Home

The CCP4 Suite

1. o o 9 3 Automated model building o 9 4 Difference Map generation o a 9 5 Why is protein refinement difficult 9 6 Free Rfactor s bue 0 un a ee ee ee 9 7 Validation gross and overall errors oo 9 7 1 Validation v2 424 0b o A A 9125 ATT OUS u e say sect xh gn A re gie 9 1 3 Bad practice 2 20 22 BA Sea Re Ee es 10 The Fast Fourier Transformation 19 1 Introduction 2 a fr WA A eae ee a ete ed 10 2 How it works 200 a ee ae ie 10 2 1 Example au sen ee een SO A SR A S 10 2 2 Why is it last 3 224 Kerr da e ad 10 3 Crystallographic FFT 2 mu onen 10 4 Programs 2 2 0 e ee er A III The propagation care and feeding of a CCP4 installation 11 Installation Wel First Off Sea a Bo St en Bee eb BE g 11 2 Directory structure ee ee 11 3 Building under Unix e e 11 3 1 Preliminaries sce o wel be ee area Se we ER 11 3 2 Unpacking the files o oo 11 3 3 Environment variables and ccp4 setup 11 3 4 Configurati0n e 11 35 Building s ree E ine nee nd 11 3 6 Testing 22 220883 Bo a ee les 11 3 7 Problems 2 eg 4 eve Bee See Re ee Ha 11 3 8 Saving space and shared libraries 11 3 9 Installing Updates 11 3 10 Installation summary LA MCCPAD e ra RA a he ou 11 5 X windows programs 2 2
2. Now you re ready to configure the source directory for the peculiarities of your version of Unix Here we assume that you are using the normal vendor supplied compilers and that you ve used the default paths suggested in ccp4 setup so that the suite is built and installed in the CCP4 directory If you have more complicated requirements consult the full story on configure in 11 6 If you wish to do multiple installations for multiple machine types then see below With CCP4 as your current directory type configure system where system describes one of the known operating systems corresponding to what you re running This might be something like irix sunos etc To find the possibil ities try configure help configure will grind away and after some sanity checks e g on the values of the environment variables you ve defined will establish the correct directories and Makefiles Uf configure doesn t support your system you ll have to consult the porting guide Chapter 15 and or seek advice from CCP4 If you wish to do multiple installations for multiple machine types then you should run the script dupt ree first This will produce a further script containing informa tion on the source directory tree This second script should then be run once for each machine type in each case a duplicate source tree is created with soft links to the original source The configure script should then be run within each duplicate s
3. 1980 Goodfel low et al 1989 Dodson et al 1996 see also Tronrud 1994 The principal CCP4 refinement program is now refmac which is undergoing active development Infor mation in this chapter is liable to become out of date and the reader is recommended to read the latest re fmac documentation 9 1 Least squares structure refinement The CCP4 suite provides the program restrain for least squares structure refine ment The old program prolsq has now been made obsolete by the Maximum Likeli hood program refmac see section 9 2 The least squares program rest rain incor porates a number of differences from re fmac see section 9 1 1 and may be useful Alternative refinement programs include TNT CNS SHELX93 9 1 1 Comparison of restrain and prolsq The CCP4 manual Here is some discussion of the differences between the least squares refinement pro gram restrain and the obsolete prolsq Most of these points also apply to a com parison of restrain and refmac restrain does constrained anisotropic thermal parameter refinement using the TES translation libration screw rotation model Unconstrained anisotropic refine ment is not feasible without atomic resolution data i e 1 Aor better so this is out of the question for all but a handful of very small proteins The end result of the TLS analysis should give some insight into secondary structure or domain motions refmac5 now does TLS refinement as well A major differen
4. Sawyer L Isaacs N amp Bailey S eds 1993 Data Collection and Processing CCP4 Daresbury Study Weekend nos DL SCI R34 ISSN 0144 5677 Warring ton WA4 4AD UK Daresbury Laboratory for Daresbury Laboratory Sheldrick G M 1991 Heavy atom location using SHELXS 90 In Wolf et al 1991 Singh A K amp Ramaseshan S 1966 The Determination of Heavy Atom Positions in Protein Derivatives Acta Cryst 21 279 280 Smith David amp Howell Lin 1992 Identification of Heavy atom Derivatives by Nor mal Probability Methods J Appl Cryst 25 81 86 Stout George H amp Jensen Lyle H 1989 X ray Structure Determination a practical guide Second edn New York John Wiley Ten Eyck L F 1985 Fast Fourier Transform Calculation of Electron Density Maps In Wyckoff et al 1985 Ten Eyck Lynn F 1973 Crystallographic Fast Fourier Transforms Acta Cryst A29 183 191 Ten Eyck Lynn F 1977 Efficient Structure Factor Calculation for Large Molecules by the Fast Fourier Transform Acta Cryst A33 486 492 Terwilliger Thomas C amp Eisenberg David 1983 Unbiased Three Dimensional Re finement of Heavy Atom Parameters by Correlation of Origin Removed Patterson Functions Acta Cryst A39 813 817 Terwilliger Thomas C amp Eisenberg David 1987 Isomorphous Replacement Effects of Errors on the Phase Probability Distribution Acta Cryst A43 6 13 Tickle I J 1985 Review of space group genera
5. The ftnchek program is useful if your compiler is unhelpful about warning of non standard non portable things see The CCP4 manual 14 4 Stealing code 83 PROGRAM name declarations C advertise the version CALL CCPRCS LUNSTO 1 name Date 2006 02 05 16 53 41 C call the pre processor CALL CCPFYP C initialise the MTZ routines CALL MTZINI C open input MTZ file CALL LROPEN 1 HKLIN C open input map file CALL MRDHDR IN2 MAPIN C close the files CALL LRCLOS 1 CALL MRCLOS IN2 CALL CCPERR 0 Normal termination END Figure 14 1 Skeleton example CCP4 program using MTZ and map files 13 4 If tnchek objects seriously to your code we ll want it fixed The free Tool pack software is even more useful or the commercial NAGware Tools development of it 14 4 Stealing code There are many advantages to stealing code from reputable sources but make sure it is freely distributable and you acknowledge it as required In particular note that the copyright conditions in Numerical Recipes Press et al 1986 prevent code from there being used directly and anyway it is not well thought of by the experts Free solid numerical code can be had in abundance from the Netlib archives by Internet ftp or mail server 14 5 Using the pre processor in programs To use the pre processor in existing code the only thing you need to add to you
6. output may contain error messages and warnings If the program en counters a fatal error it will print a message at the end of the output along with timing information If it runs successfully a message like Normal termination will be printed with the timing information The condition code set by the program will indi cate success or failure as appropriate for the operating system e g in Unix zero for success non zero on error Many of the programs produce graphical information Output from the programs should be redirected to a log file which can be viewed with Loggraph or the older xloggraph Some log files are marked up in HTML and can therefore be viewed in a browser such as Netscape Graphs appear as Java applets 3 5 Rudiments of reflexion files The MTZ format for reflexion data has some features which it is helpful to understand since CCP4 depends heavily on the standard file formats or the information the files contain and the MTZ one in particular The data are stored in the files which usually have the extension mtz notionally as columns of real or integer numbers each with a label The first three columns are the Miller indices of the reflexion with labels H K and L The programs expect you to set up a correspondence between the labels in your data file and names of data columns they recognise Standard MTZ files have one record per reflexion During the initial data processing unmerged data are s
7. 16 1 1 Orientation data 16 1 2 Standard column names and types 16 1 3 Missing Data Treatment 22 o 16 2 Maps 16 3 Coordinates 16 4 PLOT84 graphics files 16 5 Library data files 16 5 1 Symmetry operators syminfo lib V Appendices References Program index Index The CCP4 manual 85 85 85 85 95 97 103 109 Preface Credits The CCP4 manual gt Alice in Wonderland the best book on computer science for the layman The best book on anything for the layman Anon Usenet book recommendations This manual attempts to provide documentation of the CCP4 program package at a level above that of detailed input requirements for individual programs Sadly over the years material has become incomplete in places Suggestions for improvement are welcome particularly in the form of new passages for inclusion Although there is tutorial material included here novices will need background material e g from textbooks like Blundell amp Johnson 1976 Glusker amp Trueblood 1985 Luger 1980 Stout amp Jensen 1989 Drenth 1994 Giacovazzo et al 1992 Perutz 1992 Rhodes 1993 and sources such as Hahn 1992 It is also worth noting that the HTML formatted documentation included with the CCP4 suite itself contains much supplementary material see for example the docu ments under the General section of the main program index Cross re
8. 45 46 phased 40 translation search 44 45 unique reflexions 7 Unix 1 2 15 16 19 21 65 67 73 77 82 85 common error 19 112 pitfall 16 shells 19 unmerged data 19 update notifications 73 URL 76 Usenet 19 77 verbose output 21 version numbers 73 VMS 15 73 82 83 85 web site 65 66 The CCP4 manual Program index Wilson plot 7 23 windows 85 Winn Martyn 1 Wolf Wojtek vii World Wide Web 77 writeups 15 WWW see World Wide Web X Windows 7 12 13 27 66 69 70 107 Xentronics 7 Zelinka Jan 87 Colophon Colophon 113 This manual was typeset using I4TFX in mainly Times Helvetica Courierand Com puter Modern maths rendered by dvips at 300dpi Various common style options and special definitions were used see the source The bibliographic information was produced using bibtex and the index using makeindex some figures were drawn with xfig and converted to PostScript which was included with the psfig macros Since the editor considers himself a competent IATEX hacker the source might serve as a useful model for producing similar documents such as theses The bibliography file in BibTEX format might also be useful And to answer the question on everyone s lips the structure on the front cover is the DHFR example from the ribbon distribution 4Look it up The CCP4 manual
9. Sheila on the phone and then THIS happened but be sure to include any thing relevant It is sometime helpful to know what scientifically you were hoping to do if this is unusual Richard Stallman writes The fundamental principle of reporting bugs usefully is this report all the facts If you are not sure whether to state a fact or leave it out state it This holds true across all problem reporting systems for computer software or social injustice or motorcycle maintenance It is especially important in the software field due to the major differences seemingly insignificant changes can make a changed variable a missing semicolon Submit only one problem with each problem report If you have multiple prob lems use multiple reports This aids in tracking each problem and also in analysing the problems associated with a given program The CCP4 manual Chapter 13 Resources Could you tell me please which way I ought to go from here Alice s Adventures in Wonderland 13 1 CCP4 Web Pages 13 2 Anonymous ftp The CCP4 Web Pages are located at http ww ccp4 ac uk We now make extensive use of these pages and this should be your first stop when searching for information on CCP4 resources and activities The CCP4 programs can be obtained from Daresbury via Internet anonymous ftp It should be noted however that this in no way relieves the end user of the need to return a licence agreement A copy can
10. This is implemented by t fc Some of the published results are impressive The function has the form T x TO x O x 7 14 where TO x is like the translation function T 7 4 2 1 with a normalisation factor and has the form Xn Jobs h k x h TO x gt 7 15 n Is h O x is a packing function that measures the interpenetration of the molecules Ic h O x Eo 7 16 NYnFm h The CCP4 manual Chapter 8 Phase Improvement by Density Modification Solvent Flattening and Molecular Averaging etc You boil it in sawdust you salt it in glue You condense it with locusts and tape Still keeping one principal object in view To preserve its symmetrical shape The Hunting of the Snark 8 1 Density Modification 8 1 1 The Problem 8 1 2 The Method The CCP4 manual Approximate electron density maps phased by such methods as MIR or anomalous scattering may or may not be enough to deduce the structure of the molecule Exami nation shows that they fall into two groups 1 Those maps in which there are visible structural features and which it is possible at least to attempt an interpretation Once a significant portion of the map has been interpreted it is often possible to use this information to improve the rest of the map 2 Those maps which do not appear to contain structural motifs from which it is not possible to attempt an interpretation The distinction may seen obvious but in fact we could
11. XtalView is a complete package for solving a macromolecular crystal structure by isomorphous replacement including building the molecular model X Windows based Mc Ree 1993 The CCP4 manual Chapter 3 General CCP4 usage Tt s not a modern iconic user friendly menu driven color PC compatible user interface replied the queen in a tone that would need to come up two notches to be vaguely considered condescending It happens to suit me just fine retorted Alice What are you an engineer or something asked the 7 of spades Alice in DIGITALand anon 3 1 Program documentation Writeups for the programs are currently in HTML one per program and can be viewed through the local Web browser by typing ccp4help Formatted versions are in the CDOC directory and can be viewed conveniently with an editor However the CCP4 installation makes it possible for you to use the man utility you should be able to use things like man fft orman k fft ona Unix system this is providing the ccp4 setup file has been modified correctly see 811 3 3 2 Running the programs Information flows to and from the programs in two ways e input and output data files are connected as specified by command line argu ments given after the name of the program to be invoked e parameters and option specifications are read on the standard input stream and a log file of printed output is produced on the standard output stre
12. bin pxbin lib ccp4 lib make Figure 11 1 configure example Don t follow this slavishly There are corresponding Make file variables that you can override e g make FOPTIM g Suppose you support several types of system and share the CCP4 source and system independent data files across a network with a distributed file system such as NFS AFS or RFS Then it may be convenient to have a sub directory in CCP4 for each system in which to build and possibly in which to keep the resulting binaries Directories for binaries can be anywhere though As an example a configuration used on the Daresbury Iris is given in fig 11 1 which includes debugging options and scattered directories Some people require only the CCP4 library to support other systems and don t want to configure install the CCP4 programs as such Usually configure will expect to find the subdirectory src in the directory specified using the srcdir flag containing at least the makefile skeleton You can avoid this requirement so you don t have to fetch src tar Z by ftp for instance by using the flag onlylib as an argument to configure This will avoid configuring the src and unsupported directories Build the library using make onlylib at the top level The onlylib flag may be abbreviated uniquely The CCP4 manual Chapter 12 Support bug reports etc 12 1 Release policy Mine is a long and a sad tale said the Mouse turning to Alice an
13. data sftools 12 Reflection data file utility program including some density map handling sigmaa 8 30 49 50 54 55 Phase combination Read 1986 sketcher 10 Monomer library sketcher and graphical interface to libcheck solomon 9 Modifies the electron density maps by averaging solvent flipping and protein truncation sortmtz 12 18 23 25 Sorting MTZ file s sortwater 11 26 Assign waters to nearest protein atoms stereo 12 Reconstruction of 3D coordinates from measurements of a stereo diagrams stgrid 9 Generates stereographic projection from polarrfn for measuring angular coordinates stnet 9 Stereographic net for use with polarrfn plots superpose Secondary structure allignment surface 11 Accessible surface area symfit Unsupported Fit best molecular transformations to sets of crystallo graphic coordinates e g heavy atom coordinates related by non crystallographic symmetry tffc 9 39 40 46 Translation function tlsanl 10 54 TLS analysis after restrain topdraw 12 Sketchpad for creating protein topology diagrams topp 11 An automatic topological and atomic comparison program for protein structures tracer 12 Reduced cell calculations truncate 7 23 51 89 Converts intensities to amplitudes French amp Wilson 1978 undupl 7 Remove duplicates from madnes data after absurd unique 7 24 55 Generate unique reflection data set uniqueify 7 55 Script to complete dataset and add free R flags v
14. estimate is refined during the maximum likelihood phase refinement The influence of such errors was originally discussed by Blow and Crick 1959 See also Blow and Rossmann 1961 Bricogne 1991 Otwinowski 1991 and Read 1991 One hopes that the use of more than one derivative will solve the phase ambiguity unfortunately it happens that in many cases there is a strong tendency for the probabil ity distribution of the phases to be bimodal and a strategy for making the best choice needs to be developed In fact the obvious choice of taking the maximum of the prob ability distribution can lead to big errors where other choices have a certain probability of being correct We need a compromise between these different choices in such a way as to minimise the mean squared error in the electron density This best Fourier The CCP4 manual 38 Isomorphous Replacement is calculated with phases corresponding to the centroid of the distribution weighted by the figure of merit m which is a measure of the reliability of the phase determination If the probability is sharp and unimodal m will be close to 1 but if the probability is almost uniformly distributed along the circle m will be very small Therefore the best Fourier has coefficients mFpe est 6 18 It can be shown Dickerson et al 1961 that m is related to the mean value of the cosine of the error in the phase angle for the reflexion m cos AQ 6 19 for
15. the model molecule in the crystal lattice Having found the correct orientation and position we can calculate phases from the model and combine them with the observed structure factor amplitudes The structure factors thus obtained and the corresponding electron density map contain a strong bias towards the starting model but they are usually sufficiently close to the correct values for a successful refinement More information on MR may be found in the Study Weekend proceedings Machin 1985 Dodson et al 1992 7 3 The rotation function The rotation function is used to find the correct orientation of the model in the crystal lattice This may be achieved by testing the agreement between the Patterson functions calculated from the model and from the data at various relative orientations The con cept is simple the rotation function can be thought of as a convolution in a rotational The CCP4 manual 7 3 1 Model 7 3 1 Model 7 3 2 Resolution limits 41 space At the relative orientation that superposes the model onto the crystal the product of the two Patterson functions which are maps of interatomic vectors should have a large value The function was originally formulated as R C f Pi x P x dV 7 1 where P and P are the two Patterson functions C is a matrix defining a rotation of x with respect to x and U is the volume of integration For mathematical convenience this can be reformulated as an integral o
16. weight of some sort phase probability coefficients Hendrickson Lattman BATCH number M ISYM packed partial reject flag and symmetry number any other integer any other real DOXW gt SUTUZACOOU TEE 16 1 3 Missing Data Treatment In a typical series of diffraction experiments not all Bragg reflexions for a given resolution range are in fact recorded Hence after truncate some reflexion data records may be entirely missing from the MTZ file although the reflexion indices lie within the measured resolution range It is strongly recommended that index sets are made complete within the desired resolution range a script to do this is provided in CETC uniqueify The MTZ file will then contain records where there are indices but no measured data e g 0 0 2 MD MD 0 0 4 5170 230 0 0 6 1567 0 57 0 This means that it is easy to estimate completeness and later programs can restore values if possible Furthermore a particular reflexion may be recorded for the native protein but not for a derivative and the corresponding combined reflexion data record should indicate missing data for the derivative To date missing data has been indicated in a variety of ways For example a zero standard deviation is taken to mean that the corresponding datum e g structure factor amplitude is missing In all cases however the indicator is a number upon which arithmetic operations can erroneously be performed This convention has now
