCE BLOCK 2 0 64 0 2402 0 58 SOMEGA STANDARD BLOCK 2 nm730 doc 21 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 0 8 0 24 0 762 SOMEGA STANDARD CORRELATION BLOCK 2 0 8 0 394 0 762 SOMEGA VARIANCE CORRELATION BLOCK 2 0 64 0 394 0 58 SOMEGA CHOLESKY BLOCK 2 0 8 05 3 0 7 SSIGMA 0 3 STANDARD 0 8 STANDARD 0 3 VARIANCE These input options do not affect how estimated OMEGAs and SIGMAs are outputted With NONMEM 7 3 0 there are new features for abbreviated code and the ABBR record Each is discussed in greater detail in the on line help and Guide VIII Repeated SAME BLOCK for OMEGA and SIGMA Records NM73 No need to repeat multiple SAME block segments OMEGA BLOCK 2 SAME 3 Is equivalent to OMEGA BLOCK 2 SAME OMEGA BLOCK 2 SAME OMEGA BLOCK 2 SAME The SAME m feature is also available for SIGMA SIGMA BLOCK 2 SAME 3 Repeated Value Inputs for STHETA OMEGA and SIGMA NM73 As of NM73 repeated inputs of STHETA be entered as follows Long hand STHETA 2222 0 001 0 1 1000 0 001 0 1 1000 0 001 0 1 1000 0 5 FIXED 0 5 FIXED Short hand STHETA 2 x4 0 001 0 1 1000 x3 0 5 FIXED x2 Where xn means to replicate n times The item to be repeated must always be in parentheses and the xz must always be immediately after the item not before it 4x 0 2 is no
274. ortance Sampling EM Initial etas may be introduced in the control stream file or from an external source See 1 49 ETAS and PHIS Record For Inputting Specific Eta or Phi values NM73 For the DATA record EQN may be used in the IGNORE ACCEPT option to indicate a numerical comparison rather than a literal comparison as is done for EQ and NE See Numerical Equality Comparison for IGNORE option in DATA Record NM73 in section 4 Expansions on Abbreviated and Verbatim Code NM72 NM73 Informative record names for prior information of thetas omegas sigmas provide easier entry of NWPRI prior information See 29 A Note on Setting up Prior Information Maximal number of numerical integration steps is now easy to modify for ADVAN9 and ADVANI3 See discussion on MXSTEP in 1 14 SUBROUTINES New Differential Equation Solving Method nm730 doc 12 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Mu model checking by NMTRAN can be turned off If you wish to turn this off checking mu statements can take a long time for very large control stream files then include the NOCHECKMU option on the ABBR record SABBR NOCHECKMU NMTRAN will allow amp as a continuation marker on abbreviated code lines Furthermore the total length of a control stream record whether on a single line or continued on several lines using amp may be up to 67000 characters long See Continuation indicator is allowed in abbreviated code non verbatim l
275. pi_ling for gfortran For communication across computers make sure you also have a network file allocated just as with the FPI method If the MPI library files do not match the version which you downloaded or there are linking difficulties when you run nmfe73 then copy the appropriate a file from the MPICH2 installed nm730 doc 154 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 directory mpich2 lib to the NmpiWnpi lini directory Keep in mind that we have supplied 32 bit versions of libraries Environments with 64 bit processing may require libraries from the mpich2 web site For easy access of the mpi utility programs you should expand the PATH to include the path to the bin directory of the MPICH2 system if it is not there already You can insert the following line in the manager s SHOME bashrc file for example export PATH S HOME MPICH2 LINUX mpich2 install bin PATH During the parallelization process NONMEM sends a copy of its program in nonmem exe on Windows nonmem on Linux to the worker computer and then loads it there Therefore the worker computers must be of the same operating system although not necessarily same version as the manager computer For Intel fortran the worker computer does not have to have Intel Fortran installed For gfortran static option for the MPI method cannot be used in the nmfe73 script as it prevents the MPI components from being properly linked Thus the gfortran version of
276. porary directory for the PREDPP compilation for a given problem by using runpdir option at the nmfe73 command line For example You may run problem A as nmfe73 mycontrolA ctl myresults res runpdir mycontrolA and then follow with problem B as nmfe73 mycontrolB ctl myresults res runpdir mycontrolB When you return to rerunning problem A at some later time nm730 doc 34 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 nmfe73 mycontrolA ctl myresults res runpdir mycontrolA it won t need to recompile assuming your PREDPP sizings and PREDPP model did not change for problem A as its PREDPP recompile directory was not overwritten by the intervening call to problem B Finally if you feel that it is sufficient to use default sizes in sizes f90 for the various PREDPP parameters and therefore use the precompiled routines in pr of the NONMEM installed directory you may use the prdefault option nmfe73 mycontrol ctl myresults res prdefault As of nm73 you may also use the tprdefault option which tests if prdefault is acceptable and if so will use it otherwise it will perform a PREDPP recompile nmfe73 mycontrol ctl myresults res tprdefault If you enter nmfe73 mycontrol ctl myresults res tprdefault prcompile then if prdefault is not acceptable and will act on the prcompile option If you enter nmfe73 mycontrol ctl myresults res tprdefault prsame then if prdefault is not acceptable a
277. que to PREDPP and prsizes f90 All may be changed with SIZES For example SIZES MAXFCN 9000000 might be used with General Non Linear models ADVAN6 ADVAN8 ADVANO9 ADVAN13 to request more function evaluations than the default value in resource SIZES f90 which is MAXFCN 1000000 As of NM73 PCT and PIR are assessed by NMTRAN and submitted to NONMEM if prdefault is not used nm730 doc 32 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Usually a parameter value needs to be specified in SIZES when the problem is bigger than what is specified in sizes f90 For example if LTH 40 in sizes f90 and your problem needs only 35 thetas then NONMEM executable will be built to size for 35 thetas and SIZES was not needed If however the problem requires 45 thetas then SIZES LTH 45 or greater needs to be specified and then NONMEM will be set to a size of LTH 45 as well For the following parameters LTH LVR PD PC DIMTMP MMX DIMCNS and or PDT NMTRAN must anticipate a maximum size because it needs to set up internal arrays that stores the information it will gather from the control stream file It will get this maximum size from the values in sizes f90 or from the user specifying the required size in SIZES If the user does not specify in SIZES then NMTRAN will determine the best size for the problem and construct the NONMEM executable accordingly But if the user specifies a size in SIZES then this is also the size by whi
278. r process may be a new Linux operating system with a GLIBC that is new while a worker computer may be Linux with an older operating system with an old GLIBC This typically is not an easy environment to set up but if you wish to do so it means that you would need to create the nonmem executable on the Linux machine ahead of time name it nonmem2 or some other name so it is not copied over with the nonmem executable of the manager process and use that nonmem2 on the worker SCOMMANDS line 2 beolaunch sh wrk ftif nonmem2 workerl out One would do something similar if the manager were a Windows process and the worker were a Linux process for example but it is up to the user to find a means of launching a remote Linux process The psexec launcher only works between Windows computers Installing MPI on Linux If you are communicating across computers make sure you set up a share drive and the ssh system as described earlier Go to the web site http phase hpcc jp mirrors mpi mpich2 and select the appropriate tar gz file Or select the mpich2_1 2 1 1 orig tar gz file in the MPI directory given in the NONMEM installation disk On the manager computer unpack the tar gz file tar xfz mpich2 1 2 1 orig tar gz Follow the instructions in section 2 2 of mpich2 1 2 1 installguide pdf and verify that the MPI system is working NONMEM comes with the MPI library files they are located in mpi mpi_lini for Intel Fortran and mpi m
279. re directories Your particular environment will be different The computer name of the worker computer should be displayed You may be required to enter a user name and password If this is the case you should make sure that your user account and password on your manager computer is the same as on the worker computer so that user name and password is not requested Otherwise when you run the NONMEM program the run will be continually interrupted for this information During the parallelization process NONMEM sends a copy of its program nonmem exe on Windows nonmem on Linux to the worker processes s directory and then loads it there Therefore the worker computers must typically be of the same operating system although not necessarily same version as the manager computer but see below to get around this The worker computer does not have to have Intel or gfortran installed For a quick test on a single multi core computer try the following Copy foce_parallel ctl and examplel csv from the NONMEM examples directory fpiwini8 pnm from the NONMEM run directory and psexec exe from the NONMEM run directory into your standard run directory Then execute the following from your standard run directory Nmfe73 foce parallel ctl foce_parallel res parafile fpiwini8 pnm nodes 4 where the values of nodes should be no greater than the number of cores available on your computer A parafile example set up for FPI method on Window
280. rencing described below can be several fold faster than FOCE with more complex problems such as 3 compartment models and differential equation problems The iterative two stage method is specified by EST METHOD ITS INTERACTION NITER 50 where NITER default 50 sets maximum number of iterations For all new methods it is essential to set INTERACTION if the residual error is heteroscedastic nm730 doc 65 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 25 Monte Carlo Importance Sampling EM Importance sampling evaluates the conditional posterior mean and variance of parameters of individuals etas by Monte Carlo sampling integration expectation step It uses the posterior density which incorporates the likelihood of parameters relative to population means thetas and variances etas with the individual s observed data By default for the first iteration the mode and first order approximation of the variance are estimated called mode a posteriori or MAP estimation as is done in ITS or FOCE and are used as the parameters to a normal distribution proposal sampling density From this proposal density Monte Carlo samples are generated then weighted according to the posterior density as a correction since the posterior density itself is generally not truly normally distributed and conditional means and their conditional variances are evaluated For subsequent iterations the normal density near the mean of the posterior
281. represented in the number list in the braces is at least as many as the number list before the colon Another example 2 4 7 WORKER 1 3 Expands to 2 WORKER1 4 WORKER2 7 WORKER3 Another method is to use the expression offset which directly substitutes the process number listed before the colon into the place at the braces with an offset added to it So 2 8 wdir cd wk 1 hosts 1 localhost 1 nonmem exe Expands to 2 wdir cd Nwkl hosts 1 localhost 1 nonmem exe 3 wdir cd wk2 hosts 1 localhost 1 nonmem exe 4 wdir cd Nwk3 hosts 1 localhost 1 nonmem exe 5 wdir cd Nwk4 hosts 1 localhost 1 nonmem exe 6 wdir cd Nwk5 hosts 1 localhost 1 nonmem exe 7 wdir cd Nwk6 hosts 1 localhost 1 nonmem exe 8 wdir cd Nwk7 hosts 1 localhost 1 nonmem exe Similarly 2 4 7 wdir cd wk 11 hosts 1 localhost 1 nonmem exe Expands to 2 wdir cd Nwk13 hosts 1 localhost 1 nonmem exe 4 wdir cd wk15 hosts 1 localhost 1 nonmem exe 7 wdir cd wk18 hosts 1 localhost 1 nonmem exe Easy to Use Parafiles For easy use there are a series of pnm files in the vun directory that can take any number of cores on a single computer These are fpiwini8 pnm mpiwini8 pnm fpilinux8 pnm and nm730 doc 142 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 mpilinux8 pnm for MAC OSX use the linux8 pnm files located in the NONMEM run directory The 8 refers to the default number of nodes proces