17. 2 5 Map and structure factor calculation 9 2 2 4 3 Direct methods 2 2 4 4 MAD phasing almn Crowther fast rotation function Obsoleted by amore polarrfn Kabsch s fast polar rotation function with plot stnet stgrid Generate stereographic net plots for use with polarrfn rfcorr analyzes correlations between cross and self rotation functions getax Real space searching for rotation axis of a D n or C n multimer tffc space group general translation function mapsig peak search and statistics on signal noise for translation function map Also sum product ratio of two maps rsearch R factor search MERLOT a complete molecular replacement package by Paula Fitzgerald Fitzgerald 1988 CNS rotmat gives X PLOR MERLOT amore equivalent rotation angles acorn ab initio procedure for the determination of protein structure at atomic resolu tion Yao Jia Xing rantan Direct Method module for the determination of heavy atom positions in a macro molecule structure or to determine a small molecule structure sapi SHELX 97 Comprehensive package of ab initio structure solution Patterson peak analysis unconstrained refinement etc MULTAN Find phases by direct methods Most useful for finding heavy atom sites HySS Hybrid Substructure Search mlphare abs madsys Geometric approach by Wayne Hendrickson B Weis SHARP Statistical Heavy Atom Refinement and Phasing program see http babinet globalphasing com r
18. Bulk solvent correction for translation search and rigid body refinement steps of AMoRe cad 8 12 25 29 30 40 91 Collecting assorted reflection data cavenv 11 Calculates cavities in macromolecular structures ecp4i 11 13 54 ccp4mg 10 11 chainsaw 8 Molecular Replacement model truncation cif2mtz 12 Program to convert mmCIF structure factors e g from PDB to MTZ cif2xml 13 Program to convert mmCIF files to XML combat 7 23 24 Produces an MTZ file in multirecord form suitable for input to SCALA replacement for ROTAPREP compar 11 Unsupported Comparison of coordinates contact 11 26 Calculates various contacts in protein structure coordconv 12 25 27 Interconverts various coordinates formats cooT 10 11 cpirate 9 Statistical phase improvement cross_validate 13 Validate harvest files for deposition crossec 8 Interpolates X ray cross sections and computes anomalous scattering factors f f crunch2 Direct method phase extension Data Harvesting Manager 13 103 104 Program index data harvesting manager tool for managing data harvesting files detwin 7 Detwin merohedrally twinned reflection data difres Unsupported Compares two files of atomic coordinates in PDB for mat distang 11 26 40 Calculates distances and angles in protein molecule dm 9 10 48 49 Various density modification techniques canned solvent flatten ing Sayre s equation histogram matching and iterative skeletonisa
19. Figure 6 1 MIR structure solution The main steps in obtaining initial phases by MIR are shown in fig 6 1 and a tutorial example is available in CEXAM tutorial These steps are e Scale derivative data to native using either scaleit or fhscal Kraut scal ing probably better Be particularly careful about the observations standard The CCP4 manual 29 30 Isomorphous Replacement deviations turn on weighting by them iff you believe what your data processing produces e Find heavy atom sites 1 Solve the Patterson for the first derivative to give the major site s see fig 6 1 and rsps vecsum 2 Use direct methods to find sites SHELX pat see multan Note however that these methods will not in general be able to decide the enan tiomorph of the heavy atom site Also you must check that all coordinates refer to the same origin e Refinement and phasing 1 Heavy atom refinement and phasing mlphare Maximum Likelihood Phasing and Refinement has several options for this 2 Heavy atom refinement vecref This vector space refinement may be better than mlphare s reciprocal space refinement e Difference Fouriers to find new heavy atom sites in first and other derivatives fft peakmax gt Npo with fft calculating difference maps e Back to calculate and check Pattersons for other derivatives It is advisable to do the same procedure starting with a different derivative to check for errors or bias do try to start
20. If the integration radius is close to one of the cell dimensions then the integration in Patterson space could include an origin peak one lattice translation from the origin introducing large and meaningless artifacts into the rotation function The best way to avoid the problem is to keep the radius of integration smaller than the smallest cell dimension minus twice the resolution limit i e the diameter of the origin peak Alternatively we can subtract the origin peak This can be done by dividing the data into resolution shells and subtracting from each shell its average In fact this can conveniently be combined with the normalisation because by definition the average E is 1 so all one has to do to get an origin subtracted normalised Patterson is to use E 1 instead of just E 7 3 8 Non crystallographic symmetry In MR it is usually helpful to have non crystallographic symmetry Firstly knowing the local symmetry can help us pick the correct solution of the cross rotation function 7 3 8 1 Native Patterson In general the effect of a non crystallographic rotation axis is to produce a pseudo Harker plane through the origin and perpendicular to the local symmetry axis The Harker plane transforms in the reciprocal space to produce a more easily interpretable peak in the rotation function in the direction of the rotation axis In some cases how ever it is instructive to examine the native Patterson separately In particular when t
21. amp Jones T A 1996b Efficient Rebuilding of Protein Structures Acta Cryst D52 829 832 Kraulis P 1991 MOLSCRIPT a Program to Produce Both Detailed and Schematic Plot of Protein Structures J Appl Cryst 24 946 950 Kraut J Sieker L C High D F amp Freer S T 1962 Chymotrypsinogen a Three dimensional Fourier Syntheis at S Resolution Proc Nat Acad Sci 48 1417 1424 Lamzin Victor S amp Wilson Keith S 1992 Automated Refinement Procedure In Dodson et al 1992 Laskowski R A Arthur M W Mac Moss D S amp Thornton J M 1993 PROCHECK a program to check the stereochemical quality of protein structures J Appl Cryst 283 291 Lattman E E amp Love W E 1970 A Rotational Search Procedure for Detecting a Known Molecule in a Crystal Acta Cryst B26 1854 1857 Leslie A G W 1988 A Reciprocal Space Algorithm for Calculating Molecular Enve lope using the Algorithm of B C Wang CCP4 Daresbury Study Weekend nos DL SCI R26 ISSN 0144 5677 Warrington WA4 4AD UK Daresbury Labora tory for Daresbury Laboratory Luger Peter 1980 Modern X ray Analysis on Single Crystals Berlin New York Walter de Gruyter Machin P A ed 1985 Molecular Replacement CCP4 Daresbury Study Week end nos DL SCI R23 ISSN 0144 5677 Warrington WA4 4AD UK Daresbury Laboratory for Daresbury Laboratory Machin P A Campbell J W amp Elder M eds 1980 Refinement of Pr
22. are substantially larger and consequently take more time to read from disc even if the decoding is relatively cheap and space isn t a consideration An efficient text encoding such as NA4 presumably defeats the object or clarity e With clear text you can t in general preserve bit patterns on reading and writing and may lose accuracy particularly on repeated i o consider 100 iterations You re anyway probably at the mercy of the compiler s i o library which may well not use the optimal algorithms for text gt binary conversion anyhow 16 1 Reflexions MTZ The CCP4 manual The MTZ reflexion file format uses fixed length records with in general four bytes for each data item REAL 4 with a minimum of 3 columns and currently a maximum of 200 columns of data per record though these limits are imposed only by the software parameters Additional information title cell dimensions column labels symmetry information resolution range history information and if necessary batch titles and orientation data is contained in labelled header records The columns of the reflexion data records are identified by alphanumeric labels held as part of the file header in formation The user relates the item names used by the program to the required data items as identified by the labels by means of assignment statements in the program control data The file contains basically two classes of records header records and reflexion data recor
23. be found on the CCP4 website at http www ccp4 ac uk ccp4license Commercial users of CCP4 must obtain the right to access the source from CCP4 be fore any transfer takes place The current Internet address of this service is ftp ccp4 ac uk IP number 148 79 112 134 Users should log in with the id ftp or anonymous and their e mail address as pass word The sub directory ccp4 contains the relevant files Transfers using this service are logged The files are held as compressed tar files to reduce the size of them There are several files described by a README A typical Unix ftp session may go as follows not all output from ftp is shown but the general idea is given with your input slanted Your ftp command may have a different interface ftp ftp ccp4 ac uk Connected to ftp ccp4 ac uk Name anonymous Guest login ok send your complete e mail address as password Password d love dl ac uk Welcome to the CCP4 ftp distribution service 230 Guest login ok access restrictions apply ftp gt cd ccp4 current linux ftp gt binary ftp gt get README more ftp gt get ccp4 core tar gz ftp gt get phaser cctbx tar gz ftp gt bye You now need to uncompress and untar the files e g cd CCP4 tar xzf ccp4 core tar gz 13 3 Electronic mailing lists 13 3 1 The CCP4 Bulletin Board The CCP4 manual The CCP4 Bulletin Board is intended to be a forum for discussion to query aspects of the CCP4 Program Suite to pu
24. be little point in including less than say half the self vectors On the other hand some cross vectors in the summation should not be fatal David Blow points out that fortunately most proteins have an approximately spherical shape and a of search radius of 75 80 of the molecular diameter includes 90 of the self vectors Ian Tickle recommends using 75 of the minimum diameter of the molecule the shortest edge of the minimal box into which it fits Jorge Navaza has suggested using a radius equal to the maximum distance of atoms from the molecule s centre of mass limited by the smallest dimension of the minimal box and such that the volume of the integration sphere doesn t exceed the volume of the box The choice of dimensions of the P1 unit cell for the model is straightforward It must be at least the size of the molecule s minimal box plus the integration radius in each direction Navaza suggests the at least should involve adding the resolution to this sum 7 3 5 Omitting weak terms 7 3 6 Normalising The rotation function is approximately proportional to F so that it will be dominated by the largest terms The small terms can usually be omitted from the calculations if computing speed is a problem Fortunately these days and with this type of problem computing time is not usually a serious limitation In general if there are no compelling reasons data should not be omitted Normalising involves dividing data into res
25. been discarded in favour of representing missing data as special values namely IEEE NaN or VMS Rop All relevant programs check for the presence of NaNs or Rops in input The CCP4 manual 90 Name H K L IC M ISYM BATCH I P SIGI SIG FRACTIONCALC IMEAN SIGIMEAN FP FC FPHn DP DPHn SIGFP SIGDP SIGFPHn SIGDPHn PHIC PHIM PHIB FOM WT HLA HLB HLC HLD FREE FreeR_flag Type wr A HATS OOO UU Tran os RG sea z A W W File formats Item Miller indices Centric flag 0 for centric 1 for acentric Partiality flag and symmetry number Batch number Intensity 7 Selected mean intensity ol standard deviation ol Calculated partial fraction of spot Mean intensity Olmean gt Native F value Calculated F F value for derivative n Anomalous difference for native data F F7 Anomalous difference for derivative n oFP standard deviation oDP OF n oDPHn Calculated phase Most probable phase Phase figure of merit weight ABCD H L coefficients free R flag as program label free R flag as file label 1256M Isym where M is the partiality flag 0 for full 1 for partial and the symmetry number sym is normally 0 with F F F 2 for anomalous data with both components measured but is 1 if F F or 2 for F F Figure 16 1 MTZ standard column labels and types 16 1 2 The CCP4 manual 16 2
26. coordinates R m s differences as a function of residue atom type and B values contact Calculates various types of contacts and analyses water hydrogen bonding See also act distang Calculates intra and inter molecular distances dyndom Determines dynamic domains when two conformations are available Hay ward amp Berendsen 1998 dssp Definition of secondary structure produces dictionary of secondary structure Kabsch amp Sander 1983 gensym Generates all symmetry related sites from a list of input atoms and produces a list of all sites which fall within the volume of the defined volume geomcalc Does various geomtry calculations on a molecule hgen Generates hydrogen atoms for a protein coordinate file with standard geometry 1sqkab Least squares fit of two sets of coordinates ncont Analyses contacts between subsets of atoms in a PDB file pdbcur A curation tool providing various analyses and manipulations of PDB files pdbset Various useful manipulations of PDB files e g add CRYST and SCALE lines generate symmetry related subunits rename chains renumber residues transform coordinates polypose Superposition of multi domained structures procheck Comprehensive stereochemistry checking rotamer List amino acids whose side chain torsion angles deviate from Richardson s Penultimate Rotamer Library Dirk Kostrewa rwcontents Count atoms by type and other analyses sc Program to analyse shape complementarity o
27. down to alanine Similarly differences in the loop regions the insertions in particular should be removed On the other hand one should not remove too much of the model and should bear in mind that structure is generally conserved better than sequence and that two cor responding pieces of chain in related proteins may differ chemically but still have a similar fold Including many wrong atoms in the model increases the overall noise level whereas removing too many correct atoms reduces the signal we are looking for Here one tries to balance having a complete model and having a wrong one If several related structures are known one could try to construct the best model by assembling from previously superposed models the pieces that are most homologous to our protein It is sensible to try several models if they are available Interpretation is easier if they are all oriented in the same way to begin with Next we consider how best to maximise the signal we want to extract from the cal culation First of all what resolution limits are most useful For a case of identical structures we would use the data to the highest available resolution In practice the model is not identical with the expected structure and a resolution limit of at least two or three times the expected mean coordinate difference should be used Low resolution terms should be omitted as they contain information mostly about solvent rather than protein This means that the
28. finement are usually required as well as several rounds of map inspection and manual readjustments of the model An alternative to least squares refinement is to minimise a residual based on maximum likelihood ideas In recent years techniques employing molecular dynamics in crystallographic refinement have become increasingly popular Thermal motions are simulated to help the refinement overcome the local minima The programs used in refinement are summarised in 2 2 8 Throughout the process of refinement one often needs to assess the quality of the model Various programs for the analysis of coordinates are summarised in 2 2 9 The complexity of macromolecules makes clear pictorial presentation of their struc tures particularly tricky The plotting programs are summarised in 2 2 10 CCP4 provides a collection of useful utility programs which may be used at vari ous stages including those useful in collecting together information for deposition of macromolecular structures data harvesting There is also a graphical user interface to many of the programs The CCP4 manual 2 2 Program overview 7 2 2 Program overview 2 2 1 Data processing Programs mentioned below which are marked with an asterisk are not part of the CCP4 suite although this classification is not always well defined but are listed here for completeness Some of these are what we term aggregated to the suite distributed with CCP4 and some are completely s
29. from the system type are controlled by command line flags to configure of the form flag value where flag may be abbreviated uniquely and may be replaced by whitespace The currently implemented flags of this form are are sredir The CCP4 source directory referred to as CCP4 in 811 3 7 Those familiar with GNU software will recognise the influence behind this configure but beware that it isn t quite like the GNU configure scripts The CCP4 manual 70 Installation bindir Where the program binaries get installed libdir Where the library files get installed Other common options are controlled by flags of the form with feature or disable feat with f2c Use the free 2c compiler rather than a native one with shared lib Build a shared version of 1ibccp4 and link against it with x Build the X Windows based programs on some platforms disable feature Disable build components Among these options are phaser cctbx clipper pdb_extract configure currently assumes that the shareable plain text library files remain in CCP4 1ib include rather than being installed elsewhere If the build is not done in CCP4 configure creates src and lib directories mirroring those in CCP4 and does the compilations there configure examines the environment variables necessary for running the suite as defined by ccp4 setup and tries to check them as far as reasonable but these are not actually used i
30. g Xy Veh E 2 LF2FZ FOOLY 2X 1 2208 LP 2 Y 1 7 2 2 where each follows the conventions of Hahn 1992 and matches an entry in the file with logical name SYMOP usually CLIBD syminfo 1lib Names are the short form given in Hahn 1992 the subscripts are typed asis and overbar is typed as a leading so that e g PT is typed as P 1 Explicit operators rarely needed may have the x y z triplets separated by spaces or and the coordinate and translation part may be in either order It is also possible to specify the operators in reciprocal space using h k instead of x y z See also 16 5 1 TITLE title The rest of the line or up to 80 characters of it is taken as a title usually for an output file ve The CCP4 manual 18 3 3 Examples 3 3 1 Unix General CCP4 usage bin csh fe sfall HKLOUT test XYZIN Brk xylose brk HKLIN xylose lt lt EOF sfall your keywords here EOF sfall fft hklin test mapout xylose map lt lt fft eof your keywords here fft eof Figure 3 1 C shell script example bin sh sfall HKLOUT HOME test XYZIN Brk xylose brk HKLIN xylose lt lt EOF sfall amp amp your keywords here EOF sfall fft hklin SHOME test mapout xylose map lt lt fft eof your keywords here fft eof Figure 3 2 Bourne shell script example Example scripts to run m
31. imagine a third case when a map is not good enough to show the correct structure then it could clearly show an incorrect structure The fact that an experienced worker can distinguish an interpretable map from an uninterpretable map suggests that it may be possible in some manner to quantify the property of map quality If we the modify the map in such a manner to increase this quantity then it may be possible to convert a uninterpretable map into an interpretable one How are good electron density maps distinguished from poor ones Even without knowing the structure of a particular protein there are some features which can be expected to appear in the correct map such as Flat solvent The solvent in which the protein is crystallised will fill the gaps caused by imperfect packing of the molecules This solvent is usually disordered from cell to cell and so in the diffraction structure the electron density will be constant to a first approximation This is applied to an estimated map through the process of solvent flattening Wang 1985 Leslie 1988 Predictable distribution of density values in the protein Most proteins have fairly similar proportions of atomic types distributed throughout the cell according to known constraints of atomic spacing As a result the distribution of density values in the protein region is fairly similar from protein to protein This infor mation is applied through the process of histogram matching Zhang amp
32. interactively in three dimensional views O provides general interactive graphics and model building etc there are various programs from Uppsala associated with it hk lview and hk1plot can be used to plot precession pictures molscript and ribbon and others can plot ribbon diagrams from coordinates The CCP4 manual Chapter 6 Isomorphous Replacement One pill makes you larger And the other pill makes you small And the ones that Mother gives you don t do anything at all Grace Slick White Rabbit 6 1 MIR structure solutions steps and programs 1 Find heavy atom sites a Solving Patterson for first derivative to give major site s NAS Nat deriv mtz Calc diff gt CAD gt SCALEIT FFT aa Deriv Fs mtz Patterson map Put all available FHSCAL derivatives together scale Fnat and with anomalous Fderivs and assess quality differences ve PEAKMAX VECTORS HAVECS e NPO generates expected Checking Patterson to plot interesting vectors from coordinates sections b Use Direct Methods to find site s Use Direct Methods to find site s SHELX MULTAN PATSEE 2 a Heavy atom refinement MLPHARE VECREF b Heavy atom phasing MLPHARE Maximum Likelihood PHAsing and REfinement 3 Difference Fouriers to find new heavy atom sites in first and other derivatives FFT PEAKMAX NPO calculate difference maps 4 Back to calculate and check Pattersons for other derivatives
33. number of files Otherwise use mt zut ils to merge sorted files To combine low and high resolution data use scala with different run numbers for the different resolutions See the scala document for more hints To merge unscaled data with a set of previously scaled data produce a multi record file from the scaled MTZ file using combat with input type MTZF and go to step 4 Your files may have different columns if you are using say mos 1m and combat output together The CCP4 manual Chapter 5 Manipulating and Displaying Data Files Graham Knuth and Patashnik Concrete Mathematics More information on data manipulation and display can be found in 2 2 5 1 hkl manipulation Reflexion datasets are stored as MTZ files which are binary These contain hkl Fs os phases etc in a labelled column format See also 16 1 Utility programs are available to produce and handle these files f2mtz reads a formatted file and outputs an MTZ file mtz2various exports reflexion files to formats for other systems mtzdump dumps part or all of any MTZ file as plain text You will always get the header information 1 e column names types and range Plus number of zero observations etc cad Collect Assorted Data is the best program for combining several data sets to gether It reads up to 9 input files with up to 20 assigned columns in each assigns c
34. of Primitive op erations Integer Lattice Type Character 1 Space Group Number In teger Space Group Name Character 10 Point Group Name Charac ter 10 SYMM Symmetry operations in International Tables style RESO Minimum smallest number and maximum largest number resolution stored as 1 d Real 2 VALM Value with which Missing Number Flag is represented COL Column Label Character 30 Column Type Character 1 for each col umn Minimum and Maximum value in each column Real ID of corre sponding dataset Integer NDIF Number of datasets represented in the file PROJECT ID of dataset Integer Project Name Character 64 Normally one for each structure determination DATASET ID of dataset Integer Dataset Name Character 64 May be sev eral for each structure determination e g for different crystals or deriva tives DCELL Cell dimensions for a particular dataset Real 6 DWAVEL Wavelength for a particular dataset useful e g for MAD Real BATCH Batch Serial Number for each batch present Integer This line is not present if number of batches is 0 END of main header card Up to 30 Character 80 lines containing history information For multi record files Batch title Character 70 and optionally orientation data for each batch present in the file End of all headers record N b Column Types are an extra check that the user input assignment for a requested program label is of the correct typ
35. of reciprocal space to another Other special func tions allow for the generation of input data for expansion of the data to a lower symmetry if required and for the generation of data for input to rsearch cif2mtz program to convert mmCIF structure factors e g from PDB to MTZ Martyn Winn coordconv Interconverts various coordinate formats 2mtz Converts free formatted reflection files to MTZ format freerflag Tags each reflection in an MTZ file with a flag for cross validation hklplot Plots precession pictures from reflexion files hklview is better than hklplot if you have X Windows Aggregated matthews_coef Misha Isupov s Jiffy to calculate Matthews coefficient mt zMADmod Generates F F or F D from other for anomalous data mt zdump List header and reflections to terminal or printer Unix script mt zdmp runs it more simply mtzmnf Identify missing data entries in an MTZ file and replace with missing number flag e g NaN mtztona4 Converts MTZ files to portable NA4 ASCII format For exchange with another machine mtzutils Edit columns title or labels combine two reflection files See also cad mtz2various Produce a file in suitable form for SHELX X PLOR CNS etc or in mmCIF or a user defined format na4tomtz Inverse of mt ztonal phistats Performs analysis of agreement between phase sets and checking against weighting factors reindex Reindexes MTZ files when you realise something is wrong Also c
36. s Jiffy to calculate Matthews coefficient lphare 8 9 26 27 30 33 35 37 49 50 Phase calculation and refinement Otwinowski 1991 olrep 8 automated program for molecular replacement osflm 7 Program for processing image plate and CCD data stamp Unsupported Add machine stamp to MTZ or map files lacking it tz2various 12 25 55 Makes ASCII reflection files from MTZ file tzdmp 12 Script for simple running of mt zdmp tzdump 12 25 91 Displays the contents of the MTZ file tzMADmod 12 Generates F F or F D from other for anomalous data tzmnf 12 91 Identify missing data entries in an MTZ file and replace with missing number flag e g NaN mtztona4 12 20 87 Transforms MTZ file to transferable NA4 file mtzutils 8 12 24 25 87 Reflection data file utility program na4tomtz 12 20 Transforms NA4 file to MTZ one ncont 11 Computes various types of contacts in protein structures ncsmask 10 Performs operations on non crystallographic symmetry masks e g before dm npo 11 12 25 27 29 30 Plot molecules end electron density maps oasis 8 Program for breaking phase ambiguity in OAS or SIR omit 10 Calculate omit maps according to Bhat procedure overlapmap 10 11 25 Calculate the overlap of two maps p842asc 19 Unsupported Converts PLOT84 type meta file to ASCII pdb extract 13 Programs for extracting harvest information from program log files pdbcur 11 Curation and manipulation tool for coordina
37. same in both How heavy should the heavy atom be to cause useful changes in the diffraction pattern Perutz Green et al 1954 first pointed out that although one might doubt that the presence of a couple of Hg atoms would produce significant intensity changes in the diffraction pattern generated by 5000 C N and O atoms many of the scatter ing contributions of the light atoms cancel out by interference while 80 Hg electrons scatter in phase The CCP4 manual 6 4 Determination of heavy atom positions 31 In a more rigorous treatment Crick and Magdoff 1956 compared the average change in intensity produced by adding extra atoms to the crystal with the average change in intensity produced by small shifts of the protein molecules and slight changes in the unit cell dimension They showed that the r m s fractional change in intensity Z produced by the stoichiometric binding of Ny heavy atoms of scattering power fy to a protein of Np atoms having a mean scattering power fp is AP 1 Na No 2 fia fo 6 1 This analysis is useful not only to give a clue about the size of the heavy atom required but also to check how flexible the definition of isomorphism should be Their conclusion was that small shifts either of the molecule or of the cell dimensions may produce changes in the intensity sufficient to interfere seriously with the changes due to the heavy atoms at high values of 1 d They also proposed some methods to identify lack o
38. stating at least the pro gram name and the date on which it was run Documentation This should preferably be in HTML format Plain text versions are produced at Daresbury using the Lynx browser for viewing with editors or as Unix man pages Output It is helpful to instrument the output for x loggraph see its documentation 14 3 Non portable features There are a number of things to be wary of if your Fortran code is to be portable and thus acceptable to CCP4 in particular Uninitialised variables Don t assume you can use variables that haven t been as signed to or initialised in DATA Static allocation assumptions Don t assume that you can avoid SAVE where it is re quired by the Fortran standard e g for local variables maintained between sub routine calls DATA which are altered COMMON not declared in MAIN varying length COMMON blocks Assumptions about storage layout Don t assume any more about storage than the standard dictates you may not assume anything about the internal representation of values of different types or detailed argument passing conventions any bit twiddling must be done with library routines Assumptions about arithmetic accuracy You can t assume either IEEE standard arith metic or VAX type arithmetic use double precision if necessary Otherwise you need to observe the items indexed under restriction in the standard ANSI 1978 such as on association of entities
39. t want the aggregated software just CCP4 use tar xvf device ccp4 11 3 3 Environment variables and ccp4 setup The suite depends on some environment variables being set up appropriately This is usually done by the script CCP4 include ccp4 setup where CCP4 is CCP4_MASTER ccp4 Copy the supplied ccp4 setup di st in the include directory to ccp4 setup and edit it appropriately for your needs according to the instructions in the comments Note particularly the remarks in it about setting up for X Windows programs this is highly system dependent With ccp4 setup edited to your satisfaction execute it by source CCP4 include ccp4 setup to define the variables in the environment Each user wishing to use CCP4 should have a similar line in their Login file where CCP4 is replaced by the appropriate absolute path with any necessary qualification for shared filesystems Where the implies that this is the value of an environment variable ccp4 setup and the examples below mostly assume your shell is csh or tcsh where there is a distinction from sh and derivatives like ksh and bash but if you don t use a csh like shell you should have no trouble in adapting the instructions There is however accp4 setup bash version of ccp4 setup which bash users can use as a basis for a version to be sourced in bash profile The CCP4 manual 11 3 4 Configuration 11 3 4 Configuration 11 3 5 Building 11 3 6 Testing 67
40. terms lower then 10 A should be omitted The CCP4 manual 42 Molecular Replacement 7 3 3 Patterson integration radius 7 3 4 Model unit cell size As we have previously said the Patterson function is a collection of interatomic vec tors some will be vectors between atoms within the same molecule i e self vectors and some will be vectors between neighbouring molecules i e cross vectors In the rotation function we try to include the self vectors and to eliminate the cross vectors as much as possible We start by placing the model molecule in a large cell with no symmetry space group Pl If the smallest intermolecular distance is larger than the largest dimension of the molecule all the cross vectors are eliminated as the limit of the Patterson synthesis does not normally exceed the dimension of the molecule For data it is impossible to eliminate all the cross vectors as the molecules in a real crystal lattice are closely packed but the optimal radius of summation can be estimated This depends a lot on the shape of the molecule For a spherical molecule the optimal radius of summation would be the diameter of the molecule as it would contain all the self vectors For a very elongated molecule the choice is difficult The longest self vector will be equal to the length of the molecule but at that radius there will also be many cross vectors In such a case one would have to test various shorter radii but in general there seems to
41. took The CCP4 manual 7 4 2 Translation functions 45 4 3 minutes on a microVax 3 ona 1 A grid The R factor search for the same problem but using a coarser grid 2 A took 34 hours CPU time On a 1 grid it would have taken 275 hours Clearly in such a case the method of choice should be the Translation Function 7 4 2 Translation functions 7 4 2 1 The 7 function The translation function is conceptually related to the rotation function Here we in vestigate the correlation between the observed intensities and the cross vectors between the symmetry related molecules of the model as it is moved about the unit cell When the model is positioned correctly the function should have peaks at values correspond ing to the translation vectors between the symmetry related molecules There are sev eral forms of the translation function to choose from but the T function is normally used A brief description of the main options follows A more detailed description of the various translation functions is given in Tickle 1985 Tickle 1992 The T function was first proposed by Crowther and Blow 1967 It is derived from the general form of the translation function T t Y Loos h F h t 7 6 h where F h is the calculated intensity and t is a vector defining the position of the test molecule The calculated intensity can be expressed in terms of the molecular transforms of the individual symmetry related molecules Thus for a simpl
42. with anisotropic temperature factor see also fhscal icoefl Scaling F to two or more Fes for bulk solvent etc rstats Least squares scaling between Fo and Fy fhscal Scales native to derivative data 2 2 4 Obtaining ab initio phases 2 2 4 1 Heavy atoms MIR MAD SAD See also scaling by fhscal and scaleit above abs Determines the absolute configuration hand of the heavy atom substructure vecsum Patterson search peak search vectors Generates all Patterson vectors from a list of input atoms and produces a list of all vectors which fall within the volume of the Patterson calculated havecs Generates Patterson vectors from fractional sites writing them as orthogonal coordinates in Brookhaven format mlphare Calculates phases and refines heavy atom parameters for refinement of parameters see also vecref See also SHARP oasis Program for breaking phase ambiguity in OAS or SIR from Quan Hao professs Determination of NCS operators from heavy atoms Kevin Cowtan findncs Detect NCS operations automatically from heavy atom sites vecref Refines heavy atom parameters in vector space See also mlphare sapi use direct methods program to find heavy atom sites sigmaa Phase combination of isomorphous and calculated phases calculation of Sim weight etc Read 1986 MULTAN Direct methods package Can be used for finding heavy atom sites from the Patterson map SHELX 97 Direct methods package Can be used for finding
43. 2 Comm nen 11 6 The full story on configure 2 m nme 12 Support bug reports etc 12 1 Release poncy 4 3 hae ei a a ee a A 12 2 Support and bug reports o 12 2 1 Reporting bugs oo onen 13 Resources 13 1 CEPA Web Pages e 3 2 ro a n E ea E 13 2 Anonymous ftp ss sf a6 Sek ee OS 13 3 Electronic mailing lists 2 2 oo on 13 3 1 The CCP4 Bulletin Board o 13 3 2 The Bulletin Board Subscriber Service 13 3 3 The developers list 2 2 2 2 oo o o 13 34 Summary u ann ee nr en ns 13 4 Other crystallographic software 2 2 2 o The CCP4 manual 53 53 54 54 54 54 55 55 56 56 56 56 59 59 59 60 60 61 61 63 65 65 65 66 66 66 66 67 67 67 68 68 68 69 69 69 69 73 73 73 74 vi 13 5 Crystallographic information discussion 13 6 Sources of general free software 13 6 1 Free compilers IV Hackers bit 14 Writing and contributing programs 14 1 General advice 14 2 Requirements for CCP4 code 14 3 Non portable features 14 4 Stealing code 14 5 Using the pre processor in programs 14 6 Library modules 15 Porting CCP4 15 1 Portability features 15 2 Fundamental operating system dependencies 15 3 Unix flavour dependencies 16 File formats 16 1 Reflexions MTZ
44. 2000 RY80M203 Rahway New Jersey USA 07065 lt paula_fitzgerald merck com gt molscript Per Kraulis lt per kraulis sto pharmacia se gt mosflm ftp ftp mrc lmb cam ac uk pub O RAVE Alwyn Jones lt alwyn xray bmc uu se gt See also http kaktus kemi aau dk PDB information http www rcsb org http www ebi ac uk msd PHASES W Furey VA Medical Center and University of Pittsburgh lt 300531 vm2 cis pitt edu gt Photon Factory software e g weis ftp pfweis kek jp Protein W Steigemann lt steigema biochem mpg de gt raster3D ftp stanzi bchem washington edu The CCP4 manual 13 5 Crystallographic information discussion 77 SHELX PATSEE George Sheldrick Institut f r Anorganische Chemie Universit t G ttingen Tammannstr 4 D 37077 G ttingen Germany fax 49 551 393373 lt gsheldr shelx uni ac gwdg de gt sftools ftp mycroft mmid ualberta ca Turbo Frodo http afmb cnrs mrs fr TURBOFRODO turbo html X PLOR Axel Briinger lt xplor laplace csb yale edu gt Xtal Syd Hall lt sydt crystal uwa edu au gt XtalView lt ccms help sdsc edu gt 13 5 Crystallographic information discussion Bionet provides a number of discussion information forums by both e mail and network news equivalently The bionet xtallography group discusses protein crystallography there is also a network news group sci techniques xtallography The jobs group might be of interest to some For m
45. 4 is currently maintained for various flavours of Unix and for windows It should be straightforward to port it to anything else with a reasonably tolerant FORTRAN77 compiler C preferably ANSI C and POSIX 1 bindings for C subject to the problems noted in 15 2 A POSIX 2 environment will probably allow you to use the Unix scripts There should be nothing fundamentally preventing it running under other op erating systems but there are undoubtedly would be some practical problems doing this 15 1 Portability features There are several abstractions used to promote portability and implemented in the li brary ccplib diskio and the routines they call These include e Command line specification of file connexion by associating logical names with file names Calls hiding file opening specifics rather than using Fortran OPEN one of the least portable Fortran features e Binary seekable stream files e Bit twiddling routines for packing unpacking for instance bytes to from words e Access to some system services such as timing 15 2 Fundamental operating system dependencies Should you wish to undertake a port to a completely different operating system to find operating system dependencies which should be confined to the basic library mod ules the first thing to do would be to look for occurrences of the VAXVMS call in the Fortran code This will mainly concern filing system things Some possibilities for assump
46. 5 sketcher is part of ccp4i and provides a graphical interface to libcheck arp_waters version 5 0 of the Automated Refinement Procedure Lamzin amp Wil son 1992 See http www embl hamburg de ARP for the most recent version of arp_warp restrain tlsanl restrained geometry rigid body use of amplitude and phase ob servations group anisotropic displacement parameters disordered solvent cor rections Driessen et al 1989 rdent Create dictionary entries for rest rain from PDB file CNS X ray and molecular dynamics refinement also for rigid body TNT Ten Eyck and Tronrud s refinement package also for rigid body SHELX 97 Least squares refinement 2 2 9 Coordinate analysis act Coordinate checking angles Calculates angles and bond lengths Ramachandran plot anisoanl Analyses of anisotropic displacement parameters areaimol Finds solvent accessible area or area differences of atoms in a Brookhaven coordinate file or files The CCP4 manual 2 2 10 Pictorial presentation of results 11 baverage Average B values for main and side chain atoms Very useful program which gives average r m s Bs for main and side chain atoms Much simpler alternative to Branden real space R factor Can be used to truncate Bs to exclude wildly too small or too high values See also overlapmap bplot Plots B factors by residue cavenv Calculates cavities in macromolecular structures compar Compares two or three sets of atomic