282. results in ever increasing objective function values without stabilization even though the SAEM result is reasonable The usual adjustment of options in nm 7 1 2 fails to correct the problem In NONMEM 7 2 some internal scaling parameters have been adjusted Also the user can further adjust these scaling parameters 4 For certain estimation problems ADVAN 5 and ADVAN7 provide inaccurate prediction values which are sensitive to the initial thetas The work around for earlier releases is to use ADVAN6 or ADVANO 5 During a simulation problem if symmetric band matrix patterns are used in the OMEGA including a block matrix which has all covariances of 0 the first simulated data set will be correct but subsequent data sets will be incorrect This occurs because the banding information is re initialized after the first sub problem simulation This is corrected in NONMEM 7 2 Asa work around for earlier releases during simulations replace the 0 valued covariances with very small values of covariances such as 1 0e 05 6 During an estimation with FO or FOCE and the last subject in the data set has non influential etas for example with interoccasion variability if the last subject had no data during the last inter occasion the eta for that last inter occasion is non influential the estimation may become inefficient due to incorrect gradient assessments 7 If DROP is used in INPUT to not include a data item in any problem this DROP attribut
283. ring the same directory as the manager and because of the NONE directory designation in DIRECTORIES the executable nonmem will not be copied as it should not since the worker processes are pointing to the manager directory and therefore the nonmem executable in the manager directory is already available to worker processes as well Furthermore the option wnf is given This option tells the nonmem process that it is a worker MPI method and the nf tells it not to make any file buffers nf no files The worker process has all the information it needs to launch without requiring any file based communication with the manager and minimizes the footprint on the drive directory The next 4 processes are launched on a remote computer with similar settings Notice that only one of the processes among the 5 to 8 had to have a SDIRECTORY defined that of mnt workerl which they all are pointing to The HOME directory of the worker computer is the directory mnt workerl that the manager has a share connection to This means that NONMEM has a path direction to copy the nonmem executable from its current directory to the HOME directory on the worker computer If all processes DIRECTORIES entries were NONE then the most recently built nonmem executable cannot be copied to the remote computer You may want that if for example you have arranged for a nonmem executable to be there already that was previously built with the identical control stream file Ma
284. rk must exist or be concerned with populating them with the appropriate files including the nonmem executable NONMEM will take care of this automatically For example while w share needs to exist before the run as it was the share directory that needed to be set up w share worker3 did not have to exist before the NONMEM run Make sure that the managers and workers have appropriate read write access to these directories and proper privileges to load on remote computers The COV statement also allows a PARAFILE setting to turn on or off parallel computing for the COV step for classical NONMEM methods or changing the parallelization profile Examples of PARAFILE files are given in NONMEM s run directory as a list of pnm files Examples are shown in the next sections as well The files fpiwini8 pnm fpilinux8 pnm mpilinux8 pnm and fpilinux8 pnm are particularly versatile in that they are useful for multiple cores on a single computer and are designed to be used in any run directory Substitution Variables in the parafile Substitution variables provide flexibility in the use of the parafile Certain substitution variables are reserved words as follows which can be passed as arguments to the worker nonmem executable although typically this is not necessary to do so That is they are placed at the end of a COMMANDS process command line coming after nonmem exe as arguments to nonmem exe as needed control stream substitu
285. rom an MSF file or from a PHIS SETAS record and you want to calculate STABLE items based on those etas rather than from a new estimation For example SPROB SINPUT C ID GRP AMT TIME DV1 DV CMTS EVID MDV SDATA mydata csv IGNORE C SMSFI myresults MSF SEST METHOD 1 FNLETA 2 STABLE ID TIME DV IPRED CMTS MDV EVID NOAPPEND NOPRINT FILE mytable tab 1 21 Bootstrap Selecting a Random Method and Other Options for Simulation NM73 BOOTSTRAP NM73 SIML BOOTSTRAP 1 SUBP 100 EST METHOD 1 INTERACTION The above example requests a bootstrap rearrangement with replacement of an existing data set followed by analysis of that data set The BOOTSTRAP number refers to how many subjects are to be randomly selected from the data set Setting 1 or to a value larger than the number of subjects in the data set means to randomly select as many subjects as are in the data set For example if 400 subjects are in the simulation template data set then 400 subjects are randomly selected with replacement so some are selected more than once others not at all In this case NONMEM s simulator does not perform the usual activity of randomly creating DV values for a new data set but rather selects a random set of subjects of an existing data set which must already have legitimate DV values uniformly selected using seed1 with replacement This results in some subjects not being selected at all and some subjec
286. rs Guide Introduction to NONMEM 7 3 0 The MPI Windows installation guide section 9 may offer other ways to supply user name and password via the program mpiexec For example from the manager computer mpiexec register Enter name Enter password During the parallelization process NONMEM sends a copy of its program in nonmem exe on Windows nonmem on Linux to the worker computer and then loads it there Therefore generally the worker computers must be of the same operating system although not necessarily same version as the manager computer For Intel fortran or gfortran the worker computer does not have to have the compiler installed In addition the MPI system needs certain executable files available on the worker computer A minimal installation on the worker computer can be implemented by copying smpd exe found in the bin directory of you manager s MPICH2 directory to the worker computer and executing Smpd exe install See section 9 of the MPI Windows installation guide about the full use of smpd exe Also the MPI system needs certain dll library files placed in each worker processor s directory of the worker computer or in the windows system32 directory more generally in systemroot system32 Fmpich2 dll intel or fmpich2g dll gfortran Mpich2 dll Mpich2mpi dll The dll files are located in the manager s Vosystemroot oNs ystem32 directory Once you have an MPI system set up for a quick test on a single mu
287. rt tstop and neval parameters are to be inserted nmtemplate nmtemp fnt nmtemp fnd TSTART 0 TSTOP 100 NEVAL 200 resulting in the FINEDATA control stream file nmtemp fnd SINPUT C SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT CLX V1X QX V2X SDIX SDSX SDATA nmtemp csv IGNORE C SFINEDATA AXIS TIME LIN TSTOP 100 TSTART 0 NEVAL 200 FILE nmtemp2 csv Note that only words that match the variable list at the nmtemplate command line and have enclosing brackets lt gt will be replaced with the suggested values The values may also be text with no spaces in them These two scripts could be combined to provide a means of creating individual simulated curves Consider the following DOS patch script which could also be converted to an R S PLUS script or function nmtemp bat nmtemplate exe nmtemp fnt nmtemp fnd TSTART 1 TSTOP 2 NEVAL 3 finedata exe nmtemp fnd nmtemplate exe nmtemp nmt nmtemp ctl NMID 4 nmfe73 bat nmtemp ctl nmtemp res prdefault Where 1 through 4 are the DOS command line substitution parameters So the script could be executed as follows Call nmtemp bat 0 100 200 34 Then a program such as R S PLUS or S ADAPT can read in the results from nmtemp tab and plot them Another feature of nmtemplate is that the user may request a random number to be generated to serve as a value by referring to R al a2 a3 R al a2 a3 is a special function of nmtemplate whic
288. rted one with CMT 1 EVID 2 and the other with CMT 3 EVID 2 Or FINEDATA CMT 1 1 3 3 EVID 0 2 0 2 Inserts four records per time point with the following CMT EVID values in the order specified CMT EVID nm730 doc 173 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 WW NON CO MISSING comma delimted list of missing symbols By default a period and space s are considered missing values Values such as 0 or 99 may be present in the data as symbols for missing values They may be described with MISSING 0 or MISSING 99 During interpolation missing values will be skipped and only records with non missing values will be used for interpolation If NEVAL 1 only the inserted records will have filled in interpolated values and the original records will remain untouched When NEVAL 1 then original records will be filled in for the specified items but no inserted records will be added Thus filling missing values in original records is done as a separate action from inserting records They may not be done simultaneously in finedata with a single PROB but these two actions can be accomplished by two sequential PROB records See finetest7 ctl to first fill in original records with interpolated values followed by using the resulting data file as the input for the next PROB in which additional records are inserted SPROB RUN example6 from r2compl SINPUT C SET ID JID TIME DV CONC DOSE AMT RATE EVI
289. s between TERM and TERE to a summary file OBJT Indicates that following it on the same line is the text describing the objective function such as Minimal Value Of Objective Function OBJV Indicates that following it on the same line is the objective function value However a more efficient way of extracting numerical results from the analysis is from the raw output file see below OBJS Indicates that following it on the same line is the objective function standard deviation MCMC Bayesian analysis only However a more efficient way of extracting numerical results from the analysis is from the raw output file see below OBJN nm73 Indicates that following it on the same line is the nonparametric objective function value CPUT nm73 Total cpu time in seconds This is an accurate assessment of CPU usage of the entire problem whether done in single or parallel mode Shrinkage and ETASTYPE NM73 Inter subject variance shrinkage ETAshrink for each eta is evaluated as 100 1 SD eta i sqrt omega i 1 Eta shrinkage is averaged for all subjects if ETASTYPE 0 Should you wish to correct for some subjects not contributing at all to one or more etas this may or may not be desirable depending on your needs the shrinkage can be recalculated as follows nm730 doc 116 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 5 E S 100 1 Soa Naua _ 1 Ei Ny 1 Nu s 100 Np D 2 W