47. 52 fax gnu gnu org To search for a particular program in the sites registered in the system get hold of the invaluable archie program Usenet users might see also the FAQ posting for the Usenet news group comp sources wanted and postings to news answers There is free literature on the network too A case in point is http www Germany EU net books carroll alice html If your Unix system has an unbundled C compiler i e one doesn t come with the system you may be able to use GNU gcc and the GNU C libraries if you can get binaries or bootstrap them on another system If you want support there are several companies willing to sell it to you see the distribution CCP4 will compile with a suitably recent version of the free f2c compiler so you don t need a Fortran compiler either f 2c s home is ftp netlib att com netlib f2c but it s available elsewhere The latest 2as in freedom not necessarily price The CCP4 manual 78 Resources version of the GNU Fortran compiler g77 will compile CCP4 with some effort and remaining problems should be ironed out soon The CCP4 manual Part IV Hackers bit Chapter 14 Writing and contributing programs Tt s my own invention Through the Looking Glass CCP4 is a collaborative project and such collaboration is intended to encompass pro tein crystallographers anywhere You re encouraged to write new software or adapt the CCP4 programs and donate the r
48. ANSI 1978 American National Standard Programming Language FORTRAN Amer ican National Standards Institute Inc Baker D Bystroff C Fletterick R J amp Agard D A 1994 Acta Cryst D49 Blow D M 1958 Proc Roy Soc A247 302 Blow D M amp Crick F H C 1959 Acta Cryst 12 794 802 Blow D M amp Rossmann M G 1961 The Single Isomorphous Replacement Method Acta Cryst 14 1195 1202 Blundell T L amp Johnson L N 1976 Protein Crystallography London Academic Press Bricogne G 1974 Geometric Sources of Redundancy in Intensity Data and Their Use for Phase Determination Acta Cryst A30 395 405 Bricogne G 1991 A maximum likelihood theory of heavy atom parameter refinement in the isomorphous replacement method In Wolf et al 1991 Br nger A T 1992 The Free R value a Novel Statistical Quantity for Assessing the Accuracy of Crystal Structures Nature 355 472 474 Castellano E E Oliva G amp Navaza J Fast Rigid Body Refinement for Molecular Replacement Techniques Collaborative Computational Project Number 4 1994 The CCP4 Suite Programs for Protein Crystallography Acta Cryst D50 760 763 Cowtan K D amp Main P 1993 Improvement of Macromolecular Electron Density Maps by the Simultaneous Application of Real and Reciprocal Space Constraints Acta Cryst D49 148 157 Crick F H C amp Magdoff B S 1956 The Theory of the Method of Isomorphous R
49. DB file Xorth Xfrac Y Yirac Scalei x eae gt 5 1 Zorth Zfrac 1 where Scale Scalejz Scale 3 Scale a Scalei Scales Scalezz Scalez3 Scaleog 5 2 Scale3 Scalezz Scalez3 Scalez4 the matrix at the head of a PDB file i e Xfrac Scale11Xorih Scale 2Yorm Scale13 Zorn Scale 4 Vtrac Scale21Xori Scalez2Yorth Scale23Zor n Scalez4 5 3 Zfrac Scale31Xorth Scalez2Yorin Scale33Zorth Scale34 Therefore extending Scalei to ScaleiExt so that the 4 x 4 inverse matrix can be gener ated Scale Scale 2 Scale z Scale 4 Scaler Scale Scalez3 Scale SealeiExt Scale3 Scalez Scale33 Scalez4 6 0 00 0 00 0 00 1 00 Xfrac Xorth Yoh ScaleiExt x frac 5 5 Zfrac Zorth 1 The extension line 0 00 0 00 0 00 1 00 is necessary to cope with Scale 4 Scalez4 Scaleza the trans lation component of the transformation The CCP4 manual 5 4 Plotting 5 4 Plotting 27 The programs coordconv vectors and havecs will all convert various for mats of fractional coordinates to orthogonal ones havecs s PHARE input type cor responds to mlphare s output coordinate format mapslicer can display sections from map files interactively npo can plot map and coordinate files e g for inspecting map sections It produces a PLOT84 metafile which can be displayed with xplot84driver under X Windows or plotted with pltdev astexviewer can display maps and coordinates
50. Main 1990 Histogram matching employs the predictable atomic makeup of biological macro molecules to predict the histogram of density values in the protein region The current density map can then be systematically modified to bring it into consis tency with the predicted histogram This technique is complementary to solvent flattening since solvent flattening operates on the whole of the solvent region and histogram matching operates on the whole of the protein region The known 47 48 Phase Improvement density histogram is a weaker constraint on the electron density than solvent flat ness but the volume involved is usually larger and histogram matching has more power for phase extension than solvent flattening The combination of solvent flattening and histogram matching usually converges in 10 20 cycles Connectivity of electron density Proteins consist of linearly connected chains of pep tide units A good map will show this linear connectivity and it is this feature which provides the starting point for building a model of the structure The con nectivity of the map can be enhanced by a process of iterative skeletonisation as developed in the PRISM package Baker et al 1994 Known properties of the structure Some features of the structure may be known from analysis of the X ray data or from biochemical sources These include the presence of non crystallographic symmetries relating different parts of the map or relatio
51. Maps 16 2 Maps 16 3 Coordinates 91 MTZ files and take appropriate action In particular when displaying MTZ files using the program mt zdump or the script CETC mt zdmp missing data can be identified and are subsequently represented in the output in an unambiguous manner Currently missing data are given the value 1 230 which should be representable on all boxes but which will overflow the format such that the output field is filled with asterisks All programs will now output NaNs or Rops where appropriate Where such values occur in an input MTZ file they will be carried through to the output Alternatively NaNs or Rops may be generated when for some reason no value can be calculated for a particular reflection and column A new program mt zmnf has been provided to convert old style MTZ files to the new convention This program relies on being able to identify missing data and to this end the following cases are checked a If SIGF 0 0 then SIGF and the corresponding F and D SIGD if present are replaced by NaN or Rop b If SIGI 0 0 then SIGI and the corresponding I are replaced by NaN or Rop c If SIGD 0 0 and the reflection is acentric then SIGD and the corresponding D are replaced by NaN or Rop d If the weight WT 0 0 then PHIB WT and the corresponding Hendrickson Lattman coefficients A B C D if present are replaced by NaN or Rop e If the calculated structure factor FC 0 0 then FC and PHIC are
52. Specifies cell dimensions in and optionally angles in degrees which default to 90 END Terminate input and start the program going End of file will usually have the same effect GO RUN Are sometimes provided in potentially interactive programs with the same effect as END HEADER NONE BRIEF HIST FULL Controls printing of MTZ information as NONE no header output BRIEF brief header output HIST brief with MTZ history FULL full header output from MTZ reads HISTORY history Adds a record to the history stack in an MTZ file pushing older ones off the bottom if there are too many LABIN program label file label Associates the column labels that the program expects with column labels in the input MTZ file If there is no ambiguity the program and file labels can be swapped on the other side of the LABOUT program label file label Associates column labels in the output file with labels used by the program sim ilarly to LABIN ESOLUTION limit limit Specifies resolution limits If only a single limit is given it is an upper limit otherwise the upper and lower limits can be in either order They are in A unless both are lt 1 0 in which case they are in units of 4sin 0 ee SYMMETRY number name operators Specifies symmetry in terms of e g P2 2 2 either number spacegroup number e g 19 name spacegroup name e g P212121 operators explicit symmetry operators e
53. T84 graphics files 93 Field Column Fortran A a paan ri 1 Record ID e g atom het atm Atom serial number residues in order beginning with amino terminus Blank Chemical symbol right justified e g N Remoteness indicator e g D Branch descriptor e g 1 Alternative location indicator if any Standard 3 letter amino acid code for residue Blank Chain identifier Residue sequence number Code for insertion of residue if any Blank Atom s x coordinate A Atom s y coordinate A Atom s z coordinate A Occupancy value for atom B temperature factor Blank Footnote number Figure 16 3 PDB file coordinate record format 16 4 PLoT84 graphics files CCP4 currently uses a homegrown graphics metafile format called PLOT84 imple mented by the routines in p1ot841ib Its format might be relevant if you wanted to write a driver for a different graphics device in which case see existing programs like pltdev as an example or consult the documentation in doc plot84 doc 16 5 Library data files Various data files used by the library routines and specific programs live in the CLIBD directory font84 ascii data for creating the binary PLOT84 font file in CCP4_LIB with fontpack see the PLOT84 code and documentation atomsf lib atomsf_neutron lib X ray and neutron formfactors for every known atom for sfall See documentation code for the programs mentioned for further information 16 5 1 Symmetry operato
54. The CCP4 Suite Computer programs for protein crystallography Overview and manual Edition of February 2006 also available as part of the CCP4 distribution Copyright 1993 2006 The CCP4 Project CCLRC Daresbury Laboratory Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copy ing provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one Permission is granted to copy and distribute translations of this manual into another language under the above condi tions for modified versions except that this permission notice may be stated in a translation approved by CCP4 You can get extra printed copies of this document from Daresbury at cost The Secretary to CCP4 CCLRC Daresbury Laboratory Warrington WA4 4AD UK Telephone 44 1925 603929 direct line Facsimile 44 1925 603825 e mail ccp4 ccp4 ac uk Contents The CCP4 manual Preface I Introduction 1 Introduction I 1 1 CCP4 constitution 1 2 Philosophy of the CCP4 suite o 1 3 Referencing CCP4 etc Tutorial material Overview Steps in structure solution 2 2 Coon 2 1 2 2 Program overview 2 2 1 Data pro
55. a bones output file to to PDB file for ncsmask extends extend Fourier maps and the compute standard uncertainty of electron den sity fffear A package which searches for molecular fragments in poor quality electron density maps O General model building manipulation display graphics program Jones et al 1991 OOPS Facilitates process of rebuilding protein structure inside electron density Kleywegt amp Jones 1996b mama2ccp4 Convert RAVE MAMA format masks to CCP4 format map2fs Convert a CCP4 map to Xtal View fsfour format mapmask map mask extend program maprot map skewing interpolating rotating and averaging program mapsig can do arithmetic on two maps omit calculate omit maps according to Bhat procedure ncsmask Performs operations on non crystallographic symmetry masks e g before dm RAVE and associated programs see above PyMOL Python enhanced molecular graphics program see http pymol sourceforge net overlapmap Map summation averaging and subtraction real space correlation co efficients and R factors peakmax Pick peaks on map e g for searching for water 2 2 8 Refinement of protein models refmac5 Refine or idealize structures using intensity or amplitude based least squares or loglikelihood residuals Murshudov et al 1997 libcheck Generates and manages the library files which provide complete chem ical and geometric descriptions of residues and ligands used in refinement by refmac
56. al line see fig 3 2 In this case the metachar acter isn t available use HOME the terminating word for the lt lt redirection should not be quoted if the introducing one is and it is convenient to chain programs contin gent on exit status with amp amp or to use set e at the top of the scrip in modern sh implementations An example of using sh and substituting inside the here is document is the sortmt z example in examples unix runnable 3Such substitutions can be useful in re usable scripts To turn them off enclose the delimiters in quotes This will avoid nasty surprises with shell metacharacters if you don t want the facility In csh it seems not to be possible to use amp amp with multiple here is documents in this way The CCP4 manual 3 4 Output 19 Which Unix shell you use at least for scripts is a somewhat religious issue but authorities recommend sh for various reasons If you are of the csh faith at least get the improved implementation t csh and use it for writing scripts See the FAQ for Usenet group comp unix shell Unix gotcha beware of calling an executable script the same name as the program it s running Depending on your command path you may end up running the script recursively until you run out of processes or memory A good case can be made for avoiding in your path and invoking executables in the current directory with a leading 3 4 Output The log file
57. am see 83 4 This input is usually keyworded see 83 2 2 This mechanism works uniformly under Unix but only if certain system dependent definitions have been made usually by executing a command file as part of the login procedure see 811 Users of CCP4 Microsoft Windows should use the programs via the graphical user interface ccp 41 which hides many of these details see 83 8 3 2 1 Command line arguments file connexion Input and output data files are specified by associating the with logical names This ab straction was influenced by VMS with the command line processing following more the Unix model Logical names are alphanumeric possibly also including underscores They are associated with file names of a form appropriate for the operating system but including a name part and possibly an extension type directory specification and ver sion The association is usually made by specifying name value pairs on the command line following the program name Thus the format of a program invocation is program name logical name file name gt where Y indicates place holders and indicates optional repeated items i e zero or more here and elsewhere in the documentation may also be used to describe syntax and indicates alternatives between the items on each side A concrete example is fft hklin native Fs mtz mapout 2Fo Fc map The notation name refers to the value of a Unix environment variable
58. amore ecalc may be used to produce normalised structure factors Es which may give better results also for the translation function In the CCP4 implementation this has diverged from Navaza s The CCP4 manual 39 40 7 2 Molecular Replacement amore uses the the externally or internally calculated Fs or preferably Es calcu lated externally and selected radii and resolution ranges to evaluate the rotation function This is described by the Eulerian angles a B y all symmetry equiv alents are listed NB the rotation function in almn is inferior to that in amore amore CNS MERLOT calculate the cross rotation function Translation function pdbset rotates the coordinates of the model structure and outputs a new set of coor dinates sfall takes the resulting coordinates and calculates structure factors for this model in the new cell without applying the symmetry operators cad collects Fopss and F s for all symmetry equivalents tffc calculates translation function Fourier coefficients A phased translation func tion may also be calculated using fft rsearch may also be used in particular cases where you know the solution must be somewhere special but this method is computationally inferior fft then calculates the map amore s translation function may be run for many possible RF solutions at once but is somewhat inferior to t f fcs at present mapsig prints TF map statistics Checking the results Intr
59. and solvent flattening are carried out until convergence is reached The relevant software is summarised in 2 2 6 Having satisfied ourselves that we have obtained the best possible phases we should now attempt to interpret the electron density map Tracing the polypeptide chain of a new structure and the subsequent readjustments are very demanding on both good judgement and time It is therefore very important that the graphics software we use is convenient to use and reliable It is summarised in 82 2 7 2 2 10 Next follows a process of refinement If our starting model comes from molecular replacement it is likely to contain large systematic errors in the coordinates A prelim inary step of rigid body refinement should be carried out to reduce the largest errors in coordinates In this technique the model is divided into a number of constrained rigid bodies The number of parameters refined here is relatively small so one can refine using low resolution data and thus increase the radius of convergence The classical restrained least squares refinement has a relatively small radius of convergence This is because it uses a linear approximation to a strongly non linear problem Discrepancies are minimised between the observed structure factors and those calculated from the model The positions of the atoms are shifted while maintaining geometric constraints on the known bond lengths angles and Van der Waals contacts Many cycles of re
60. and the axis defined as 0 but it is not depen dent on the misorientation angles Typically U will be a permutation matrix The CCP4 manual 16 1 2 Standard column names and types 89 4 the mounting axis as defined for oscgen 1 2 3 for h k D 5 Six misorientation angles rotation axis mis setting six real numbers 6 Oy Ozi 0x 0y 02 The first three refer to the orientation at the beginning of the rotation range at Ostart the second three angles if present refer to to the end of the rotation range at Dena The second set should be set equal to the first set if no crystal slippage is assumed 6 Start and stop rotation angles and range of values 3 real numbers Ogtart Dena Orange 7 Crystal mosaicity in degrees A 8A A correlated dispersion 9 Scale and temperature factors for this batch from scala etc The correction to be applied is oo scaleexp 2B sin A 7 16 1 16 1 2 Standard column names and types Standard names are normally used for the items as in fig 16 1 The column types are as follows index h k 1 intensity structure amplitude F anomalous difference standard deviation of anything J F D or other structure amplitude associated with one member of an hkl h k pair F or FG standard deviation of a column of type G intensity associated with one member of an hkl h k 1 pair I or IG standard deviation of a column of type K phase angle in degrees
61. any programs are available in subdirectories of CEXAM Those in CEXAM unix runnable can actually be run using the distributed data in CEXAM toxd and CEXAM rnase This is not a discussion of the details of any particular program so only the script framework is given the keywords are omitted The programs are usually run from scripts command procedures with the keyworded command data read inline on the standard input stream rather than coming from separate files but can be run interac tively of course One such script com file often runs several programs in sequence Programs can also be run using the graphical user interface ccp4i see 3 8 En tries in ccp4i correspond to program keywords Tutorials using the inteface are dis tributed in SCEXAM tutorial The line bin csh fe in the C shell example in fig 3 1 ensures the script is executed by the C shell without reading chsrc and exiting if a command fails The input is re directed from the here is document comprising the lines between the delimiters EOF sfall Between these delimiters the shell substitutes environment variables preceded by and various other things see csh 1 The exit status of the first program is tested and the second not executed if it fails The terminating avoids problems if there isn t a trailing newline in the file There are some differences if you use the Bourne shell sh 1 or a derivative like ksh or bash apart from the initi
62. as CCP4 but is properly The CCP4 Suite CCP4 is overseen by two committees Working Group 1 largely comprising heads of groups normally meets annually and makes general policy Working Group 2 meets more frequently and coordinates the developers The current chairmen of WG1 and WG2 are Jim Naismith University of St Andrews and Phil Evans MRC LMB re spectively Working Group 2 has a representative of industrial users currently Tadeusz Skarzynski of GlaxoWellcome The various CCPs are coordinated by a steering com mittee The core activities of the CCP4 project are run by a group at Daresbury Labora tory which currently February 2006 comprises Martyn Winn Peter Briggs Charles Ballard Francois Remacle Norman Stein Daniel Rolfe Ronan Keegan and Maeri Howard CCP4 also fund a number of developers in the UK through the BBSRC grant and receipts from industry 1 2 Philosophy of the CCP4 suite The CCP4 manual Unlike many other packages particularly for small molecule crystallography the CCP4 suite is a set of separate programs which communicate via standard data files rather than all operations being integrated into one huge program This has some disadvan tages in that it is less easy for programs to make decisions about what operation to do next though it is seldom a problem in practice and that the programs are less con sistent with each other although much work has recently been done to improve this The advan