290. s is as follows set TRANSFER_TYPE 0 SGENERAL NODES 2 PARSE TYPE 3 PARSE NUM 200 TIMEOUTI 60 TIMEOUT 10 PARAPRINT 0 TRANSFER TYPE 0 NODES number of nodes that is process whether cores or computers SINGLE node NODES 1 MULTI node node means process whether cores or computers NODES gt 1 WORKER node NODES 0 parse num number of subjects to give to each node parse type 0 give each node parse num subjects parse type 1 evenly distribute numbers of subjects among available nodes parse type 2 load balance among nodes nm730 doc 144 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 t LM r S parse type 3 assign subjects to nodes based on idranges parse type 4 load balance among nodes taking into account loading time This setting of parse type will assess ideal number of nodes If loading time too costly will eventually revert to single CPU mode timeouti seconds to wait for node to start if not started in time deassign node and give its load to next worker until next iteration timeout minutes to wait for node to compelte if not completed by then deassign node and have manager complete it paraprint 1 print to console the parallel computing process Can be modified at run time with ctrl B toggle Regardless of paraprint setting control stream log always records parallelization progress transfer type 0 for file transfer unl
291. s may prefer to request NPRED CRES CWRES or NPRED RES CWRES The conditional weighted residual will not differ from the non conditional weighted residual if FO was selected in the previous EST command In NM72 if EST INTERACTION was not specified prior to requesting TABLE CWRES then the population variance covariance is evaluated at eta 0 C q 0 In NONMEM 7 1 0 and 7 1 2 regardless of INTERACTION setting in a previous EST statement C f is used nm730 doc 43 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 CPREDI CRESI CWRESI These are conditional with eta epsilon interaction pred res and wres values The conditional mode or conditional mean etas must be available from a previous EST MAXEVAL gt 0 command EPRED ERES EWRES The EPRED ERES EWRES are Monte Carlo generated expected or exact pred res and wres values and are not linearized approximations like the other diagnostic types The expected diagnostic items are evaluated using predicted function and residual variances evaluated over a Monte Carlo sampled range of etas with population variance Omega Define EPRED f n p 0 Q aq is the expected predicted value for data point j of subject i for a given subject evaluated by Monte Carlo sampling overall possible eta The probability density of eta p 0 0 2n is a multivariate normal distribution with eta variance Q The 1xz vector of EPRED for a given subject where n is the number o
292. s particular set of individual parameters were accepted So you may record them if you wish IF BAYES EXTRA 1 AND ITER_REPORT gt 0 AND TIME 0 0 THEN IF FIRST WRITE 0 THEN OPEN unit 50 FILE C NONMEM WORKA_ TRIM TFI PNM_NODE NUMBER FIRST WRITE 1 ENDIF WRITE 50 112 1X F14 0 5 1X 1PG12 5 ITER_REPORT ID CL V1 Q V2 OBJI NIREC 1 ENDIF SERROR include nonmem reserved general BAYES EXTRA REQUEST 1 Y F F EPS 1 IF BAYES EXTRA 1 AND ITER_REPORT gt 0 THEN IF FIRST WRITE2 0 THEN OPEN UNIT 51 FILE C NNONMEMNWWORKB TRIM TFI PNM NODE NUMBER FIRST WRITE2 1 ENDIF WRITE 51 I12 1X F14 0 2 1X 1PG12 5 ITER REPORT ID TIME F ENDIF Initial values of THETA STHETA 0 001 2 0 LN CL 0 001 2 0 LN V1 0 001 2 0 LN Q 0 001 2 0 LN V2 INITIAL values of OMEGA SOMEGA BLOCK 4 0 15 P 0 01 F 0 15 P 0 01 F 0 01 F 0 15 P 0 01 F 0 01 F 0 01 F 0 15 P Initial value of SIGMA SIGMA nm730 doc 201 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 0 6 P THETA 2 0 FIX 2 0 FIX 2 0 FIX 2 0 FIX OMEGA BLOCK 4 10000 FIX 0 00 10000 0 00 0 00 10000 0 00 0 00 0 0 10000 Prior information to the OMEGAS SOMEGA BLOCK 4 0 2 FIX 0 0 0 2 0 0 0 0 0 0 0 0 0 0 2 0 0 2 THETA 4 FIX SEST METHOD BAYES INTERACTION FILE example8b ext NBURN 10000 NITER 1000 PRINT
293. s process works well with the methods such as importance sampling SAEM or BAYES but works only partially for classical NONMEM methods or ITS If using with classical NONMEM methods or ITS it is better to set LAPLACE NUMERICAL although it does not solve the problem perfectly Classical methods rely on NMTRAN creating symbolic derivatives of the residual variance components with respect to eta which is used to create the proper individual objective function For this to occur NMTRAN has to see all of the relevant equations in the control stream file or the user must have the eta derivatives evaluated This method has some of the SDE differential equations and RVAR components calculated in subroutines SDE DER and SDE CADD hidden from NMTRAN Despite this problem classical NONMEM methods provide parameters using the SDE call routines that are similar although not identical to those when the SDE equations are placed in line into the control stream file To see how the SDE call routines work for each of the analysis methods see sde9 res that uses SDE f90 and compare the results with sde10 res which uses the in line equations The new methods except ITS do not need these NMTRAN constructed components so they work with the SDE call routines quite well As of NM73 numerical eta derivatives are now available for FOCE ITS so that it is not necessary for NMTRAN to see all the code or for the user to supply evaluation of the eta derivatives In th
294. s well In the main search algorithm finite central difference methods are also used These are evaluated as OO 1 h O 0 0 h 20h Numerical analysis of central finite difference methods recommend that the ideal relative step size h for the parameter theta 1 should be no greater than SIGL 3 If h is set to SIGL 3 then the resulting finite difference value itself will have approximately 2 SIGL 3 precision nm730 doc 54 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 The main search algorithm also utilizes pseudo second derivative type evaluations using forward difference methods For these calculations an ideal h would be 10 65 resulting in precision of second derivative constructs of about SIGL 3 Thus it is safest to set the step size h as specified by NSIG to be no more than SIGL 3 An internal SIGL in NONMEM specifies the precision to which the objective function itself actually the individual subject objective functions which sum to the total objective function is to be evaluated This internal SIGL is set to 10 As long as NSIG was set to a value less then or equal to 10 2 or 10 3 then the gradients would be evaluated to an appropriate precision to make the gradient search algorithm work efficiently With many subjects if SIGL 10 is the precision to which each individual objective function is evaluated and they are all of the same sign then the sum objective function could have a resulting precision of logi
295. scrambling 3 Owen plus Faure Tezuka type scrambling Other examples RANMETHOD SI Indicates sobol sequence with Owen scrambling for eta vector generation Since there is no integer in the first position of RANMETHOD indicated the general random number generator remains unchanged from the RANMETHOD specification previously specified or ran method 3 if none was specified earlier RANMETHOD 1S2 Indicates ranl type random number generator for general purposes sobol sequence with Faure Tezuka scrambling for eta vector generation The sobol sequence method of quasi random number generation can reduce the Monte Carlo noise in the objective function evaluation during importance sampling under some circumstances When the sampling density fits the posterior density well such as with rich continuous data the sobol sequence method does not reduce the Monte Carlo noise by much If you are fitting categorical data or sparse data and perhaps you are using the t distribution DF gt 0 for the importance sampling density then sobol sequence generation may be helpful in reducing Monte Carlo noise The RANMETHOD specification propagates to subsequent EST records in a given problem but does not propagate to SCHAIN or STABLE records In NM72 only DIRECT and IMP IMPMAP methods could utilize the Sobol quasi random method As of NM73 Sobol may be used for BAYES and SAEM methods as well From experience The SO and S1 methods produce considerable bias
296. section 1 47 EST Additional Output Files Produced on root phi for additional information one can obtain about eta shrinkage for each subject Residual error shrinkage EPSshrink for each residual error is evaluated for simple problems as 1009c 1 SD IWRES see 13 For more complicated problems the data and individual predicted values that contribute to assessing the shrinkage for each epsilon is not as straight forward For example if EPS 1 is proportional error to PK data and EPS 2 is proportional error to PD and they are not connected by an off diagonal sigma then EPS1 shrinkage pertains to PK data residuals and EPS2 shrinkage pertains to PD data residuals If they are related by an off diagonal SIGMA then their shrinkage is related and they will have similar or identical shrinkage values If two epsilons pertain to the same data such as proportional EPS and additive EPS for PK data Y F F EPS 1 EPS 2 nm730 doc 117 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Then the same epsilon shrinkage is associated with EPS 1 and EPS 2 However if F 0 for some data then such values contribute to EPS 2 shrinkage assessment but not to EPS 1 shrinkage assessment In such cases shrinkage to EPS 1 and EPS 2 may differ slightly where EPS 1 shrinkage incorporates only residuals to data with predicted values that are non zero and EPS 2 shrinkage incorporates residuals to all PK data 1 46 EST Format of Raw Ou
297. sed as follows For E field xPEw d indicates w total characters to be occupied by the number including decimal point sign digits E specifier and 2 digit magnitude d digits to the right of the decimal point and x digits to the left of the decimal point Examples E12 5 0 12345E 02 2PE13 6 12 12345E 02 If you are outputting numbers that are less than 1 0E 99 such as 1 22345E 102 there will be one less significant digit displayed to make room for the extra digit in the exponent To make room for a three digit exponent you may set the format as follows xPEw dEe where e is the number of digits to be provided for the exponent For example 1PE12 4E3 2 3456E 002 For F field Fw d indicates w total characters to be occupied by the number including decimal point sign and digits d digits to the right of the decimal point Examples F10 3 0 012 234567 123 For G field xPGw d For numbers gt 0 1 will print an F field number if the value fits into w places showing d digits otherwise will resort to xPEw d format For numbers lt 0 1 will always use xPEw d format If the user defined format is inappropriate for a particular number then the default format will be used for that number An example TABLE record could be nm730 doc 50 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 STABLE ID CMT EVID TIME NPRED NRES PREDI RESI WRESI CPRED CRES CWRES CPREDI CRESI CWRESI ZABF EPRED ERES EW
298. ses being 8 if it is not specified on the command line or in a defaults pnm file An example of its use is as follows Nmfe73 foce parallel ctl foce parallel res parafile mpiwini8 pnm nodes 4 The example control stream file foce parallel ctl is in the examples directory WINDOWS Setting up a network drive on Windows for multiple Computers Both FPI and MPI methods require the user to set up network drives to pass files between manager and worker computers If you are running your multiple process on multiple cores of just a single computer then you may skip this section From the worker computer select a directory or create a directory which you would like to have shared with the manager computer Suppose it is called c share On windows XP open my computers or right click on Start gt Explore go to directory tree right click on c share select properties then select Sharing and click on share this folder On other Windows systems there may be a different menu path to follow A suggested share name will be given You may keep this as is or change to a name you prefer Click on Permissions for user Everyone select Full control click on apply Consult your IT representative if you are not able to obtain privileges From the manager computer right click on the my computer icon and select map network drive Select an available drive letter which for this example will be w Then enter the computer name of the remote
299. sh component must be set up Check that you have ssh installed on both manager and worker computers From the manager run the standard Linux date program on the worker computer ssh n any computer date enter password If the date is returned from the worker computer you have ssh connection You might have to enter user account name ssh n my account any computer date For ssh to work in parallel computing you need to set up ssh so it does not always ask for your password From the manager computer ssh keygen t dsa Respond yes to writing to ssh and enter in a passphrase Copy id_dsa pub from the manager to the worker computer possibly via the share drive you had set up cp ssh id_dsa pub mnt share Then concatenate this manager created id_dsa pub to the authorized_keys file on the worker computer cd HOME chmod w ssh authorized_keys touch ssh authorized_keys cat id_dsa pub gt gt ssh authorized_keys chmod 400 ssh authorized_keys From the manager computer repeat the command ssh n any computer date it should ask you for the pass phrase then give you the date Do it again ssh n any computer date nm730 doc 151 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 the pass phrase should not be requested this time nor should a password be requested and a date from the worker computer should return During the parallelization process NONMEM sends a copy of its program to the
300. sider the following problem which is example 6 at the end of this document Data were simulated with 17 PK and 18 PD observations for each of 50 subjects receiving a bolus of drug followed by short infusion a week later The PK model has 2 compartments Vc k12 k21 with first order k10 and receptor mediated clearance Vmax Kmc The PD model is indirect response with receptors generated by zero order process k03 and removed by first order process k30 or via drug receptor complex Vmax Kmc There are 46 population parameters variances covariances and intra subject error coefficients and thee differential equations In the table below are listed the estimation times not including a COV step using various SIGL NSIG and TOL values Note that when not setting SIGL NM 6 method the problem would take a very long time When SIGL NSIG and TOL were set properly estimation times were much less with successful completions Of course as they say in the weight loss commercials individual results may vary and such great differences in execution times will not occur for all problems INSIG 3 INSIG 2 INSIG 1 TOL 6 TOL 6 TOL 4 Advan method SIGL 100 NM6 style SIGL 6 SIGL 3 9 gt 30 22 10 6 gt 24 17 13 new gt 20 8 5 2 1 17 The SIGLO level NM72 As of NONMEM 7 2 0 the user may obtain even greater control of the precision at which various parts of the estimation are performed by using the SIGLO option
301. sity the proposal density is centered at the previous sample position New samples are accepted with a certain probability The variance of the proposal density is adjusted to maintain a certain average acceptance rate IACCEPT This method requires more elaborate sampling strategy but is useful for highly non normally distributed posterior densities such as in the case of very sparse data few data points per subject or when there is categorical data In the first phase called the burn in or stochastic mode SAEM evaluates an unbiased but highly stochastic approximation of individual parameters semi integration usually 2 samples per individual Population parameters are updated from individual parameters by single iteration maximization steps that are very stable and improves the objective function usually in 300 5000 iterations In the second mode called the accumulation mode individual parameter samples from previous iterations are averaged together converging towards the true conditional nm730 doc 70 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 individual parameter means and variances The algorithm leads to population parameters converging towards the maximum of the exact likelihood The SAEM method is specified by EST METHOD SAEM INTERACTION Followed by one or more of the following options NBURN 2000 Maximum number of iterations in which to perform the stochastic phase of the SAEM method default 1000
302. some remain due to a ctrl C interrupt rerun the problem and look again for any non zero sized FILE again Repeat as needed By default maxlimz0 NMTRAN will set the LIM values to those listed in sizes f90 or to the minimum required whichever is larger As of NM73 if you set maxlim 1 on the command line then LIMI LIM3 LIM4 LIM13 and LIMIS those used during estimation and therefore by workers in a parallelization problem will be set to the size needed to assure no buffer files are used and everything is stored in memory for the particular prolem If you set maxlim 2 then LIMI LIM2 LIM3 LIM4 LIM5 LIM6 LIM7 LIM8 LIMII LIMI3 LIM15 and LIM16 are also sized to what is needed to assure that buffer files are not needed If you set maxlim 3 then MAXRECID will also be sized to MAXDREC the largest number of records in any individual MAXRECID sizes arrays involved in storing state variables during partial derivative estimates of sigmas and sigma like thetas to improve efficiency of the EM and Monte Carlo methods When setting maxlim 3 it is preferred to also use tprdefault or prcompile but not prdefault as NMTRAN s optional resizing of the PREDPP size parameter MAXRECID may conflict with the prdefault option To specify only a subset of LIM s to be sized by NMTRAN set maxlim to a number list enclosed within parantheses such as maxlim 1 2 3 11 16 which will have NMTRAN find size requirements for L
303. specified using the PRIOR statement available since NM 6 release 2 0 and described in the html Help manual use only NWPRI option for the new EST methods then normally the analysis is set up for three stage hierarchical analysis By default NOPRIOR 0 and this prior information will be used However if NOPRIOR 1 then for the particular estimation the prior information is not included in the analysis This is useful if you want to not use prior information during a maximization METHOD IMP CONDITIONAL IMPMAP SAEM or ITS but then use it for the Bayesian analysis METHOD BA YES As of NM73 when NOPRIOR 1 is set the estimation will not use TNPRI prior information TNPRI should only be used with FO FOCE Laplace estimations In previous versions of NONMEM NOPRIOR 1 did not act on TNPRI priors 1 29 A Note on Setting up Prior Information Prior information is important for MCMC Bayesian analysis but not necessary for maximization methods Of greatest importance are priors to the Omegas As a general rule if your data set consists of fewer subjects than 100 times the dimension of the Omega matrix to be estimated then you should have at least uninformative OMEGA prior information Priors to THETAS are nm730 doc 78 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 assumed multivariate normal and priors to OMEGAS and SIGMAS are assumed Wishart distributed Alternatively a residual variance in the form of its square root may b
304. st 4 times that of what one would normally set NSIG If evaluating only the S matrix then SIGL SIGLO TOL should be at least 3 times that of what one normally sets NSIG For example during EST NSIG 2 SIGL 6 TOL 6 may be sufficient but during COV you may need SIGL 12 TOL 12 to avoid positive nm730 doc 105 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 definiteness issues The MATRIX TOL and SIGL have no relevance to the variance results for a BAYES method which are derived from samples generated during the estimation step If TOL is set in the COV record but SIGL and or SIGLO are not then the TOL is not changed Also if TOL is set for the COV record then this TOL is used for all compartments ATOL NM72 The absolute tolerance option pertains to using ADVAN13 and as of NM73 to ADVANO as well where ATOL is the accuracy for derivatives evaluated near zero The same ATOL value is set for all compartments The ATOL by default is 12 Usually the problem runs quickly when using ADVANIS with this setting On occasion however you may want to reduce ATOL usually equal to that of TOL and improve speeds of up to 3 to 4 fold ATOL may be set at the EST or COV command Keep in mind that ATOL is changed for the COV step only if SIGL and or SIGLO are also specified at the COV record NOFCOV NM72 No COV step for any classical estimation steps This would be useful if you wanted EM estimation analyses with variance cov
305. st useful for very sparse sparse or rich data and for data with non normal likelihood such as categorical data The iterative two stage ITS method is best for rich data and rapid exploratory methods to obtain good initial parameters for the other methods The FOCE method is useful for rich data and in cases where there are several or more thetas that do not have ETA s associated with them nm730 doc 97 of 210 DIRECT IMP IMPMAP SAEM BAYES NONMEM Users Guide Introduction to NONMEM 7 3 0 1 37 Composite methods Composite methods may be performed by giving a series of SEST commands The results of the estimation method are passed on as initial parameters to the next SEST method Also any settings of options of the present method are passed on by default to the next SEST method One suggestion is to perform in the following order although trial and error is very important 1 Iterative two stage for rapid movement of parameters towards reasonable values 10 30 iterations 2 SAEM if model is complex or data are very sparse with 300 3000 iterations depending on model complexity Obtain maximum likelihood parameters 3 Importance Sampling if model is complex with 300 3000 samples 50 100 iterations depending on model complexity Obtain maximum likelihood parameters 4 Evaluate at final position by importance sampling Obtain maximum likelihood value and standard errors 5 Perform MCMC Bayesian analysis on your favori
306. state are as follows 1 Means this is the first call for the problem initializations will be done l See note below 2 Means this is not the first call and the calculation is to continue normally with no change in any input parameters except possibly TOUT and ITASK If ITOL RTOL and or ATOL are changed between calls with ISTATE 2 the new values will be used but not tested for legality 3 Means this is not the first call and the calculation is to continue normally but with a change in input parameters other than TOUT and ITASK changes are allowed in NEQ ITOL RTOL ATOL IOPT LRW LIW JT ML MU and any optional inputs except HO MXORDN AND MXORDS see IWORK description for ML and MU Note A preliminary call with TOUT T is not counted as a first call here as no initialization or checking of input is done Such a call is sometimes useful for the purpose of outputting the initial conditions Thus the first call for which TOUT T requires ISTATE 1 on input On output istate has the following values and meanings 1 Means nothing was done TOUT T and ISTATE 1 on input 2 Means the integration was performed successfully Means an excessive amount of work more than MXSTEP steps was done on nm730 doc 52 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 this call before completing the requested task but the integration was otherwise successful as far as T MXSTEP