63. assumes that the file was written in the native format of the machine with a warning If you have say MTZ files written with an old pre release 2 2 version of CCP4 you can easily insert the 5The similarity with relational database tables is intentional Tf you need to transport them use p842asc and asc2p84 The CCP4 manual 20 General CCP4 usage program starts read environ def file read default def file check extensions check existence of in files parse command line arguments check extensions check existence of in files execute program body Figure 3 3 Description of CCP4 Program startup machine stamp by running mt ztona4 and na4tomt z back to back using the current na4tomtz Similarly with map files and maptona4 3 7 Details of logical name processing etc The gory details in this section can be skipped by casual users They are relevant if you want to redefine or add to the default logical names for instance When a CCP4 program starts up it goes through a pre processing stage before the main body of the program is run see fig 3 3 It starts by reading in a file called environ def in the include sub directory pointed to by SCINCL unless you use the e switch see below This file specifies the logical names that can be expected and some information about the type of file associated with each logical name A typical line may be SYMOP in 1ib which says that the l
64. atom vectors The structure factors of the native protein Fp the derivative Fpy and the heavy atom structure Fy are related by the vector equation Fy Fpp Fp 6 3 We can only get the magnitudes of Fpy Fpp and of Fp Fp from experiment As the phase information is unavailable at this stage it is impossible to use the equation above to calculate Fy Therefore we must estimate it using the procedures defined below For centric reflexions Fpy and Fp are collinear This means Fy must equal either Fpy Fp or Fpy Fp In general the contribution of the heavy atom is small with respect to the structure factor of the protein and the phases of Fpy and Fp are very likely to be the same Therefore Fy should be equal to Fpy Fp In a few cases usually with very weak reflexions the sign of Fpy can be different from the sign of Fp and Fy should be evaluated by Fpy Fp crossing over For acentric reflexions the phases of Fpy and Fp will not in general be correlated and therefore the exact relationship would be given by Fso Fpu Fpl Op ApH Fycos Opy 01 2Fpsin 6 4 and since usually Fy is rather small compared to Fp and Fpp we can assume Op Opy and sin op Op gt 0 It follows that Fpu Fp Fh cos ApH On 6 5 This term is usually adequate for calculating an isomorphous difference Patterson see below Anomalous scattering data can also be used to estimate Fy The app
65. atterson However if the measurements are reasonably accurate and the heavy atom is well substituted both the isomorphous and anomalous difference Pattersons should be interpretable 6 4 3 2 Solving the Patterson 6 4 3 3 Direct methods Even an error free Patterson generated from several sites can be hard to interpret es pecially in a high symmetry space group Automated search procedures for locating heavy atoms in difference Patterson maps have also been developed Terwilliger amp Eisenberg 1987 see also rsps and vecsum A systematic search procedure is es pecially necessary in the presence of high non crystallographic symmetry as in crys talline viruses Rossmann er al 1986 These automatic procedures can fail when two vectors that are close overlap to form one big peak Thus the heavy atom may be placed incorrectly at a special position Solution by hand may be the only way of avoiding this It is good practice to find consistency between Pattersons calculated from different resolution ranges rather than to rely on just one Difference Pattersons can determine whether your solution is correct see below The location of a larger number of heavy atoms by difference Patterson becomes in creasingly difficult and the use of direct methods can help in solving the problem Standard direct method programs for example SHELX86 Sheldrick 1991 or MULTAN Germain et al 1970 can sometimes find heavy atom sites consistent with the i
66. ave been successful with structures of up to about 150 atoms However they have sometimes been used to find positions of the heavy atoms in some difficult cases of heavy atom derivatives The programs are summarised in 82 2 4 3 A basic crystallographic tool is the Fourier transform Using it Patterson maps are calculated from structure factor amplitudes and electron density maps are calculated from structure factor amplitudes and phases transformations from reciprocal space to real space Quite often one also has to calculate structure factors from coordinates via an Foaie Map real space to reciprocal Most of the relevant programs use the fast Fourier transform FFT algorithm They are summarised in 2 2 5 One may attempt to improve the phases by density modification A variety of methods can be used to modify the electron density map which are summarised in 2 2 6 For structures where there is a non crystallographic symmetry we can also use molecular averaging This is a powerful means of improving phases especially in cases of high non crystallographic symmetry but if carefully used in combination with solvent flattening even a two fold averaging can lead to a substantial improvement of the map One first has to determine the symmetry transformations for averaging either from heavy atom positions or from molecular replacement Then a molecular envelope is determined either automatically or from an existing model Cycles of averaging
67. bbard R eds 1989 Molecular Simulation and Protein Crystallography CCP4 Daresbury Study Weekend nos DL SCI R27 ISSN 0144 5677 Warrington WA4 4AD UK Daresbury Laboratory for Dares bury Laboratory Green D W Ingram V M amp Perutz M E 1954 Proc Roy Soc A 225 287 Hahn Theo ed 1992 International Tables for Crystallography third edn Vol A Kluwer Hao Quan amp Woolfson M M 1989 Application of the Ps function Method to Macromolecular Structure Determination Acta Cryst A45 794 797 Harada Y Lifchitz A Berthou J amp Jolles P 1981 A Translation Function Com bining Packing and Diffraction Information An Application to Lysozyme High Temperature Form Acta Cryst A37 398 406 Harker D 1956 Acta Cryst 9 1 Hayward S amp Berendsen H J C 1998 Systematic Analysis of Domain Motions in Proteins from Conformational Change New Results on Citrate Synthase and T4 Lysozyme Proteins Structure Function and Genetics 30 144 Helliwell J R Machin P A amp Papiz M Z eds 1987 Computational As pects of Protein Crystal Data Analysis CCP4 Daresbury Study Weekend nos DL SCI R25 ISSN 0144 5677 Warrington WA4 4AD UK Daresbury Labora tory for Daresbury Laboratory Hendrickson W A 1991 Determination of Macromolecular Structures from Anoma lous Diffraction of Synchrotron Radiation Science 254 51 58 Hendrickson W A amp Lattman E E 1970 Representa
68. be readable It s derived from the literate program check out comp prog literate in the file library nw which can be used directly if you have the noweb tools mentioned in the code It s done that way to be more maintainable at Daresbury 85 86 Porting CCP4 the target into configure and try to run it and use the configuration to make test at the top level When you ve got that working all should be OK try it out e g on the data in the examples toxd directory It s worthwhile contacting the CCP4 staff before starting any porting work for your benefit and ours The CCP4 manual Chapter 16 File formats The information below about the format of reflexion map and coordinate files is not normally required since libraries are provided for handling files in these formats In formation is stored in the headers of MTZ and map files to allow transparent reading on architectures other than that on which a file was written see 3 6 For details of the na4 ASCII exchange format for mtz and map files see the documentation and code for the program mt ztona4 The obsolete LCF format for reflexion files is only relevant for converting old datasets see the 1cflib code The file formats are binary We don t subscribe to the school of thought which advocates text only files in this case of large amounts of floating point data although they are a good idea in other cases There are two basic reasons e Clear text files
69. ble to remove the installed bina ries from the previous one with make uninstall In particular this will remove obsolete binaries 6You can get patch from the usual archives of free Unix software it s distributed by the GNU project for instance or from the Daresbury info server if necessary see 13 The CCP4 manual 11 3 10 Installation summary 11 4 CCP4I Here s a summary of the simplest procedure for installation again depending on the platform this may be performed automatically cd CCP4 gunzip if necessary tar x edit include ccp4 setup configure system make testlib not normally called for make optional step set path CCP4 src Spath runsome tests and re set path op tional make install e make clean or realclean or empty targets optional to tidy up From version 4 0 the suite includes a graphical user interface called ccp4i The files for this reside in sub directory CCP4 ccp4i Necessary environment variables are set in the main setup file ccp4 setup ccp4i exists as a set of tcl scripts and requires no compilation Once the setup file has been sourced it can be run simply by typing ecp4i ccp4i does however require a local installation of Tcl Tk with the blt extension Details on this and other installation issues can be found at http www ccp4 ac uk ccp4i_main html 11 5 X windows programs If you wish to use the X Windows based programs such as xloggraph you wi
70. blicise or announce meetings etc or to ask the user community for help It is not intended for bug reports which should be sent to ccp4 dl ac uk To send messages to and receive messages from the Bulletin Bulletin board is a historical misnomer this is just an electronic mailing list 75 76 Resources Board you need to subscribe see below The address to send messages to the Bul letin Board is ccp4bb dl ac uk and the system re broadcasts these to the list of subscribers 13 3 2 The Bulletin Board Subscriber Service In order for you to use the Bulletin Board you must first inform the system that you wish to subscribe to the list subscriptions are handled by an automatic system In its simplest form send the message SUBSCRIBE CCP4BBto Majordomo dl ac uk Note take care with the address to manipulate your access to the list a common error is to send the message to ccp4bb dl ac uk which results in the user com munity getting the message and no action taken by the system Messages to Ma jordomo dl ac uk have blank lines ignored and all commands are case insensitive Your e mail address will be worked out by the server but can be changed with the RETURN PATH command To remove yourself from the list send UNSUBSCRIBE CCP4BBtoMajordomo dl ac uk Ifnecessary just send HELP for guidance 13 3 3 The developers list 13 3 4 Summary As well as the CCP4BB list there is a list for people who consider themse
71. by playing around with the solvent multiplication factor and the truncation level see the keywords TRUNC and SLVMUL If the crystals contain a lot of solvent 70 to 80 you might try phase extension but make sure the map actually improves by doing so If you have got very weak phase information at higher resolution try including it from the start but keep the RADIUS at the lower resolution If you have got reasonable phase information to 3 7A and very weak information to 3 2A use a radius of 3 7A and a high resolution cutoff of 3 2A don t extend to 3 2 but use all information from the first cycle Identify and refine non crystallographic symmetry if present The Uppsala pro grammes O mama and imp are ideally suited for this purpose You must remove overlap between symmetry related masks The mask can be either in O or CCP4 format e The mask will encompasses a monomer Each mask has a set of associated symmetry operators which descibe how the density within this mask is related to other density within the asymmetric unit This is a bit different from the way things are done in the rave package and has some advantages include the non crystallographic restraints in Solomon and run the script again You might need to reduce the level of truncation a bit and the solvent multiplier should probaly be a bit less negative 8 4 Estimating solvent content Estimating solvent content Th
72. case the correctly oriented model is used to generate a new set of structure factor magnitudes and phases The agreement between the observed and model magni tudes is then used to generate weights for the model phases typically using sigmaa s PARTIAL option The calculated phase and resultant weight are then input into the density modification calculation The use of sensible defaults means that the program can be operated very simply In the case of a non crystallographic symmetry averaging calculation it is necessary to generate a mask covering one non crystallographic subunit usually one molecule and in addition the non crystallographic symmetry matrices must be included in the command file The most common application of the program involves use of the histogram match ing and solvent flattening methods These are very quick lt 1 minute for 10 cycles with a unit cell volume of 10 A on a typical workstation and in the majority of cases lead to a good map improvement In the case of very poor starting maps the histogram matching can lead to a very broken map in which case it may help to apply a temper ature factor to smooth the map or else repeat the calculation with solvent flattening alone Averaging is always useful if the symmetry elements and mask can be deter mined Sayre s equation is usually only helpful when the starting map is very good and its application is very slow Other features available in dm include input
73. ce of approach is that prolsq uses an FFT for structure factor and derivative calculations whereas rest rain uses slow FT s This means that prolsq takes many cycles 50 to converge but each cycle is very fast whereas restrain takes only a few cycles 5 but each one is much slower Normally prolsq has the advantage here restrain s functionally is much like prolsq s with some small differences e prolsq treats all main chain peptide residues as though they had identical geom etry recent data indicates that glycine and proline are different from the others restrain treats them differently e restrain uses individual distance constraint weights based on the estimates of the standard deviations of Engh amp Huber 1991 these values are all in the dictionary together with the ideal distances prolsq uses blanket values for the weights because its dictionary doesn t contain the s d s restrain has a coupled occupancy refinement option for disordered sidechains restrain has a full matrix option for estimating individual positional standard deviations However it requires a lot of memory and at present needs to be compiled with a re parameterised include file e prolsq has the option of applying non bonded intermolecular repulsion restraints i e between symmetry related molecules as well as the intramolecular ones At present rest rain only applies the intramolecular repulsions e There is a subtle difference in the way th
74. cessing 2 2 2 Data scaling and reduction 0 0 0 00004 2 2 3 Data combination and scaling different sets 2 2 4 Obtaining ab initio phases 2 2 5 Map and structure factor calculation 2 2 6 Molecular averaging and phase improvement 2 2 7 Map interpretation and manipulation 2 2 8 Refinement of protein models 2 2 9 Coordinate analysis o 2 2 10 Pictorial presentation of results 2 2 11 Utility programs 2 2 12 Data Harvesting Utilities 2 2 13 Graphical user interface onen 2 2 14 Other systems General CCP4 usage 3 1 3 2 3 3 3 4 3 5 3 6 3 7 3 8 Program documentation Running the programs 3 2 1 Command line arguments file connexion 3 2 2 Keyworded input 2 2 2 moon Examples 3 3 1 Unix Output 2 2 2 Rudiments of reflexion files Transporting data files Details of logical name processing etc o o 3 7 1 Command line Graphical User Interface Data processing and reduction 4 1 42 Data processing Data reduction vii OMANI YANN 95 NR Rh a e e e e WwWWwWNnNnrF OOO oO 15 15 15 15 16 18 18 19 19 19 20 21 21 5 Manipulating and Displaying Data Files 5 1 5 2 5 3 5 4 hkl manipulation 2 Coon Map file utilities a
75. cession pictures 27 program initialisation 81 program options 21 program output 19 program startup 20 program termination 82 python 65 R factor search 6 Ramachandran plot 10 12 56 references 2 refinement 5 Hendrickson Konnert 53 least squares 53 reflexion files 19 relational database 19 87 release policy 73 Remacle Francois reporting bugs 74 ribbon diagrams 11 12 27 Richard Stallman 74 rigid body fitting 39 The CCP4 manual 111 rigid body refinement 6 40 Rolfe Daniel 1 Rop 89 rotation function 39 43 Rypniewski Wojtek vii Sayre s equation 9 48 49 104 scaling 5 7 8 23 scratch files 21 self rotation function 5 self vectors 42 shared libraries 68 shell login 66 shells Unix 19 Sim weight 8 Skarzynski Tadeusz 1 skeletonisation 9 10 48 104 small molecule drawing 12 solvent content 50 51 solvent flattening 6 9 47 49 sorting 12 25 spacegroups specifying 17 standard input 15 standard output 15 standards C 81 Fortran 81 82 POSIX 81 stealing code 83 Steele Guy L 113 Stein Norman 1 stereochemistry 11 subroutine library support 73 symmetry equivalents 23 symmetry operators 17 syntax description 15 Tcl Tk scripting language 21 Tektronix 12 temperature factor 8 termination program 82 Terry Howard 87 Tickle Ian vii TLS refinement 53 54 Toolpack 83 tools Fortran 83 translation function 6 40
76. ct sites Note that mlphare does not implement a correct approach to maximum likeli hood estimation which should involve a straightforward optimisation of the likeli hood function after integrating out the unknown native F and phase Bricogne 1991 Read 1991 This optimisation is carried out at least as a function of the resolution dependent lack of closure non isomorphism parameters and possibly also as a func tion of the heavy atom parameters although vector space refinement see below may be a better way of determining those if anomalous data are not involved The CCP4 manual 36 Isomorphous Replacement This refinement can be quite slow and if there are sufficient centric data it is sen sible to refine against them first When there are two independent derivatives it is pos sible to use the sign of the anomalous occupancy to decide the hand of the solution If the anomalous occupancies are negative the heavy atom sites are very likely to lie at x y z Another possibility is that you have confused the intensity observations and indexed the h k I as h k 1 The final and essential check is to calculate difference Fouriers or residual maps with coefficients Y AF so Fg Je 6 15 h These may show some peaks corresponding to new heavy atom sites 6 5 2 Vector space refinement 6 5 3 Fy refinement Vector space refinement as implemented in vecref has advantages ove
77. d odd are tests of successive least significant bits of n If we now rearrange the data into the bit reversed order the transform can be constructed by combining adjacent pairs of points to get two point transforms then the adjacent pairs of pairs are combined to get four point transforms and so on until the first and the second half of the data is combined into the final transform If you have a mathematical or functional programming bent and want a deep un derstanding of the algorithm consult Jones 1989 for a calculation of it Let s take a simple example of an 8 point transform The coefficients of Fs of the successive divisions of data are e even o odd ee eo oe 00 eee eeo eoe eoo oee oeo ooe 000 In binary this becomes 000 001 010 O11 100 101 110 111 and with the bits reversed 000 100 010 110 001 101 O11 111 which in decimal corresponds to elements numbered 04261537 This means that we first have to combine element 0 with 4 adjacent pairs on the above list 2 with 6 1 with 5 and 3 with 7 to get two point transforms N b each element is normally a complex number Next we combine the resulting elements 0 with 2 1 with 3 4 with 6 and 5 with 7 to get 4 point transforms Finally we combine the resulting elements 0 with 1 4 with 5 2 with 3 and 6 with 7 to get the complete 8 point transform Done Let s now see how fast the FFT is compared with the traditional brute force approach First the data have to b
78. d sighing Alice s Adventures in Wonderland Continuous updates of the suite are no longer made Instead there are intermittent releases of well defined versions Patches for serious bugs in the current version are made available by ftp in the prerelease directory and advertised on the ccp4bb mail list see 13 3 and the Problems Page on the web see 13 1 New major ver sions are also announced on ccp4bb of course and possibly by posts to appropriate network news groups Update notifications to non commercial users are only made electronically and in the occasional CCP4 printed newsletter Releases may have ver sion numbers of the form n m p where n m is the major version and p indicates the number of the patch to that version but at the time of writing this system has had to be dropped this is a bug and releases just have major version labels A change of n in the version is intended to be used for substantial changes e g when the file formats change in an incompatible way The sources distributed by ftp and tape always correspond to the latest version but patch files of differences between different minor versions are available by ftp for the convenience of users wishing to upgrade from an earlier minor version 12 2 Support and bug reports The CCP4 manual Remember non trivial programs generally contain bugs You may or may not feel that CCP4 has more than similar systems of the same size but you do have th
79. defined in the standard startup scripts In Unix it is substituted using the leading The procheck documentation is a PostScript file in the manua 1 directory as well as a plain text doc file The CCP4 manual 15 16 General CCP4 usage Here an FFT is performed on the MTZ reflexion dataset associated with the logi cal name hklin filename native Fs mtz and a map is output to logical name mapout file 2Fo Fc map Logical names are case insensitive remember that the file names are case sensitive in Unix though Each standard logical name has a default extension associated with it which is appended is one isn t specified Thus the mtz and map extensions in the above example are redundant since they are the defaults for the logical names in question Unix pitfall Beware of trying to default the extension for file names which already contain a dot since the part of the name after the dot will be taken as an explicit extension This could arise if for instance you were generating filenames with a resolution included as a real number perhaps in some parameterised script There is a library of standard logical names with associated default extensions and access modes read write or read write Names from this set which are appropriate to each program are detailed in the programs writeups Some will default to reading a library file e g of symmetry operators or writing a file in the scratch directory and usually are not s
80. diation where it is possible to select the wavelength to optimise the anomalous con tribution It is possible to obtain the hand of the heavy atoms from one anomalous derivative at one wavelength Woolfon amp Yao 1994 6 6 3 Multi wavelength Anomalous scattering Data The values of f and f are wavelength dependent Thus if data are collected at sev eral wavelengths there will be small isomorphous changes between the Fhg s and the anomalous differences will also have different magnitudes The advantage of this is that only one derivative needs to used and the data sets are all isomorphous if they are from the same crystal The disadvantages are that one needs a tunable X ray source and the experiments are more complex The increased complexity is due to the fact that the effective signals are even smaller than normally present in MIR or SIRAS The signal and thus the wavelengths need to be optimised This is achieved by using wave lengths around an absorption edge for the heavy atom in question See Hendrickson 1991 for more details on experimental procedures These differences can be used to determine the protein phase either using madsys or by maximum likelihood phasing mlphare the latter is probably the method of choice Although there is no native ina MAD experiment mlphare requires an effective native to be defined This can lead to negative occupancies To avoid this use the data set with the largest negative va
81. ds A standard reflexion data file contains the following items in the order given Not necessarily all items have to be present e MTZ identification record containing the machine stamp encoding the number formats of the architecture it was written on as a full word at byte 8 from the start of the file e Reflexion data columns of data held as REAL 4 e A beginning of header record e A set of keyworded records containing VERS Version stamp Character 10 currently MTZ V1 1 TITLE File Title short identification of file Character 70 NCOL number of columns number of reflexions in file number of batches Integer If the number of batches gt 0 this indicates a multi record file CELL Cell Parameters Real 6 SORT Sort order of 1st 5 columns in file Integer 5 MTZ s Terribly Zany but Means Tilde Zero Mc Claughlin Terry and Zelinka were its progenitors This may bring to mind tables or relations in relational databases intentionally so The model for an MTZ file is two relations one the header keyed on keywords such as SYMMETRY the other comprising the reflexions keyed on the H K and L attributes columns This formalism has been largely forgotten but clarifies some things particularly operations in mt zut ils or a replacement which should implement some of relational calculus 87 88 16 1 1 Orientation data File formats SYMINF Number of Symmetry operations Integer Number
82. e banner that the programs print on startup and if there might be some programming or compiler problem what operating system you are using There is no orthodox standard for submitting effective bug reports but some infor mation that can typically be useful as well as mistakes to avoid is given in this advice adapted from Cygnus Support In general common sense assuming such an animal exists dictates the kind of in formation that would be most helpful in tracking down and resolving problems in soft ware e Include anything you would want to know if you were looking at the report from the other end There s no need to include every minute detail about your envi ronment users have been known to provide the values of every variable in their environments although anything that might be different from someone else s environment should be included your path for instance Often we will want to see the complete output log file from the run e Narratives are often useful given a certain degree of restraint If a person re sponsible for a bug can see that A was executed and then B and then C knowing that sequence of events might trigger the finding of an intermediate step that was missing or an extra step that might have changed the environment enough to cause a visible problem Again restraint is always in order I set the build running went to get a cup of coffee Columbian cream but no sugar talked to
83. e Normally the Miller indices will be in the first three columns though in the defi nition of the format there is no restriction on the use of the columns of the reflexion data records However the subroutines which output the MTZ header information in a formatted way presume that the first 3 columns of a standard MTZ file are the Miller Indices and the first 5 columns of a multi record MTZ file are H K L M ISYM and Batch number The orientation block contains values of the orientation parameters constraint flags for cell parameters and the relative scale and temperature factor to put this batch on the same scale as the reference intensity 1 Six cell dimensions ngstr m and degrees real numbers These define an orthogonalization matrix B to convert indices hkl to Cartesian crystal axes xyz x Bh a b cosy c cos B B 0 b siny c sinB cosa 0 0 A c This gives x parallel to a and y in the a b plane Six cell dimension flags integer defining which parameters may be varied 1 parameter is freely variable 0 parameter is fixed i e cell angle 90 or 120 i gt 0 parameter is constrained to have the same value as parameter number i Orthonormal matrix U 9 real numbers in order U11 U12 U13 U2 etc to ro tate the crystal Cartesian coordinates to the camera Cartesian coordinates The idxref matrix is A UB The matrix U defines a standard orientation and is a function of the mounting axis
84. e A AAA es 7 1 2 Cross rotation function oaoa aa 7 1 3 Translation function aoaaa aaae 7 1 4 Checking the results ooa a Mtroduct ON se el a Ren ah The rotation function 2 200 002 e eee eee TIA Model os Sh gS Dee dk eg 7 3 2 Resolution limits 2 22 CC Como 7 3 3 Patterson integration radius 2 2 2 onen 7 3 4 Model unit cell Size o 7 3 5 Omitting weak terms o 73 6 Normalising e EST Ongm Removals ce os A A oe 7 3 8 Non crystallographic symmetry The translation search 2 2 Co ee ee 741 Rfactorsearch o e 7 4 2 Translation functions o non Phase Improvement Density Modification 2 2 2 2 o o e e e 8 1 1 The Problem ia 6 lt 2 222 8 Sr 28 8864 8 1 2 The Method s s 2 20 28 E ae ei Using dm 2 Bisher GAR Be A ae ee Using Solomon esca ee nr ss Estimating solvent content o e The CCP4 manual 25 25 25 25 26 27 29 29 30 30 31 32 32 33 35 35 36 36 36 36 36 37 37 37 38 38 39 39 39 39 40 40 40 40 41 41 42 42 42 42 43 43 44 44 45 47 47 47 47 49 50 50 9 Refinement and validation 9 1 Least squares structure refinement 9 1 1 Comparison of restrainandprolsQ 9 2 Maximum Likelihood refinement 9 2 1 TLS refinement in refmac 5
85. e CCP4 manual 46 Molecular Replacement This rests on the approximation that the search model and the structure have the same self vector set It should be pointed out that correct scaling is critical for the self vector subtraction to work The complete formula for the T1 function as it is usually defined can now be writ ten T1 t Y Loros MB Fm Fine ht 7 12 h where Leross h are the observed intensities with the the self vectors subtracted Fy is the Fourier transform of one of the molecules in the crystal Fr is the complex conjugate Fourier transform of a second molecule h is the reciprocal vector hkl t is the translation vector between molecule M1 and molecule M2 The function 71 defined in this way is an ordinary Fourier summation and can be evaluated by a fast Fourier algorithm 7 4 2 3 T2 function Whereas the T1 function has to be calculated for every pair of symmetry related molecules the 72 function combines all the symmetry It has the form T t Y Loros 1 x 2 Fui 7 13 h i j lt i The 72 function applies phase shifts so that all the peaks pile up at the same place defined by a single translation vector t for the reference molecule so the signal noise is much better than for a single T1 function This is implemented by t f fc 7 4 2 4 The TO O function This is an attempt to generalise the translation search by combining a 3 dimensional translation function with a packing function Harada et al 1981
86. e as Matthews formula Vp 1 66v Vm 1 66 Ay Na 1 Dp is Matthew s v 0 74cc g Alternatively Vp NpAv N where N is the number of protein atoms in the unit cell including hydrogens There are about the same number of hydrogens as C N O etc If Vp is the fraction of protein volume in the asymmetric unit the density PpVp Ps 1 Vp 1 35V gt 1 0 1 Vp 0 35V 1 0 8 3 p where p is the solvent density 1 0 for water Therefore Vp p 1 0 0 35 8 4 If you know the density you can work backwards and find the number of molecules in the asymmetric unit exactly axissearch will tell you the volume of the unit cell in A Now you get this useful information e Possible molecular weight per asymmetric unit e Fraction of protein solvent from 8 4 e If you know the number of atoms in your protein subunit roughly 14 times the number of residues including hydrogens and saying roughly again that the volume per atom is 10 A it is possible to get an estimate of number of subunits asymmetric unit Even if you do not know the density accurately it is often possible to guess how many subunits there might be per asymmetric unit Often there is only one possible solution for Nyubunits The CCP4 manual Chapter 9 Refinement and validation I don t believe there s an atom of meaning in it Alice s Adventures in Wonderland Refinement has been covered in three Study Weekends Machin et al
87. e contributed both to the software and documentation see the program writeups but Eleanor Dodson Phil Evans and Ian Tickle deserve special mention The quotations by or relating to Lewis Carroll have some significance because Carroll Dodgson was born at Daresbury administrative base of CCP4 Also the original editor was never previously far from either Looking Glass House or Dodgson s college vil Part I Introduction Chapter 1 Introduction Where do you come from said the Red Queen And where are you going Through the Looking Glass 1 1 CCP4 constitution The CCP4 program suite is a collection of disparate programs covering most of the computations required for macromolecular crystallography They have been collected and developed under the auspices of the Collaborative Computing Project Number 4 in Protein Crystallography supported by the UK Science and Engineering Research Council SERC since 1979 and currently the Biotechnology and Biological Sciences Research Council BBSRC and coordinated at Daresbury Laboratory The Project aimed to support collaboration between those working on such software in the UK and to assemble a comprehensive collection of it to satisfy the computational require ments of the relevant UK groups The results of this effort are also made available for distribution to academic and commercial users outside the UK The distribution described herein is often loosely referred to
88. e environ def file Graphical User Interface As an alternative to running the programs via scripts as outlined above a graphical user interface ccp4i is also available The interface is written in the scripting language Tcl Tk and will run on any platform which supports an interpreter for Tcl Tk This includes both Unix and Microsoft Windows platforms The main function of the first version of the interface is to make running the pro grams easier by hiding many of the details of logical names and keyworded scripts that are outlined above It also presents options for running the programs in clearer ways than just bare keywords which makes it particularly accessible for inexperienced users At the same time since the programs are independent of the interface more ex perienced users can jump in and out of ccp4i to run one or two programs without becoming locked in to the system Within ccp4i access to programs is organised by task where a task usually cor responds to one main program but may also use other helper programs For example the interface to f ft may also run the mapmask program to extend the map then use other utilities to convert the map file format to something appropriate for a graphics program Related groups of tasks are further organised into modules corresponding to different parts of the structure solution process such as Experimental Phasing or Molecular Replacement In addition ccp4i also has a simple project ma
89. e example of two molecules in the unit cell related by a two fold axis we get F h Fe M F h Em Fu2 Ein Fin 7 7 where Fy and Fm2 are the molecular transforms of the two molecules sampled at point h in reciprocal space and denotes complex conjugation For a model molecule shifted by x from an original position only the phase of its molecular transform changes and Fy becomes Fye2 t The calculated intensities can now be expressed as follows F h Em h ehti L Fyn h ento Fin h e 2h Fin h e th 7 8 We should keep in mind that the two molecular transforms and their putative posi tions t and tz are related by symmetry We can now substitute the above expression into the translation function T t Ios h Fan Fo Fy Fyn Fi Fo ero y 7 9 h where t now equals t t2 We are only interested in terms determining the cross vectors i e dependent on the relative positions of molecules so we can omit the first two F terms which represent the self vectors for the two molecules For simplicity one of the cross terms is also omitted The result is known as the T function T t Y Los B Fi Fine rt 7 10 h 7 4 2 2 Removing the self vectors the T1 function In an attempt to improve the signal to noise ratio we may try to subtract the coefficients for the self vectors from the the observed intensities Icross Lobs k Fan Fin 7 11 The time was mainly spent in the FFT Th
90. e needed In some cases just one derivative can be enough The best results can be obtained if data from anomalous scattering are available but there are limitations due to the fact that anomalous differences are small and would require the diffraction data to be measured with a degree of accuracy that is often unob tainable SIRAS phases are normally considered the first step in an MIR determination instead of the final result A different approach that utilises a procedure known as noise filtering has been proposed by Wang see 88 Analogous procedures which make use of different phys ical constraints have been used to improve the electron density maps they have been reviewed Podjarny et al 1987 However all these approaches are more useful for cleaning a bad MIR map than actually solving phase ambiguities in SIR phases There are a number of techniques used to break the phase ambiguity of one wavelength anomalous scatterering with some success Ralph amp Woolfson 1991 Fan et al 1990 Hao amp Woolfson 1989 The CCP4 manual Chapter 7 Molecular Replacement He must have imitated someone else s hand said the King The jury all brightened up again Alice s Adventures in Wonderland 7 1 Molecular replacement steps and programs 1 Detection of non crystallographic symmetry using Crowther s FFT rotation function self rotation function Native Fs mtz gt gt AMORE gt Runtrials
91. e planar groups peptide groups plus PHE ARG etc sidechains are treated prolsq restrains to the plane calculated from the coordinates before each refinement cycle whereas rest rain restrains to the current best plane this should allow the planes more flexibility of move ment 53 54 Refinement and validation 9 2 Maximum Likelihood refinement 9 2 1 9 3 9 4 The maximum likelihood approach to model refinement has been implemented in the CCP4 program refmac The refmac program is used for the restrained or unre strained refinement or idealisation of a macromolecular structure It minimises the model parameters to satisfy a Maximum Likelihood residual There are options to use different minimization methods refmac produces an MTZ output file containing sigmaa style coefficients suitable for the calculation of mF DF and 2mF DF maps using fft The latest version of REFMAC is significantly different to earlier versions and is known as refmac5 The key functionalities of refmac5 are e Restraints are calculated within the main program A large dictionary of standard geometries is included Restraints can also be created for novel ligands see in particular the Monomer Library Sketcher in ccp4i which is an interface to the libcheck program e TLS refinement can be used This is particularly useful when there is signif icant anisotropy but the resolution does not warrant refinement of individual anisotropic disp
92. e rearranged into the bit reversed order That takes very little time and requires no extra memory as it only involves swapping pairs of data The transform itself is constructed in log N sweeps through the data i e the number of times all the V data points can be factored three in the above example At each level data points are combined pairwise This means the number of operations in each sweep is proportional to N Thus the whole algorithm is of order Nlog N Such O NlogN asymptotic complexity is typically associated with divide and conquer algorithms like this The transform can be parallelised and vectorised What about the non FFT approach From the expression N 1 Ray N 10 3 20 we see that for every Fp numbering N we cycle over all f s also numbering N Thus the number of operations is proportional to N What is the difference between N and Nlog N For N 1000 the factor is 100 For N 1000000 the factor is 50000 roughly the difference between one minute and one month Moral beware of algorithms with quadratic or higher complexity The CCP4 manual 10 3 Crystallographic FFT 61 10 3 Crystallographic FFT 10 4 Programs So far we have only discussed the case where the number of data points N is a power of two What if N cannot be factored all the way to single points There are other highly optimised algorithms tailored for handling transforms of various lengths In fact the crystallographic FFT does no