307. stem and MPI system that is appropriate for NONMEM may be retrieved from a script file that could have the following environment variables defined compilerpath mpibinpath mpilibpath mpilibname Comments in these files are provided for instructions about each of these environment variables These paths will be temporarily added to the front of the PATH environment variable so that the appropriate compiler or MPI system is called to service NONMEM In the past conflicts with other installed fortran compilers from other applications would prevent the appropriate compiler from being used for the NONMEM system This location file method allows NONMEM to be forced to look in a particular location The location file should be called nmloc bat or nmloc by convention It may be specified at the nmfe73 command line by the locfile option for example nmfe73 myfile ctl myfile res locfile nmloc bat If locfile is not specified the nmfe73 script looks in the present working directory for nmloc bat windows or nmloc linux If this file is not found it looks in the top directory of the NONMEM installed directory Thus the file nmloc bat Windows or nmloc Linux in the top nonmem installed directory serves as the default location file and may be modified or used as a template and placed in the working directory or specified in the locfile option on the command line If a particular environment variable in the above list is not found or is not d
308. t then the method of earlier versions of NONMEM will be used with the cholesky elements uniformly varied over the interval l iaccept initial value and 1 iaccept initial value If DFS gt one million then SIGMA is fixed at the initial values If DFS 0 then the dimensionality of the entire SIGMA matrix is used as degrees of freedom CHAIN Record Any initial settings of THETA OMEGA and SIGMA that are read in by S EST METHOD CHAIN are applied only for the estimation step The SIML command will not be affected and will still use the initial settings given in THETA OMEGA and SIGMA statements or from an MSFI file To introduce initial THETAs omegas and sigmas that will cover the entire scope of a given problem use the CHAIN record nm730 doc 128 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SCHAIN FILE examplel previous txt NSAMPLE 0 ISAMPLE 1000000000 The following options are available for CHAIN and have the same actions as for SEST METHOD CHAIN FILE NSAMPLE ISAMPLE SEED RANMETHOD FORMAT ORDER CTYPE DF DFS IACCEPT NOLABEL NOTITLE Setting SEED or RANMETHOD in a CHAIN record does not propagate to EST METHOD CHAIN or any other SEST record ISAMPEND NM73 has a different action with CHAIN then with EST METHOD CHAIN If the option ISAMPEND is set to a value greater than ISAMPLE then NONMEM uniformly randomly selects one of these samples between ISAMPLE and ISAMPEND This is particularly
309. t EVID used is not 0 then the MDV value is set to 1 for that call If an XVID is 1 then the call to PK ERROR for that XVID is not made nor for the remaining XVID s If there is an EVID column the value in this column is not passed to PK ERROR unless XVID1 1 in which case a normal call on that record occurs The following is a control stream file to a stochastic differential equation SDE problem courtesy of Dr Christoffer Tornoe that uses the XVID data items examples Wde8 ctl in the examples SPROBLEM PK ODE HANDS ON ONE SINPUT ID TIME DV AMT CMT FLAG MDV EVID SDE QA XVID1 QB XVID2 QZ XVID3 SDATA sde8 csv IGNORE SSUBROUTINE ADVAN6 TOL 10 DP SMODEL COMP CENTRAL nm730 doc 161 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 COMP P1 STHETA 0 10 ALICE THETA 0 32 2 VD THETA 0 2 4 SIGMA STHETA 0 1 SGW1 SOMEGA 0 1 PloCL SOMEGA 0 0 2 ND SSIGMA 1 FIX PK SPK IF NEWIND NE 2 OT 0 TVCL THETA 1 cL TVCL EXP ETA 1 TVVD THETA 2 VD TVVD EXP ETA 2 SGW1 THETA 4 IF NEWIND NE 2 THEN AHT1 0 PHT1 0 ENDIF IF EVID NE 3 THEN Al A 1 A2 A 2 ELSE Al Al A2 A2 ENDIF IF EVID EQ 0 OBS DV IF EVID GT 2 AND SDE EQ 2 THEN RVAR A2 1 VD 2 THETA 3 2 K1 A2 1 VD RVAR AHT1 Al K1 OBS A1 VD PHT1 A2 K1 RVAR K1 ENDIF IF EVID GT 2 AND SDE EQ 3 THEN AHT1 A1 PHT1 0 ENDIF IF EVID GT
310. t distribution has larger tails and is useful for situations where the posterior density has a highly non normal distribution For very sparse data or highly non linear posterior densities such as with categorical data you may want to set DF to somewhere between 2 and 10 RANMETHOD nISImIP NM72 default n 3 Where n 0 4 m 0 3 By default the random number generator used for all Monte Carlo EM and Bayesian methods use the Knuth method ran3 of reference 5 We feel this is the best random number generator for many purposes However you may choose alternative random number generators n as follows n 0 4 0 ranO of reference 5 minimal standard generator 1 ranl of reference 5 Bays and Durham 2 ran2 of reference 5 3 ran3 of reference 5 Knuth 4 NONMEM s traditional random number generator used in SSIMULATION For special purposes a sobol 5 sequence method with or without scrambling 9 may be called upon and only for the purpose of creating quasi random samples of eta vectors To select the sobol method without scrambling add an 5 to RANMETHOD For example RANMETHOD 2S nm730 doc 68 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Selects random number generator ran2 for general purposes and sobol sequence for the eta vector generation The number m is reserved for the type of scrambing desired m 0 3 0 no scrambing so SO is the same as S 1 Owen type scrambling 2 Faure Tezuka type
311. t permitted Repeated inputs of SOMEGA or SIGMA may be entered as follows SOMEGA BLOCK 6 0 1 0 01 0 1 0 01 x2 0 1 0 01 x3 0 1 0 01 x4 0 1 0 01 x5 0 1 nm730 doc 22 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 The VALUES diag odiag feature allows one to set up initial values with diagonals diag and off diagonals odiag The above example could have been entered as SOMEGA B OCK For SOMEGA B OCK 6 6 VALUI FIX VALUI ES 0 1 0 01 fixed block such as for omega priors ES 0 15 0 0 A BBR DECLARE feature for abbreviated code NM73 Integers and arrays may be declared and used in abbreviated code SABBR DEC ARE DOS SABBR DEC ARE INT EG ER E 100 DOS I ETIM E 100 A BBR REPLACE feature for abbreviated code NM73 Any character string may be replaced In particular this allows for symbolic labeling to thetas etas and epsilons As an example subscripts to THETAS and ETAS can be given symbolic names SABBR REPLACE THETA CL THETA 4 SABBR REPLACE ETA CL ETA 5 CL THETA CL EXP ETA CL Replacement with selection by data item and parameter is permitted ETA 4 ETA 7 ETA 1 SABBR REPLACE TH SPK KA THETA OCC which is equivalent to SPK IF OCC
312. t replacement Should the sample selection become exhausted which would occur if CHAIN or CHAIN records are utilized for more than ISAMPEND ISAMPLE 1 times subsequent sample selection then occurs with replacement SELECT 3 uniform random selection of sample with replacement this is equivalent to SELECT 0 for SCHAIN 1 49 ETAS and PHIS Record For Inputting Specific Eta or Phi values NM73 Sometimes it is desired to bring in specific eta or phi values and using them as initial values just as is done for thetas using the THETA record The simplest syntax is to enter a single set of etas SETAS 0 4 3 0 3 0 5 0 from the control stream file All of the subjects in the data set will be given these set of initial values of etas Alternatively enter them as phi values convenient for EM methods PHIS 0 4 3 0 3 0 5 0 The eta values will then be evaluated as eta i phi i mu i for each eta where mu i 2mu i is evaluated according to their definitions in the PK section Alternatively enter initial etas and or phis for an entire set of subjects from a phi or phm in the case of mixture problems of a previous analysis SETAS FILE myprevious phi FORMAT s1pE15 8 TBLN 3 Where FORMAT should at least have the delimiter appropriate to read the file and TBLN is the table number in the file If TBLN is not specified then the first set of etas phis are brought in In matching the etas phis to the data set given in DATA of the control str
313. tas and Reporting Thetas NM73 Initial thetas in the THETA record may be functionally transformed with the THETAI or THI record and final thetas may then be reverse transformed for report purposes using nm730 doc 99 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 THETAR or STHR This has particular value when it is desired that the thetas by estimated within NONMEM in the log domain but you want the convenience of inputting and outputting them in the natural domain such as when performing linear MU referencing For example STHETAI THETA 1 NTHETA LOG THETAI 1 NTHETA THETA NTHETA 1 NTHETA NTHP LOG THETAI NTHETA 1 NTHETA NTHP Or STHETAI THETA 1 NTHETA LOG THETAI 1 NTHETA THETAP 1 NTHP LOG THETAPI 1 NTHP Where ntheta number of to be estimated thetas and nthp number of theta priors Or leave it to NONMEM to supply the range which is by default NTHETA NTHP STHETAI THETA LOG THETAI This record will convert any initial thetas in a THETA record or thetas obtained from a chain file but will not convert thetas from an MSF file Furthermore the variance to the theta priors will be appropriately converted when using PRIOR NWPRI PRIOR TNPRI receives variance covariance information from MSF files and this information is in the model theta domain For reporting thetas the inverse function should be supp
314. te model 200 1000 burn in samples having started at maximum no more is necessary 10000 30000 stationary samples Obtain complete distribution of parameters to obtain mean standard error confidence bounds An example control stream file follows Iterative two stage with 50 iterations SEST METHOD ITS INTERACTION NITER 50 SIGL 7 NSIG 2 SAEM with 200 iterations for stochastic mode 500 iterations for accumulated averaging mode SEST METHOD SAEM INTERACTION NBURN 200 NITER 500 Importance sampling for 10 iterations expectation step only this evaluates OBJF without moving population parameters Note that SIGL 7 that was set for the previous EST command is assumed for this SEST command as well SEST METHOD IMP INTERACTION ISAMPLE 1000 NITER 10 EONLY 1 MCMC Bayesian Analysis with 200 burn in samples and 10000 stationary samples SEST METHOD BAYES INTERACTION NBURN 200 NSAMPLE 10000 Here is the full control stream file SPROBLEM Setup of Data for Bayesian Analysis SINPUT SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT CLX V1X QX V2X SDIX SDSX SDATA samp5 csv SSUBROUTINES ADVAN3 TRANS4 At least An uninformative Prior on OMEGAS is recommended for MCMC Bayesian SPRIOR NWPRI NTHETA 4 NETA 4 NTHP 0 NETP 4 NPEXP 1 SPK MU 1 THETA 1 nm730 doc 98 of 210 NONMEM Users
315. te the control stream file name given at the command line of the nmfe73 script lt licfile gt substitute the entire licfile option including its value provided by the nmfe73 script For example licfile c mynonmem license nmlicense lic is substituted into lt licfile gt background substitute background switch if given by user on the nmfe73 command line lt parafile gt substitute parafile option such as parafile2myparallel pnm given at nmfe73 command line Never use the parafile switch on a worker process Substitution variables need not be used just as arguments to the nonmem executables that are loaded In some cases they are needed in other parts of the command line of the process launch nm730 doc 139 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 or in the directory listing of SDIRECTORIES In such cases it is not desired to substitute the entire option value string but just the value portion Where the value of the option itself is to be substituted use option For example suppose the nmexec option is used to specify an alternative nonmem executable name In such cases you would specify lt lt nmexec gt gt in place of the usual nonmem exe 3 psexec d w worker2 cmd exe C lt lt nmexec gt gt control stream This principle of using lt gt versus lt lt gt gt applies to the other substitution parameters as well You may also define your own substitution para
316. ted followed by the cumulative probability CUM associated with each eta followed by the joint density probability of that support point if default or MARGINALS was selected If ETAS was selected then instead of cumulative probabilities the support point eta vector that best fits that subject ETM is listed root npe NM73 The expected value etas and expected value eta covariances ETC are listed for each problem or sub problem Because only one line is written per problem or sub problem the column header is displayed unless NOLABEL 1 only once for the entire NONMEM run However each line contains information of table number problem number sub problem number super problem and iteration number root npi NM73 The individual probabilities are listed in this file The header line unless NOLABEL 1 is written only once at the beginning of the file per NONMEM run Each line contains information of table number problem number sub problem number super problem iteration number subject number and ID This is followed by the individual probabilities at each support point of which there are NSUPP NSUPPE or NIND of them whichever is greater The line with Subject number 0 contains the joint probability of each support point the same as listed in root npd under the column PROBABILITY For each support point K the joint probability is equal to the sum of the individual probabilities over all subject numbers I Thus row of subject n
317. ted integer values LINR or LINLINR R for round produces values rounded to the nearest integer LOG or LOGLIN A covariate logarithmic time linear interpolation is used for the covariate value for the inserted records A T or R suffix results in truncated or rounded integer values respectively LINLOG A covariate linear time logarithmic interpolation is used for the covariate value for the inserted records A T or R suffix results in truncated or rounded integer values respectively LOGLOG A covariate logarithmic time logarithmic interpolation is used for the covariate value for the inserted records A T or R suffix results in truncated or rounded integer values respectively Another example SFINEDATA CMT 3 3 EVID NEXT 2 indicating to create two inserted records for a given fine time point For the first inserted record CMT 3 and EVID of the next original record For the second inserted record CMT 3 and EVID 2 Inserted records will be given the following values by default unless over ridden by a data item specification such as SFINEDATA EVID 2 DV EVID 0 MDV 1 Times may be entered as numerical values or in hh mm ss format Data sets with DATE TIME records may also be processed but then TSTART and TSTOP must be in numerical hours or hh mm ss format Once finedata produces the augmented data file in this example example6b csv then a suitable NM TRAN control stream file that would take advantag
318. ten code does not compute the eta derivatives When OPTMAP 1I is present values of G and H are ignored during eta optimization This may be used to test user coded derivatives because two runs one with OPTMAP 1 and one without it should give very similar values for the OBJV WRES etc if the user coded derivatives are correct That is the nm730 doc 58 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 analytic derivatives in G and H are ignored and this option may be used when analytic derivatives are difficult to compute e g user supplied code such as SDE 2 Nelder Mead method which uses a secant method rather than relying on derivatives ETADER 0 default NM73 In evaluating the MAP objective function the term log Det V must be evaluated to obtained the marginal or integrated posterior density where V is the eta Variance matrix based on the subject s posterior density 0 Expected value V using analytical first derivatives 1 Expected value V using forward finite difference numerical first derivatives Needed if not all code evaluating F and Y derivatives with respect to eta are available for processing by NM TRAN or in user supplied code 2 Expected value V using central finite difference numerical first derivatives Needed if not all code evaluating F and Y derivatives with respect to eta are available for processing by NM TRAN or in user supplied code That is the analytic derivatives in G and H are i
319. ter than manager psexec any computer d w c share worker3 cmd exe C nonmem exe IRECTORIES Names of directories as a manager sees them NONE FIRST DIRECTORY IS THE COMMON DIRECTORY Make it NONE if no common directory is to be used This is the best option workerlN NEXT SET ARE THE WORKER directories nm730 doc 145 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 3 worker2 4 w share worker3 This directory is on a different computer from manager SIDRANGES USED IF PARSE_TYPE 3 1 1 50 2 51 100 After an estimation step is performed the worker processes exit For the next estimation step that follows if there is one the manager will reload the worker processes For the FPI method with TRANSFER_TYPE 0 a PARAFILE file name may be given specific to a SEST command EST METHOD IMP INTERACTION NITER 20 PARAFILE myparallel imp pnm EST METHOD 1 INTERACTION PARAFILE myparallel foce pnm If no parallel file is given for an estimation method it takes the PARAFILE name of the previous EST command If no PARAFILE option was given for the first SEST method then it takes the value given in the command line switch parafile If no parafile switch was given then the default name parallel pnm is assumed If parallel pnm file does not exist then NONMEM runs on a single CPU If you want worker processes to remain resident until all estimations and problems listed in the c
320. the thetas and OMEGA 1 1 in indestb cov The perfect match of the variance between indestm and indestb was done by ensuring both performed 2 derivative information matrix analyses in indestm by selecting LAPLACE in the SEST step and in indestb by selecting MATRIX R in the COV step What adds power to this technique over the typical single subject analysis method is that some of the parameters may be shared For example in examples indestms ctl instead of each subject finding its own residual variance coefficient a shared SIGMA 1 1 is estimated SPROB THEOPHYLLINE POPULATION DATA SINPUT ID DOSE AMT TIME CP DV WT SDATA THEOPP SSUBROUTINES ADVAN2 ETA 1 MEAN ABSORPTION RATE CONSTANT 1 HR ETA 2 MEAN ELIMINATION RATE CONSTANT 1 HR THETA 3 SLOPE OF CLEARANCE VS WEIGHT RELATIONSHIP LITERS HR KG ALING PARAMETER VOLUME WT SINCE DOSE IS WEIGHT ADJUSTED CALLFL 1 A THETA 1 ETA 1 THETA 2 ETA 2 L THETA 3 ETA 3 SC CL K Q STHETA 0 0 FIXED X3 SOMEGA 1 0E 06 FIXED X3 SETAS 3 08 04 nm730 doc 182 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 SERROR IPRED F Y F EPS 1 SSIGMA 0 2 SEST METHOD 1 INTERACTION LAPLACE MAXEVAL 9999 PRINT 1 NOHABORT FNLETA 0 MCETA 1 STABLE ID DOSE TIME DV IPRED NOAPPEND NOPRINT FILE INDESTMS TAB STABLE ID KA K CL NOAPPEND FIRSTONLY NOPRINT FILE INDESTMS PAR COV MATRIX R Thus while each subject finds its own K KA and CL in the form of unco
321. thods for FOCE Laplace NM73 The SUBP option in SIML may be greater than 9999 new limit is 27 1 All EM Bayes methods are now estimated with the INTERACTION option on by default unless NOINTERACTION is specified When NOPRIOR 1 is set the estimation will not use TNPRI prior information TNPRI should only be used with FO FOCE Laplace estimations In previous versions of NONMEM NOPRIOR 1 did not act on TNPRI priors nm730 doc 13 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 New elements are available in the NONMEM report xml file termination nfuncevals termination sigdigits termination txtmsgs which catalog termination text messages by number which can be mapped to source txtmsgs f90 etabarn ebvshrink np objective function and total cputime If inputted omega or sigma elements are not positive definite because of rounding errors a value to the diagonal elements will be added to make it positive definite A message in the NONMEM report file will indicate if this was done In root ext Iteration 100000006 indicates 1 if parameter was fixed in estimation 0 otherwise See 1 46 EST Format of Raw Output File Thetas may be inputted and reported in their natural domain even when linear MU referencing See 1 38 THETAI THI AND THETAR THR Records for Transforming Initial Thetas and Reporting Thetas NM73 Covariance assessment may be turned off for a particular estimation See NOCOV 0 nm73 in s
322. timation step SEST METHOD 0 MAXEVAL 9999 PRINT 5 NSIG 3 Second estimation step SEST METHOD CONDTIONAL NSIG 4 0 2 will first result in estimation of the problem by the first order method using as initial parameters those defined by the THETA OMEGA and SIGMA statements Next the first order conditional estimation method will be implemented using as initial parameters the final estimates of THETA OMEGA and SIGMA from the previous analysis Up to 20 estimations may be performed within a problem For all intermediate estimation steps their final parameter values and objective function will be printed to the raw output file Many settings to options specified in a EST method will by default carry over to the next SEST method unless a new option setting is specified Thus in the example above PRINT will remain 5 and MAXEVAL will remain 9999 for the second EST statement whereas NSIG will be changed to 4 and METHOD becomes conditional An exception to this rule are NOTHETABOUND NOOMEGABOUND and NOSIGMABOUND in which these options pertain to all of the estimations in the series within a PROB In NM710 NM712 and NM720 these options must be given with the very first SEST record in the problem With NM73 these options may be placed with any of the SEST records but will still apply to all EST records in the problem The EM and Monte Carlo estimation methods particularly benefit from performing them in sequenc
323. tion Increased number of mixed effects levels Random effects across groups of individuals such as clinical site can be modeled in NONMEM Sites themselves may be additionally grouped such as by country etc See section 1 43 Adding Nested Random Levels Above Subject ID NM73 Easy to code inter occasion variability ETA s to be referenced by an index variable related to the inter occasion data item See section l 4 Expansions on Abbreviated and Verbatim Code NM72 NM73 Symbolic reference to thetas etas and epsilons See section 4 Expansions on Abbreviated and Verbatim Code NM72 NM73 Priors for SIGMA matrix A SIGMA prior matrix may be added assumes inverse Wishart distributed to provide prior information for SIGMAs See section 1 29 A Note on Setting up Prior Information Optimizing settings for some options in SAEM and Importance Sampling User may request an optimal ISAMPLE setting be determined for each subject by NONMEM for SAEM and IMP rather than relying on a pre specified value Similarly user may request IACCEPT and DF settings be optimized for each subject by NONMEM when performing IMP For BAYES and SAEM user may request that most appropriate CINTERV AL be determined based on the degree of Markov chain correlation across iterations rather than the user having to assess appropriate CINTERVAL by trial and error See section 1 25 Monte Carlo Importance Sampling EM and I 27 Stochastic Approximation Expectation Maximizati
324. tion inverse covariance etc This has been corrected 13 When a series of STABLE statements without FILE specification is followed by STABLE statements with FILE specification not all tables print out and an error is issued in the NONMEM report file OERROR IN WRITING FILE TABLE FILE USER FORMAT ERROR IN FORMAT SWRITE Work around is to set LFORMAT NONE and RFORMAT NONE on the first STABLE record with a FILE option 14 Problems with temporally over lapping dosing records and with SEST and COV records may fail during a parallelization run at the COV step Work around is to perform the COV step without parallelization 15 Repetition variables and data items RPTI RPTO RPT_ useful for repeated records for convolution problems did not work properly for estimation methods other than FO This has been corrected in NONMEM 7 3 nm730 doc 15 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 16 If the partial derivative of MTIME with respect to any eta is negative such as MTIME 1 THETA 5 ETA 5 then the predicted value of F and its derivatives will probably be incorrect The bug exists in all versions of PREDPP from NONMEM VI to NONMEM 7 2 IT is corrected for NONMEM 7 3 A work around is to use ALAG s in place of MTIME s but this is somewhat complicated A fix is to edit the file PRED f90 or PRED f for older versions in the pr directory Locate the characters DSUM DSUM GG UIMTGG MTPTR K 1 Change to DSUM DSUM ABS