93. e source to fix them if necessary Although CCP4 makes no commitment to support for the suite we are keen to improve it and will be grateful for reports of bugs particularly with fixes or gen eral constructive comments regarding functionality usability etc including installa tion and documentation The primary purpose of bug reports is to get problems fixed in a subsequent version for the good of all but we will often be able to furnish fixes or workarounds to avoid your research being stymied Also if you make significant en hancements to the software please pass these on as required by the licence Programs not in the unsupported directory will get the most attention as they have someone specifically assigned to look after them Our ability to maintain different flavours of Unix is limited by their availability to us but we are happy to include compatible changes supplied to us which are necessary for other platforms Unfortunately our VMS expertise and facilities are limited most CCP4 development work is now done in Unix but we do our best We cannot support operating systems other than VMS and Unix if you need to run the suite under another OS consult 15 in the first instance and we will be interested to hear of your experience Please make reports etc to the CCP4 Secretary at Daresbury preferably by elec tronic mail to ccp4 dl ac uk If you don t get a response within a few days try again but please make sure t
94. eavy atom sites for one of the derivatives are determined they can be used to calculate preliminary phases The phases then can be used to solve for heavy atom positions in the other derivatives Next the positions and occupancies of heavy atoms for all the derivatives can be refined simultaneously to give an improved set of phases It is an iterative process during which more heavy atom sites can be found usually with low occupancy and included in the refinement Finally a statistically best set of phases is obtained The relevant programs are summarised in 2 2 4 1 If we already have a reasonable model of the structure we are trying to solve e g from a homologous protein or the same protein in a different crystal lattice we may attempt a method known as molecular replacement As more and more structures are solved this is an increasingly successful method of determining preliminary phases First the molecule s orientation in the crystal unit cell is determined from a cross ro tation function If the molecule has an internal non crystallographic symmetry the orientation of the symmetry elements can be determined from a self rotation function and checked for consistency with the cross rotation The translational parameters are Overview then obtained from an R factor search or from a translation function The relevant programs are summarised in 2 2 4 2 In case of small molecules the phases are usually obtained by direct methods These h
95. ecref 8 30 36 Vector space refinement of the heavy atoms Consider also mlphare vecsum 8 30 34 Unsupported Automated Patterson solution vectors 8 27 29 Generates Patterson vectors from atomic coordinates The CCP4 manual Program index Non CCP4 programs VO watertidy watncs ll watpeak Il wi lume 11 11 26 protein atoms lson 7 23 xccpjiffy2idraw 12 xdldataman 13 xdlmapman 10 25 55 xloggraph 12 19 23 49 66 69 82 X Windows xplot84driver 12 27 29 zeroed BP 3 8 Babel 12 76 CNS 9 10 12 30 40 53 FRODO TOM 76 HySS 9 LAUE 7 AMA 10 25 p R S S S S S T T X X a MERLOT 9 40 MULTAN 8 9 34 VE 9 10 76 ARP 8 9 ELX 97 8 10 ELX86 34 ELX93 53 ELX 12 30 55 77 NT 10 30 53 cl1 Tk 69 PLOR 9 12 26 55 DS 7 23 ngel 9 archie 77 arp_warp 10 54 bash 18 66 bibtex 113 blt 69 bo nes 10 cd 69 chemnotes 26 co nfigure 67 71 corels 76 cp 70 cs h 18 19 66 67 d trek 7 demon 9 76 de nzo 7 23 The CCP4 manual Makes Wilson plot Converts xloggraph xplot84driver output to editable idraw format manipulates reflection files exchanges formats etc manipulates maps exchanges formats etc Display plots derived from log files under Polyhedral volume around selected atoms Rearranges water position Pick waters which follow NCS and sort out to NCS as
96. ed by the use of clusters In fact for a derivative with one heavy metal at Ny sites ar 1 gt Nu Np fi fp while for a complex of Ny atoms in a single site ar 1 Np Nufiu fe The advantage of using groups in contrast to conventional compounds becomes especially evident when the sites of substitutions are to be located in a difference Patterson More information on isomorphous replacement can be found in the Study Weekend proceedings Wolf er al 1991 and Blundell and Johnson 1976 Useful references to the development of the method are Green et al 1954 Harker 1956 Perutz 1956 Crick and Magdoff 1956 Blow 1958 Kendrew et al 1958 and Blow and Crick 1959 6 4 Determination of heavy atom positions In order to use the isomorphous replacement method the amplitude and phase of the heavy atom contribution to the structure factor of the derivative has to be determined This means that the sites of substitution must be worked out from the observed dif ferences in scattering amplitudes caused by the introduction of the heavy atoms The techniques used to do this are e Patterson searches of different kinds e direct methods e difference Fouriers These have to work with some estimate of Fy obtained only from the observed differ ences The CCP4 manual 32 6 4 1 6 4 2 6 4 2 1 6 4 2 2 Acentric reflexions Isomorphous Replacement Scaling between native and derivative data Estimati
97. em specific subroutines other than at the CCP4 library level CHARACTER variables in COMMON with non CHARACTER non standard ordering of declarations DATA must come last FORMAT statements without all items including line termination descrip tors being separated by commas Besides ensuring portability such restrictions allow us to use useful tools to check and maintain the code C is also acceptable for non numerical applications but discouraged for numeri cal ones Code in the library should not depend on ANSI C features since some people still have only K amp R C but for non library code they can be expected to acquire gcc if necessary Try to make C POSIX l compliant Zlotnik 1991 and fix it up for those systems without the relevant features If you have to use Pascal it will need to be a dialect that can be grokked by p2c for translation to C Data formats Programs should use the standard data formats 16 for reflexion co ordinate and map data Initialisation Fortran programs should call the routine CCPF YP initially allowing one to run them using the mechanism program hklin file hklout file The CCP4 manual 81 82 Writing and contributing programs Termination There should be no STOP statements within the code rather call the subroutine CCPERR This has the form CALL CCPERR num string where num is an exit value 0 for normal terminatio
98. ement CCP4 Daresbury Study Weekend Warrington WA4 4AD UK Daresbury Laboratory for Daresbury Laboratory Drenth Jan 1994 Principles of Protein X ray Crystallography Springer Verlag 97 98 Driessen H Haneef M I J Harris G W Howlin B Khan G amp Moss D S 1989 J Appl Cryst 22 510 516 Driessen Huub P C amp Tickle Ian J 1994 The use of normalised amplitudes in the Rotation Function Joint CCP4 and ESF EACBM Newsletter on Protein Crystal lography June Engh R A amp Huber R 1991 Accurate bond and angle parameters for X ray protein structure refinement Acta Cryst A47 392 400 Fan Hai Fu Hao Quan amp Woolfson M M 1990 On the Application of One Wavelength Anomalous Scattering II An Analytical Approach for Phase Deter mination Acta Cryst A46 659 664 Fitzgerald P M D 1988 MERLOT n Integrated Package of Computer Programs for the Determination of Crystal Structures by Molecular Replacement J Appl Cryst 21 273 281 French G S 8 Wilson K S 1978 Acta Cryst A34 517 Germain G Main P amp Woolfson M M 1970 Acta Cryst B26 274 285 Giacovazzo C Monaco H L amp Viterbo B 1992 Fundamentals of Crystallography International Union of Crystallography Texts on Crystallocraphy Oxford Glusker J P amp Trueblood K N 1985 Crystal structure analysis a primer Second edn Oxford University Press Goodfellow J Henrick K amp Hu
99. eparate external packages not distributed with CCP4 The tutorial chapters have more information on the use of these programs See also the program index page 103 and 13 4 for information on obtaining some non CCP4 programs 2 2 1 1 Film image plates CCDs mosflm mosflmisa widely used data processing program for image plate CCD and oscillation data denzo film and image plate processing package Otwinowski 1993 d trek Image processing package LAUE For processing data taken with the Laue method Aggregated ipdisp For viewing and measuring images under X Windows hklview displays zones of reciprocal space as pseudo precession images under X Windows rotgen Simulate X ray diffraction rotation images 2 2 1 2 Other area detectors 2 2 1 3 Diffractometers madnes Package for processing FAST area detector data and data from other de tectors See also CCP4 programs undupl and madlat xentronics Proprietary program for processing Xentronics data XDS Kabsch s program for processing Xentronics data Most machines provide their own software and produce an ASCII file f2mtz or combat will convert this into MTZ format respectively for merged and un merged data 2 2 2 Data scaling and reduction combat is a jiffy for converting various foreign formats of un merged intensity data to multi record MTZ format for input to scala It can easily be extended to accept additional formats scalepack2mtz Converts mer
100. eplacement for Protein Crystals I Acta Cryst 9 901 Crowther R A 1972 The Fast Rotation Function Pages 173 178 of Rossmann M G ed The Molecular Replacement Method Int Sci Rev Ser no 13 New York Gordon and Breach Crowther R A amp Blow D M 1967 A Method of Positioning a Known Molecule in an Unknown Crystal Structure Acta Cryst 23 544 548 Cullis A F H H Muirhead Perutz M F Rossmann M G amp North A C T 1961 Proc Roy Soc A 265 15 Dickerson R E Kendrew J C amp Strandberg B E 1961 The Crystal Structure of Myoglobin Phase Determination to a Resolution of 2 by the Method of Isomorphous Replacement Acta Cryst 14 1188 DOD 1978 Military Standard Fortran DOD Supplement to American Standard X3 9 1978 MIL STD 1793 US Government Printing Office Dodson E J 1976 A Comparison of Different Heavy Atom Refinement Procedures Pages 259 268 of Ahmed F R ed Crystallographic Computing Techniques Copenhagen Munksgaard Dodson E J Evans P R amp French S 1975 Pages 423 436 of Ramaseshan S amp Abramson S C eds Anomalous Scattering Copenhagen Munksgaard Dodson E J Gover S amp Wolf W eds 1992 Molecular Replacement CCP4 Daresbury Study Weekend nos DL SCI R33 ISSN 0144 5677 Warrington WA4 4AD UK Daresbury Laboratory for Daresbury Laboratory Dodson E J Moore M amp Bailey S eds 1996 Macromolecular Refin
101. er since Whenever we do something that involves a calculation of an electron density map or structure factors i e moving between real and reciprocal spaces it is almost certain that the program we use has an FFT algorithm embedded somewhere inside it We can tell if it does not the calculation is orders of magnitude slower than what we have come to expect In addition to the above uses the FFT has an important role in molecular replacement in the calculation of the rotation and translation functions This development is mostly due to Tony Crowther 1972 who showed how the rotation function can be calculated by an FFT if the Patterson function is expanded in spherical harmonics and David Blow Crowther amp Blow 1967 who applied the FFT to translation functions Although it is not necessary to know anything about the mechanism of the FFT in order to use it it might still be interesting especially if the explanation is reasonably understandable A good description of the FFT is given in Press et al 1986 It goes like this If we have a discrete Fourier transform of length N we can split it into two trans forms of length N 2 each One of them is formed from the even numbered terms and the other from the odd numbered terms N 1 Fa Fe HN j 0 N 2 1 N 2 1 Zu L Pj ezrihl 2j Ny gt Pr ezrih 2j 1 N ne 1 N 2 1 pa Pj o2Tijh N 2 W 7 fie pe2tihi N 2 ere ee 10 1 where Wp e2 h N Depending on N we can cont
102. ere s no problem with clobbering a previous CCP4 installation make realclean will delete the directories containing split files used to build the library and you will need these if you have to run a symbolic debugger with library code In this case you may prefer to make clean and then make empty targets the latter will make empty files in place of the compiled programs saving space but preserving the time information that make can use later SIf you can t keep the CCP4 directory around on disc after make empty targets you might back it up to cd with tar to preserve the date information if you subsequently get updates The CCP4 manual 68 Installation 11 3 7 Problems Since the amount of testing of different systems we can do is limited there may be problems with the installation on less widely used systems perhaps due to non standard Fortran code which hasn t yet been eliminated or peculiarities of different versions of Unix First of all check our Web page problems htm1 which has a list of known problems with the current release If there is no fix then please try to solve the problem locally if you can and let CCP4 know about what changes you needed to make if you can send us patches made with diff c electronically that would be most helpful If you successfully use without changes a configuration which reports itself as untested or not properly tested it would be helpful if you could let CCP4 know that it wor
103. ess time flexibility portability and simplicity of use Since the data formats form the core of the suite they are described in more detail in 16 The formats used in the CCP4 suite share many characteristics with those used in other packages The programs are distributed in source form so they can be studied don t look too closely modified and fixed by recipients in contrast to most commercial software In addition executables are distributed for some common platforms Otherwise you will need Fortran C and C compilers to build the suite although there are freely obtainable Fortran and C C implementations see 13 6 There are freely available Unix implementations for PC type boxes at least and the combination of a sufficiently well resourced top end PC and say the Linux operating system and compilers reput edly makes a competent environment for scientific computing There is a policy of continual technical and scientific updates to the suite Where existing programs have been incorporated into the suite they have often subsequently undergone considerable modification above that needed to use the CCP4 file formats 1 3 Referencing CCP4 etc Please see the distribution conditions for the suite available from the CCP4 website at http www ccp4 ac uk ccp4license html In particular please note that any publica tion arising from use of the CCP4 program suite should include a reference to Collab orative Computational P
104. esults to CCP4 for the benefit of the community Below are some guidelines for doing so The current software may not all obey these guidelines for one reason or another but do as we say not as we do If you re doing development work on the code you may want to join the ccp4 dev mailing list 13 3 3 14 1 General advice Please especially if you are writing new code use the library routines where appropri ate Using library routines has various advantages from the point of view of maintain ability portability consistency and possibly efficiency Abstraction is your friend If you see the need for new facilities in the library please let CCP4 know The library routines are imperfectly documented in the same place as the programs if necessary use the source 14 2 Requirements for CCP4 code The following conditions should be satisfied by code to be included in the suite We are aware that not all the code in it currently fulfils these conditions but Language Most code is expected to be written in ANSI standard FORTRAN77 ANSI 1978 although we re prepared to accept some extensions which are in Fortran90 taken from DOD 1978 primarily long variable names lower case block DO INCLUDE and IMPLICIT NONE Don t use e style comments e n qualifications on non character declarations without good reason like a serious need to save space just INTEGER REAL DOUBLE PRECISION COMPLEX or LOGICAL and no BYTE syst
105. evise estimates FM using MAD data where FM is an optimised value of the nor malised anomalous scattering BP3 Multivariate likelihood substructure refinement and phasing of S MIR AS and SAD 2 2 5 Map and structure factor calculation sfall Structure factor calculation using inverse FFT fft Map calculation using fast Fourier algorithm mapmask Extend asymmetric unit of mask or map to cover any grid volume mapdump Print map header and part or all of map 2 2 6 Molecular averaging and phase improvement cpirate Statistical Phase Improvement from Kevin Cowtan dm Density modification using solvent flattening Sayre s equation histogram match ing NCS averaging and iterative skeletonisation dmmulti A multi crystal version of DM solomon Modifies the electron density maps by averaging solvent flipping and pro tein truncation demon angel Fred Vellieux s molecular averaging and solvent flattening package there are versions for both Groningen BIOMOL files and CCP4 MTZ files RAVE The Uppsala averaging package The CCP4 manual 10 Overview 2 2 7 Map interpretation and manipulation ccp4mg The CCP4 molecular graphics program Optimised for the generation of presentation graphics and surfaces COOT The Crystallographic Object Orientated Toolkit for model building completion and validation xdlmapman Manipulates maps bones skeletonisation exchanges formats etc Kley wegt amp Jones 1996a bones2pdb Convert
106. example m 0 866 corresponds to an expected error of 30 in Gp The maximum likelihood approach of assigning each possible phase a likelihood and combining these means that weights are more realistic i e lower and phases more accurate since spurious sites are lost more completely An alternative procedure avoiding the limiting assumption that all the errors lie in the determination of Fpy has been proposed Cullis et al 1961 6 6 6 Hendrickson Lattman coefficients The formulation of the phase probability function given by Blow and Crick has the disadvantage that for every reflexion a summation must be carried out over all the derivatives This means that every time new derivatives or phase information from other sources are available everything has to be re calculated A variation of the Blow and Crick method Hendrickson amp Lattman 1970 represents the phase probability for each derivative by the equation P o ek A cosa B sina C cos20 Dsin 20 6 20 Calculating the probability function using various information is simply a matter of summing the corresponding coefficients 6 6 7 Single Isomorphous Replacement The phase ambiguity can be removed by another isomorphous compound but the need for two or more derivatives has often created difficulties since their preparation is one of the most time consuming steps in protein crystallography Therefore it is of interest to minimise the number of isomorphous crystals which ar
107. eyword value pairs of the form keyword value Such arguments are separated by spaces tabs commas or char acters The detail of the input expected is given in the documentation for each program of course However there are some general rules e Only the first four characters of keywords are significant although you are rec ommended to use complete keywords and they are case insensitive e Records may be continued across line breaks using amp or as the last non blank non comment character on the line to be continued e Text following a non quoted or is treated as a comment and ignored A continuation character may precede the comment e Strings may be single or double quoted but the quotes may be omitted if a string doesn t contain whitespace or a comment or continuation character as above or if the whole of the rest of the record is read as a single string which is the case with title information etc E g TITLE nothing special The CCP4 manual 3 2 2 Keyworded input 17 ITLE don t continue me ITLE not a comment e Leading spaces are ignored e If an input record is of the form filename with possible surrounding whites pace the contents of the file filename will be included as input at that point after which the following records will be read 3 2 2 1 Common keywords Specifications of some keywords common to various programs follow CELLabc a y
108. f isomorphism One could examine the variation of A 1 or alternatively of Fea Fp Fp with resolution if the ratio increases at high 1 d some movements of the protein have taken place The Normal analysis Smith amp Howell 1992 is a good method of assessing a derivative if the standard deviations are reasonable They suggest looking at the distri bution of isomorphous differences against standard deviations for all the observations The central limit theorem predicts that if the Fpn measurements are essentially the same as the Fp measurements 66 of the differences should lie within one standard deviation of the mean If this is not so it is possible that there is a real difference between the two sets of data Serious problems in the search for heavy atom derivatives can arise in the case of crystals with very large asymmetric units From Crick and Magdoff s analysis it is clear that in order to have a detectable difference in intensity when Np increases either Ny or fa must increase Obviously fy is limited by the number of electrons of the heavier stable atoms available It is possible to label the protein at many sites by standard heavy atom compounds but derivatives containing single atoms at multiple sites are usually difficult to solve and the risk of obtaining non isomorphous crystals is increased What is needed is the scattering power of multiple substitutions concentrated at one or a few sites this can be achiev