325. tory into your standard run directory Then execute the following from your standard run directory Nmfe73 foce parallel ctl foce parallel res parafile fpilinux8 pnm nodes 4 where the values of nodes should be no greater than the number of cores available on your computer For instructional purposes here is an example pnm file for FPI on Linux systems note TRANSFER TYPE 20 GENERAL NODES 3 PARSE TYPE 2 PARSE NUM 50 TIMEOUTI 300 TIMEOUT 20 PARAPRINT 0 TRANSFER TYPE O0 NODES number of nodes that is process whether cores or computers SINGLE node NODES 1 MULTI node node means process whether cores or computers NODES gt 1 WORKER node NODES 0 parse num number of subjects to give to each node parse type 0 give each node parse num subjects parse type 1 evenly distribute numbers of subjects among available nodes Ne Ne Ne Ne Ne Ne Ne Ne nm730 doc 152 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 7 7 parse type 2 load balance among nodes parse type 3 assign subjects to nodes based on idranges parse type 4 load balance among nodes taking into account loading time This setting of parse type will assess ideal number of nodes If loading time too costly will eventually revert to single CPU mode timeouti seconds to wait for node to start if not started in time deassign node and give its load to next worker until next iteration timeout minutes to wait for node to compe
326. tput File A raw output file will be produced that provide numerical results in a columnar format The raw output file name is provided by the user using a new FILE parameter added to the EST record A raw output file has the following format A header line that begins with the word Table such as TABLE NO 4 MCMC Bayesian Analysis Goal Function AVERAGE VALUE OF LIKELIHOOD FUNCTION This header line provides the analysis text same as given on the METH line in the main report file followed by the goal function text same as given on the OBJT line in the report file The next line contains the column headers to the table such as this is actually all on one line in the file ITERATION THETA1 THETA2 THETA3 THETA4 SIGMA 1 1 OMEGA 1 1 OMEGA 2 1 OMEGA 2 2 OMEGA 3 1 OMEGA 3 2 OMEGA 3 3 OMEGA 4 1 OMEGA 4 2 OMEGA 4 3 OMEGA 4 4 OBJ This is followed by a series of lines containing the intermediate results from each printed iteration six significant digits based on the PRINT option setting 10 1 73786E 00 1 57046E 00 7 02200E 01 2 35533E 00 6 18150E 02 1 82955E 01 3 18352E 03 1 46727E 01 4 38860E 02 2 58155E 02 1 45753E 01 4 58791E 02 6 28773E 03 5 06262E 02 1 50017E 01 2301 19773603667 For the above example each of the values up to the next to last one occupies 13 characters including the delimiter in this example the delimiter is a space The last value is the objective function which occupies 3
327. ts selected more than once NOREPLACE NM73 SIML BOOTSTRAP 50 SUBP 100 NOREPLACE EST METHOD 1 INTERACTION In the above example 50 unique subjects are to be randomly selected from the simulation template data set The NOREPLACE feature is reasonable if there are many more than 50 subjects to choose from template set for example 1000 subjects in the template and for each nm730 doc 61 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 sub problem 50 of them are randomly chosen without replacement that is without repeating a subject STRAT NM73 SIML BOOTSTRAP 50 SUBP 100 NOREPLACE STRAT CAT A single stratification data item may be entered In the above example the data item CAT serves as the stratification This splits the data set into distinct sub sets guaranteeing a specific number of subjects will be selected from each category For example if in the base data set CAT has values of 1 or 2 with 33 subjects in group 1 and 67 subjects in group 2 out of 100 total subjects then exactly 33 of of subjects from group 1 will be randomly selected out of 50 total 16 and exactly 67 of subjects will be randomly selected from group 2 34 This has value when desiring that a bootstrap analysis maintain the same proportion of subjects belonging to certain categories such as gender or age bracket To stratify by both age bracket and gender create a stratification data item that would be for example valued 1 for subjects w
328. ts using step size h that varies theta at the 3 significant digit This results in 1 significant digit precision remaining in evaluating the finite difference gradients The search algorithm is now attempting to maximize the objective function to 3 significant digits when it is working with gradients that are accurate to only 1 2 significant digits This results in inefficient advancement of the objective function causing NONMEM to make repeated evaluations within an iteration as well as iterations for which the objective function is barely moving NONMEM can then spend many hours trying to obtain precision in its parameters which are impossible to obtain Eventually it may stop because the maximum iterations were used up or when it realizes that it could not reach the desired precision With this understanding of the search algorithm process and recognizing the complex relationship between the step size needed for each parameter and the finite difference method used in each part of the algorithm the optimization algorithm was changed to allow the user to specify SIGL and for the algorithm to set up the appropriate step size for a given finite difference method based on the user supplied SIGL While some trial and error may still be required by the user for a given problem certain general rules may be considered nm730 doc 55 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 1 Set SIGL NSIG and TOL such that SIGL TOL NSIG SI
329. uation with an Application to the Population Pharmacokinetics of Gliclazide Pharmaceutical Research 2006 23 9 2036 2049 4 Nguyen THT Comets E Mentre F Extension of NPDE for evaluation of nonlinear mixed effect models in presence of data below the quantification limit with applications to HIV dynamic model J Pharmacokinet Pharmacodyn 2012 39 499 518 5 Press WH Teukolsky SA Vettering WT Flannery BP Numerical Recipes The Art Of Scientifc Programming gu Edition Cambridge University Press New York 1992 pp 269 305 6 Press WH Teukolsky SA Vettering WT Flannery BP Numerical Recipes The Art Of Scientifc Programming 2 Edition Cambridge University Press New York 1992 pp 180 184 7 Savic RM Karlsson MO Evaluation of an extended grid method using nonparametric distributions AAPS Journal 2009 11 3 615 627 8 Baverel PG Savic RM Karlsson MO Two bootstrapping routines for obtaining imprecision estimates for nonparametric parameter distributions in nonlinear mixed effects models J Pharmacokinetics and Pharmacodynamics 2011 38 1 63 82 9 Hee Sun Hong And Fred J Hickernell Algorithm 823 Implementing Scrambled Digital Sequences ACM Transactions on Mathematical Software Vol 29 No 2 June 2003 Pages 95 109 10 Lavielle M Monolix Users Manual computer program Version 2 1 Orsay France Laboratoire de Mathematiques U Paris Sud 2007 11 Bennett Racine Poone and Wakefield MCMC for
330. ulation residual assessment WRESI hence WRESI will differ from NWRESI with FO INTERACTION There are other individual residual values available mostly as place holders in the system but these have no additional statistical value They are NIPRED IPREDI NPRED IPRD CIPREDI CIPRED EIPRED EPRED NIRES IRESI NRES IRS CIRESI CIRES EIRES ERES NIWRES V n 0 y f q 0 IWRESI NIWRES IWRS EIWRES V y f pn Rd MDVRES 0 NM73 default Set MDVRES to 1 in the ERROR or PRED routine if you do not want to include a particular value for weighted residual assessment This may be useful when for example this data point is assessed by a non normal distribution likelihood such as the PHI function for below detection limit values in which F FLAG is set By default if at least one data value of a given subject is fitted with a non normal distribution likelihood then population weighted residual diagnostics are not assessed for any of the data for that subject By setting MDVRES 1 to these particular below detection values the weighted residual algorithm can assess the remaining normally distributed values for that subject For example SERROR SD THETA 5 IPRED LOG F DUM LOQ IPRED SD E I UMD PHI DUM F TYPE EQ 1 THEN F FLAG 0 nm730 doc 47 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Y IPRED SD ERR 1 ENDIF IF TYPE EQ 2 THEN F FLAG
331. umber I column of support K contains the individual probability IPROB LK The sum of the individual probabilities over all support points for any given line subject is equal to 1 NIND The format of the file is fixed at 1PE22 15 and cannot be changed It is intended for use in further analysis by analytical software and is designed to report the full double precision information of each probability nm730 doc 124 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 root fgh NM73 This file is produced if the user selects SEST NUMDER 1 or 3 The file lists the numerically evaluated derivatives of Y with respect to eta where G I 1 partial Y with respect to eta i G I J 1 Second derivatives of Y with respect to eta i eta j H I 1 partial Y with respect to eps i H j 1 partial Y with respect to eps i eta j root agh NM73 This file is produced if the user selects SEST NUMDER 2 or 3 The file lists the analytically evaluated derivatives of Y with respect to eta from the PK ERROR and or PREDO routines in FSUBS where G I 1 partial Y with respect to eta i G LJ4 1 Second derivatives of Y with respect to eta i eta j not always evaluated by FSUBS H I 1 partial Y with respect to eps i H i j l partial Y with respect to eps i eta j root cpu NM73 The cpu time in seconds is reported in this file It is an accurate representation of the computer usage whether single or parallel process The same
332. uming an analysis SPROB RUN example3 from adltrim2s SINPUT C SET ID JID TIME CONC DV DOSE AMT RATE EVID MDV CMT VC1 K101 VC2 K102 SIGZ PROB SDATA example3 csv IGNORE C SSUBROUTINES ADVAN1 TRANS1 SMIX P 1 THETA 5 P 2 1 0 THETA 5 NSPOP 2 SPK MU_1 THETA 1 MU 2 THETA 2 MU 3 THETA 3 MU 4 THETA 4 VCM DEXP MU_1 ETA 1 10M DEXP MU_2 ETA VCF DEXP MU_3 ETA 3 10F DEXP MU_4 ETA Q 1 IF MIXNUM EQ 2 Q 0 V O VCM 1 0 Q VCF Q K10M 1 0 Q K10F S1 V SERROR Y F F EPS 1 STHETA 4 3 2 9 4 3 0 67 0 7 SOMEGA BLOCK 2 0 0 027 SOMEGA BLOCK 2 05 01 06 SIGMA 0 0 PHIS FILE etafile3 phi phm FORMAT S1PE15 7 TBLN 3 SEST METHOD CHAIN FILE etafile3 chn ISAMPLE 5 NSAMPLE 0 SEST METHOD IMP MAPITER 0 CTYPE 3 INTERACTION NSIG 3 PRINT 1 NITER 3 nm730 doc 131 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 Or use FNLETA 2 to use the etas that were brought in to evaluate predicted values without performing a new population estimation SPROB RUN Example 1 from samp51 SINPUT C SET ID JID TIME DV CONC AMT DOSE RATE EVID MDV CMT CLX VIX QX V2X SDIX SDSX SDATA etafile csv IGNORE C SSUBROUTINES ADVAN3 TRANS4 PK MU_1 THETA 1 MU 2 THETA 2 MU 3 THETA 3 MU 4 THETA 4 LCL MU_1 ETA 1 CL DEXP LCL LV1 MU_2 ETA 2 V1 DEXP LV1 LQ MU_3 ETA 3 Q DE LV2 MU_4 ETA 4 V2 DEXP LV2 S1 V1 RROR IPRED F F F EPS 1 Init
333. un and NONMEM may stop the run because one of the sizing parameters was too low The user should then insert a SIZES record in the control stream file set the offending sizing parameter to the appropriate value and run the problem again SIZES f90 no longer contains parameters DIMPKS and DIMRHS and DIMRV for NMTRAN The arrays sized by these parameters are dynamically allocated to whatever size is necessary for the abbreviated code in the current control stream All other arrays for NMTRAN can be increased in size if necessary with SIZES As of NM73 NMTRAN determines the maximum number of observation records MDV 0 that occur in any subject among all data files used in the entire control stream file If this value is greater than the NO value listed in SIZES f90 it will set NO to this larger size Thus users no longer have to be conscientious of sizing the NO parameter However there is no guarantee that NMTRAN will correctly assess NO for the entire scope of the control stream file for all types of problems Should this occur NONMEM may issue an error and the user will need to set the NO value with a SIZES record When PREDPP PK ERROR INEN etc is used NMTRAN also creates a sizes file called prsizes f90 This file contains sizing and other parameters needed by PREDPP Some parameters PD LVR which sets the prsizes parameter PE are the same as in FSIZES and have the same values Some PC PCT PIR PAL MAXFCN MAXRECID are uni
334. unction such as a viral model in which a very small change in the potency of an anti viral agent results in widely varying time of return of viral load This results in standard errors being poorly assessed for thetas that do not have inter subject variances associated with them Setting DERCONT to 1 slows the analysis but can provide more accurate assessments of SE in such models The DERCONT works only for the Monte Carlo EM algorithms such as IMP and SAEM CONSTRAIN 1 NM72 A built in simulated annealing algorithm has been put in place for NONMMEM 7 2 0 Simulated annealing slows the rate of reduction of the elements of the OMEGA values during the burn in phase of the SAEM method allowing for a more global search of parameters The subroutine CONSTRAINT performs this algorithm when the option CONSTRAIN is set to 1 or 5 where 1 is the default setting This is by the constraint algorithm starting the Omegas at 1 5 times the initial values and then controlling the rate at which the Omegas shrink during each iteration CONTRAIN 2 or 6 performs simulated annealing on sigma parameters CONSTRAIN 3 or 7 performs simulated annealing on both OMEGA and SIGMA parameters CONSTRAIN 0 or 4 performs no simulated annealing on non zero valued OMEGAS The user may modify the subroutine CONSTRAINT that performs the simulated annealing algorithm The source code to the CONSTRAINT subroutine is available from the source directory as constraint f90 and the
335. us run are appropriate for the present run and testing that the present prsizes f90 is not different from the present run Typically you can expect that the nmfe73 script will do a PREDPP recompile when any of the following sizes change LVR PD PC PCT PIR PAL MAXFCN This could happen if the user changes the values via SIZES Also NMTRAN will resize LVR if the number of OMEGA entries changes and it will resize PD if the number of data items listed in SDATA changes Size changes are all listed in prsizes f90 in the PREDPP temporary recompile directory The PREDPP files selected for linking listed in LINK LNK can change if the SSUBROUTINES statement which specifies ADVAN TRAN is changed You may force PREDPP recompilation in case the run does not appear to execute properly when no recompilation occurs by setting the prcompile switch nmfe73 mycontrol ctl myresults res prcompile On the other hand if the nmfe73 script for some reason believes there is a change in the previous run from the present run but you are convinced there is not a change you may force the skipping of the PREDPP compilation step and use the compiled files from the previous run by adding the argument prsame at the end of the command line For example nmfe73 mycontrol ctl myresults res prsame If you are repeatedly going between two or more problems so that often they need to be PREDPP recompiled and you want to save time you can specify a unique tem
336. ux nmfe73 mycontrol ctl myresults res background gt console txt In Unix Linux you can additionally append amp to the command to execute it in the background you must also use background option when using amp nmfe73 mycontrol ctl myresults res background gt amp console txt amp And periodically monitor the rerouted file tail f console txt For the more adventurous user you may modify the nmfe73 scripts for alternative behaviors Additional options are available to make execution of the nmfe73 script more flexible From the nmfe73 command line the user may enter a run directory that is different from the directory in which the nmfe73 script is launched rundir c my favorite dir Where rundir is the run directory if it is different from the present working directory you must make sure all user dependent input files control stream file msf files and data files are available in that run directory The user may also enter an alternative name for the constructed executable nm730 doc 29 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 nmexec nonmem2 specifies an alternative executable name than the default nonmem exe windows or nonmem Linux To turn off production of the XML output file root xml where root is the root name of the control stream file use the option xmloff Beginning in NM73 an additional feature of the execution script file is that the path to the fortran compiler sy
337. variance V is always evaluated at conditional mode from the most recent FOCE ITS estimation or conditional mean from the most recent IMP IMPMAP SAEM analysis eta 1 so that ECC V f EPRED f n EPRED p q0 Q dn and ECWRES ECC ERES As with CWRES the eta hat conditional mode or mean values must be available from a previous SEST MAXEVAL gt 0 command Thus ECWRES is the Monte Carlo version of CWRES while EWRES is the Monte Carlo version of CWRESI In NM72 if SEST INTERACTION was not specified prior to requesting STABLE CWRES then the residual variance is evaluated at eta 0 V 0 In NONMEM 7 1 0 and 7 1 2 regardless of INTERACTION setting in a previous EST statement V 1 is used NPDE The NPDE is the normalized prediction distribution error reference 2 takes into account within subject correlations also a Monte Carlo assessed diagnostic item For each simulated vector of data yx ESRES y EPRED its decorrelated residual vector is calculated ESWRES EC ESRES and compared against the decorrelated residual vector of observed values EWRES such that 1 K pde 7 gt 6 EWRES ESWRES k 1 For K random samples where x 1 for x gt 0 Oforx lt 0 For each element in the vector Then an inverse normal distribution transformation is performed npde pde nm730 doc 45 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 NPD The NPD is th
338. we show that the addresses to the worker computers listed in the file mpd hosts could be loaded using the mpdboot f command The f option is also available in mpirun so this information may be supplied within the parafile for example l mpirun PWD n 1 0 f mpd hosts nonmem Some Advanced Technics For Defining the PARAFILE for an MPI System Because the MPI system communicates completely via ports and not via file transfer as the FPI system does one can set up a parafile in which an MPI command is repeated for several nodes even though they may point to the same directory Here is an example which makes creating a PARAFILE for an MPI system versatile GENERAL NODES 8 PARSE_TYPE 2 TRANSFER_TYPE 1 PARAPRINT 0 COMPUTERS 2 SCOMMANDS 1 mpiexec wdir SPWD n 1 nonmem 2 4 wdir SPWD n 1 host MY MANAGER COMPUTER nonmem wnf 5 8 wdir HOME n 1 host MY WORKER COMPUTER nonmem wnf SDIRECTORIES 1 8 NONE 5 mnt workerl In this example node 1 is defined as usual as the manager process Then processes 2 through 4 are defined using a command that is repeated for each of these processes it is copied 3 times in the resulting nmmpi script file that is eventually executed Yet processes 2 4 all point to the default current directory of the manager PWD Furthermore the DIRECTORIES entries for these processes is NONE That means the three worker processes which are loaded on the manager computer are sha
339. worker computer and then loads it there Therefore the worker computers must be of the same operating system although not necessarily same version as the manager computer For Intel fortran the worker computer does not have to have Intel Fortran installed For gfortran static option for the FPI is used in the nmfe73 script which makes gfortran portable to the worker computer without requiring the gfortran share library libgfortran so 3 If for some reason you needed to remove the static option then gfortran requires its share library available for the worker process and in the path designated by the manager s LD_LIBRARY_PATH setting such as LD LIBRARY PATH SHOME gcc trunk lib SHOME libg SLD LIBRARY PATH Export LD LIBRARY PATH where HOME gcc trunk lib is the library path for the manager s gfortran and HOME libef is the path on the worker computer containing at least the file libgfortran so 3 You may place these lines in the bashrc file Therefore if upon loading NONMEM on the worker computer a message is displayed indicating that certain share files are missing etc then you may need to either install gfortran or selectively make the share file available Setting up FPI on Linux For a quick test on a single multi core computer try the following Copy foce parallel ctl and examplel csv from the NONMEM examples directory fpilinux8 pnm from the NONMEM run directory and beolaunch sh from the NONMEM run direc
340. y now be used in abbreviated code DVALUE MAX VAL1 VAL2 VAL3 However this function should not be involved in evaluation of the objective function IF THEN statements should be used for those for example DVALUE VAL1 IF VAL2 gt DVALUE DVALUE VAL2 IF VAL3 gt DVALUE DVALUE VAL3 GAMLN Function NM73 The GAMLN function returns an accurate evaluation of the logarithm of the gamma function It can be used in the evaluation the factorial FAC exp gamln x 1 0 Where FAC X X X 1 X 2 1 It is more accurate that the Stirling s approximation and may be used in abbreviated code in the evaluation of the objective function Declaring Reserved Variables NM73 Some useful reserved variables are explicitly recognized by NMTRAN that can be used by the user There are however many other variables that are generally internal to NONMEM and often are not needed by users except occasionally which are not explicitly recognized by NMTRAN and so cannot be used in abbreviated code but must be used with verbatim code at beginning of line For example the variable ITER REPORT is available that contains the present iteration number as reported to the console or NONMEM report file that may be useful to be accessed within the PK SERROR or PRED code A convenient means of accessing this variable as well as letting NMTRAN allow you to use that variable in abbreviated code is to place its MODUL
341. ybe the remote computer is a different platform than the manager computer and needed a different executable nm730 doc 158 of 210 NONMEM Users Guide Introduction to NONMEM 7 3 0 MPICH2 communication between a Linux and Windows operating system has not been attempted so it is not known if this would work anyway Note that host MY MANAGER COMPUTER had to be identified on the worker processes that were being launched locally The mpiexec command gets confused if it has to deal with several lines containing different computer names So it is best not to leave the host switch up to default once you get past the manager processor line The wnf switch must be carefully used Make sure that LIMI LIM3 LIM4 LIM13 and Lim15 are appropriately sized so that the buffer files named FILEXX do not have to be used Or as of NM73 you may set maxlim 1 or higher on the nmfe73 command line Then LIMI LIM3 LIM4 LIM13 and Lim15 those used during estimation and therefore by workers in a parallelization problem will be set to the size needed to assure no buffer files are used and everything is stored in memory for the particular prolem If you set maxlim 2 then LIMI LIM2 LIM3 LIMA LIM5 LIM6 LIM7 LIM8 LIM13 LIM15 and LIM16 are also sized to what is needed to assure that buffer files are not needed If the buffer files do need to be used then use switch wf Each worker process will make a series of files named WK1_FILE for

NONMEM USERS GUIDE INTRODUCTION TO NONMEM 7.3.0

Contents

Download Pdf Manuals

Related Search

Related Contents