109. f molecular interfaces sfcheck Program for assessing agreement between atomic model and X ray data sortwater Sorts waters by the protein chain to which they belong in the case of a protein with several equivalent subunits surface Determines accessible surface area topp An automatic topological and atomic comparison program for protein struc tures volume Determines polyhedral volume around selected atoms watertidy Assigns waters to nearest subunit and residue watncs Pick waters which follow NCS and sort out to NCS asymmetric unit watpeak Lists peaks found by peakmax near to atoms 2 2 10 Pictorial presentation of results ccp4mg The CCP4 molecular graphics program Optimised for the generation of presentation graphics and surfaces COOT The Crystallographic Object Orientated Toolkit for model building completion and validation astexviewer Java program for displaying molecular structures and electron den sity maps loggraph Plots graphs of tables from many CCP4 programs log files as part of ccp4i mapslicer Section viewer for CCP4 map files O General model building manipulation display graphics program which can produce hardcopy molscript good representation of molecules often used together with raster3D Kraulis 1991 npo Plots maps and draws structure onto them Various graphical representations The CCP4 manual 12 2 2 11 Utility programs Overview ortep2 A small molecule drawing prog
110. ferences in the text of the form n m refer to chapter n section m etc This document contains material from various sources It is mostly based on the previous manual which was mainly an installation guide and Wojtek Rypniewski s overview man page The previous manual was updated by Dave Love from a ver sion containing contributions from Peter Daly Kim Henrick and Wojtek Wolf Chapter 1 contains material originating from Phil Evans and John Campbell Chapter 6 was originally based on material by Silvia Onesti Some of 8 was taken from material by Andrew Leslie and Kevin Cowtan The description of the MTZ file format was taken from Sandra Mc Laughlin s mt z1 ib documentation and that for maplib from John Campbell and Phil Evans Several of the sections outlining procedures and program usage are taken from lecture notes by Eleanor Dodson and often have accompanying example files in the examples directory derived from that course Wojtek Rypniewski s acknowledgements for what he wrote include I had asked various people for help and comments I would like to thank Ian Tickle for reading very thoroughly the chapter on molecular replacement and making several corrections Fred Antson did the same for molecular averaging Sandra McLaughlin helped me write some guidelines on programming standards Dave Love edited designed and typeset this document Subsequent updates from Martyn Winn and Peter Briggs Many people hav
111. files with extension lib dic bes or prt which do not have a full path name have the directory CLIBD put in front of them so logical name SYMOP will actually be associated with CLIBD syminfo 1lib Thus you only need specify values for logical names such as SYMOP if you want to use other than the library file Furthermore files with extension scr are deemed to be scratch files and are assigned to the scratch directory Files of type in are checked for existence and the program fails with a message if it can t find them Under Unix using output files that already exist will also cause the program to stop unless the environment variable CCP 4_OPEN is set to UNKNOWN in which case they will be overwritten with a warning message The programs can take several command line switches to modify the pre processing that is done though they are rarely useful These must be given before the other argu ments and have the Unix style prefix rather than They are n Do not read the global files default def and environ def v 0 9 Verbose output You can also use h instead of v Higher values give more debug information from the program the default value is 1 Use v 0 if you want to turn off messages about file opening otherwise v is currently only useful for debugging An error occurs if you omit the number d file Specify an alternative to the default def file e file Specify an alternative to th
112. for reading such records as indicated by the third column of fig 16 3 is A6 15 1X A4 A1 A3 1X A1 14 A1 3X 3F8 3 2F6 2 1X 13 The standard setting of the orthogonal axes for the Brookhaven format is xla yle xa ze The suite assumes these settings if the SCALE cards are not present in a coordinate file 3This format will eventually be replaced by the new macromolecular CIF format The CCP4 manual 92 PUNE oOo Du 11 13 14 16 17 18 19 20 21 22 23 24 25 26 34 35 37 38 52 53 54 55 56 57 256 NC NR NS MODE NCstart NRstart NSstart Nx Ny Nz X Y Z a B y MAPC MAPR MAPS Amin Amax Amean ISPG Nsymbt LSKFLG SKWMAT SKWTRN future use MAP MACHST ARMS NLABL LABEL 20 10 File formats number of columns fastest changing in map number of rows number of sections slowest changing in map Data type 0 envelope stored as signed 8 bit bytes in the range 128 127 the convention is that values 40 are true within the mask although masks should normally only comprise values of 0 and 1 1 image stored as 16 bit half words 2 image stored as 32 bit reals 3 transform stored as complex 16 bit integers 4 transform stored as complex 32 bit reals Note Mode 2 is the normal one used in CCP4 programs for maps and mode 0 is used for masks others haven t been tested recently and may not work Number of first column in map Number of first ro
113. ftware is briefly summarised in 2 2 1 Once we have extracted integrated intensities from the images the responsibility of CCP4 is to pick up these intensities and provide tools for further processing and analysis The data are scaled and merged Corrections are applied the data are then scaled together and reduced to the asymmetric unit Finally the intensities are converted to structure factor amplitudes The relevant software is summarised in 2 2 1 From this point solving the structure is nothing other than finding the phases 1 e the other 50 of the information we need but cannot measure directly The fundamental technique for obtaining phases ab initio is the multiple isomor phous replacement MIR It involves introducing strong scatterers heavy atoms into the crystal lattice If as a result the lattice was not significantly damaged the presence of the heavy atom should cause small but measurable changes in intensities Several different heavy atom derivatives are usually needed Data from them are collected and finally the various data sets are scaled together using the programs summarised in 86 The Patterson difference map is calculated for each derivative to see if the heavy atom positions can be determined for any of them This is usually only possible if there are a few of the heavy atoms In most cases this is done manually although programs for automatic peak interpretation or direct methods can also be used Once the major h
114. ged scalepack output into MTZ format detwin Detwins reflection data from merohederally twinned crystals dtrek2mtz Converts d trek scalemerge output into MTZ format dtrek2scala Converts integrated intensity and header files from d trek into multi record MTZ format for input to scala postref postrefinement of film data scala Replaces rotavata and agrovata Scales batches of data from procesed images together in a similar way to Kabsch 1988 Applies scale factors adds together partially recorded reflexions monitors and rejects bad agreements between re peated measurements or symmetry equivalents and averages them for output Various statistics on the averaging are produced truncate Converts from intensities to Fs by the method of French amp Wilson 1978 and does a Wilson plot to estimate B factor unique Generates a unique list of reflexions uniqueify Script for completing a dataset up to a high resolution limit and adding free R flags wilson Makes a Wilson plot rebatch Alters batch numbers in an unmerged MTZ file saves running mosf1lm again Jiffy has been adopted in this community to mean a small utility program c f mainstream jargon use as a unit of time Raymond 1993 The CCP4 manual Overview 2 2 3 Data combination and scaling different sets cad Combines Assorted Data from several MTZ files and resorts changes asymmetric unit etc See also mt zutils scaleit scaling of derivative to native data
115. hange symmetry of residue to asymmetric unit sftools Reflection data file utility program including some density map handling sortmtz Sort and or merge MTZ files stereo Reconstruct 3D coordinates from measurements of stereo diagrams tracer Lattice TRAnsformation CEll Reduction wulff ps For generating Wulff net The CCP4 manual 2 2 12 Data Harvesting Utilities 13 xdldataman Manipulates reflection files exchanges formats etc Kleywegt amp Jones 1996a 2 2 12 Data Harvesting Utilities cif2xml Convert mmCIF data harvesting files into XML format cross_validate Validate harvesting files prior to deposition Data Harvesting Manager A graphical tool to curate data harvesting files this is part of ccp4i pdb_extract A suite of programs from the PDB RCSB for generating harvest files from program logfiles 2 2 13 Graphical user interface 2 2 14 Other systems ccp4i is the CCP4 graphical user interface T hope they give us a plug BIOMOL The BIOMOL package is a set of programs developed in the Groningen protein crystallography group for most crystallographic tasks which is being con verted to use CCP4 file formats PHASES Phasing Protein The PROTEIN program system is an integrated collection of crystallo graphic programs designed for the structure analysis of macromolecules Xtal Is a package for small molecule and macromolecular calculations which is avail able for reasonably large PC systems
116. hat it is possible to reply problems we quite often encounter include invalid return paths From or Reply to fields of your mail header with a bad network address and failure of the local system to deliver mail due to a bad user id or configuration problem The ccp4bb list is not intended for bug reports Any mention of support by us in this document or elsewhere should be understood to be in quotes if not already so to indicate that it isn t guaranteed 73 74 12 2 1 Reporting bugs Support bug reports etc We don t have time to answer queries not directly related to CCP4 software please address questions about running system utilities etc to your local computing support staff Failing that if you have access to it Usenet news is the ultimate fount of wisdom but not necessarily about CCP4 and don t assume the developers read netnews See 13 6 and 13 4 for some information on obtaining non CCP4 software Before reporting a bug please check whether it s been fixed in the latest version if you don t already have the latest The distribution contains a CHANGES file which lists the major changes since the previous release Next check the Problems Page on the web site see 813 1 to see if it has been fixed since the latest release There is also a file called PROBLEMS in the distribution describing long standing known bugs In any case it will usually be useful to know what version you are using look in th
117. he local and crystallographic axes have the same order e g both are two fold and are parallel the combination of the two symmetry elements results in pure translation The Harker plane should have an identifiable peak where the intermolecular vectors coincide and the position of this peak indicates the position of the molecular axis relative to the crystal axis If the two axes are not exactly parallel the vectors do not coincide and the peaks become smeared out 7 3 8 2 Self rotation function Often there is more than one molecule subunit in the asymmetric unit A self rotation function i e a rotation function where the native Patterson is compared with itself can be calculated in such cases to find the transformation that superposes the different molecules It is often difficult to decide whether a given peak is significant The self rotation function has a massive origin peak corresponding to zero rotation and all the other peaks are relatively small It is helpful to calculate the mean and standard deviation for all the values after the origin peak has been removed We can take as a rule of thumb that if a peak exceeds five standard deviations it is significant If it is weaker it may still be significant but it must be treated with skepticism The crucial test for the correctness of a solution is that it must be consistent with the cross rotation function In addition if we are dealing with a symmetric molecule i e a dimer or tetra
118. heavy atom sites using isomorphous differences see http linux uni ac gwdg de SHELX rsps Determines heavy atom positions from derivative difference Patterson maps Can be used interactively to examine the fit of potential sites to the map crossec tabulates anomalous scattering factors f and f SHARP Statistical Heavy Atom Refinement and Phasing program Probably the program of choice for heavy atom refinement and phasing see http babinet globalphasing com BP3 Multivariate likelihood substructure refinement and phasing of S MIR AS and SAD See also mlphare 2 2 4 2 Molecular replacement amore is a complete molecular replacement system in one program from Jorge Navaza molrep automated program for molecular replacement from Alexei Vagin Vaguine et al 1999 phaser MR is crystallographic software for phasing macromolecular crystal struc tures with maximum likelihood techniques This is the molecular replacement module which replaced beast beast brute force molecular replacement with Ensemble Average Statistics Max imum likelihood based molecular repaclement Randy Read Obsoleted by phaser MR bulk applies bulk solvent correction for the translation search and rigid body refine ment steps of AMoRe fsearch performs up to 6 dimensional molecular replacement solution given a pre determined envelope ecalc calculate normalized structure amplitudes Es from Fs chainsaw model preparation The CCP4 manual 2
119. id volume to cover q Weighted map coefs molecule rather than asymmetric unit Figure 9 1 Steps in generating an electron density difference map 9 5 Why is protein refinement difficult Small molecule people manage analysis and refinement with very few problems Macro molecular crystals present several particular problems in refinement e For macromolecular crystals the unit cell is big and there are a very large num ber of X ray data to collect all of which have low signal to noise ratio It is therefore not usually possible to collect data to atomic resolution as is normal for small molecule structures The data available often suffer from both system atic and random errors These are due to the crystal size problems of mounting absorption and crystal decay e Protein crystals have an additional problem There is usually a high solvent content and the crystal forces are weak Some parts of the chain may not be crystalline at all and others may have high thermal motion This means that not all the unit cell can be properly parameterised This is true for almost all pro teins not just those which diffract to lower resolution This problem particularly reduces the intensity of the high resolution data In addition it leads to severe effects of radiation damage e These two problems mean that experimental data extend to limited resolution typically to a maximum limit in the range 3 2 A e This means that the ratio of observat
120. ild or rebuild parts of a model arp_warp updates the model by identifying and removing poorly defined atoms and adding new atoms Rejection of atoms is carried out on the basis of the density interpolated at the atomic centre the deviation of the density shape from sphericity and some distance criteria Addition of atoms is performed on the basis of difference density coupled with distance constraints CCP4 distributes an older version of arp_warp version 5 0 which has been re named as arp_waters and which should only be used for adding waters while cy cling with refmac5 Difference map generation On completing a round of refinement various types of difference map can be generated with the program fft for comparison with the current model e g using the graphics program O refmac produces weighted map coefficients suitable for mF DF and The CCP4 manual 9 5 Why is protein refinement difficult 55 2mF DF maps these coefficients reduce model bias and are recommended over the unweighted Fy F and 2F F maps To obtain weighted coefficients from the output of rest rain the program sigmaa can be used ea RESTRAIN SIGMAA model brk alternative G weighted map coefs refinement A 8 p program Native Fs mtz SS REFMAC gt Nat ive sF mtz gt FFT gt MAPMASK gt XDLMAPMAN new Fc Sim wt and ac calc 2mFo DFc map EXTEND map format exchange Fo scaled to Fc mFo DFc select gr
121. in 0 6 5 1 Maximum Likelihood Phase refinement In Maximum Likelihood phase refinement Otwinowski 1991 we find a probability distribution for values of the protein phases op and then minimise the weighted differ ences between observed and calculated values of Fpy generated with this phase with respect to the heavy atom parameters These are e relative scale and temperature factor between protein and derivative data e coordinates e real and anomalous relative occupancies e thermal factors either isotropic or anisotropic e error estimates which are used when finding the appropriate weight for each observation The parameters are updated and another cycle of refinement is carried out The function to be minimised is 2 Rj LY wp Fats FSS Op 6 14 op h where j refers to the j derivative This procedure refines scale factors relative occupancies and estimated errors for different derivatives It is not seriously biased by incorrect models Since all possible protein phases are sampled for each reflection the procedure rapidly eliminates incor rect sites For reflections with well defined phases the weight of the protein phases will be very small for all values except the best phase But there are many reflections where the protein phase is poorly defined and for these reflections the maximum likelihood approach avoids the old problem of biasing the results towards phases derived from incorre
122. included by the FORTRAN compiler C_LIBS the additional libraries included by the C compiler CXX LIBS the additional libraries included by the C compiler RANLIB dummy on SysV Unix ranlib on BSD depending on whether ar r builds a symbol table itself LNS indicates how to make symbolic links in the filesystem usually In s else cp to copy rather than link In hard links could be used in the absence of symbolic links if you don t need to operate across file systems M4 how to run the m4 macroprocessor and define a symbol to indicate the system type see configure source MAKE the name of the make program for use in recursive Make files if the sys tem s make doesn t define the symbol MAKE Usually null SETFLAGS set appropriately if individual programs need special Fortran compiler flags most likely to suppress optimisation see the configure source INSTALL_PROGRAM a command to install executables Uses install c ifa BSD version is available to avoid interfering with running programs else cp INSTALL_DATA a command to install non executables If BSD install is avail able uses install m 644 else cp The CCP4 manual 11 6 The full story on configure 71 Installing only the library cd CCP4 mkdir iris build cd iris build debugging diagnostics for C some optimisation and debugging for fortran env CC gcc XCFLAGS g Wall FOPTIM gl configure irix src ccpdisk xtal ccp4
123. ing 48 49 Ballard Charles 1 binary files 19 Bionet 77 BMCD 77 Briggs Peter 1 Brookhaven format 25 bugs 73 reporting 74 76 bulk solvent 8 bulk solvent correction 54 bulletin board 75 C 2 65 81 85 C 2 65 Campbell John vii Carroll Lewis vii CCD 7 ccp4i 21 69 centricity 90 CIF 91 citation CCP4 2 column labels MTZ 90 column types MTZ 89 combination of data 8 command line arguments 15 20 21 common error 19 common keywords 17 condition code 19 configuration 67 connectivity map 48 constitution CCP4 1 coordinate transformations 105 coordinates interpretation and manipulation 25 copyright 83 Cowtan Kevin vii cross rotation function 5 40 cross vectors 42 cross validation 55 crystallisation database 77 crystallographic software non CCP4 76 Cygnus Support 74 Daly Peter vii data collection 5 7 data combination 8 data formats 81 data reduction 7 DCL 65 85 density maps 6 density modification 47 104 density protein 51 derivative data 8 difference map 54 diffractometers 5 7 direct methods 30 distributed file system 71 documentation 15 81 82 metalanguage 15 Dodson Eleanor vii e mail 73 Easter egg 113 environment variable 15 16 20 error messages 19 Eulerian angles 40 Evans Phil vii 1 examples 65 67 FAST 7 fast Fourier transform see FFT FFT 9 59 83 file extension default 16 e
124. inue dividing the original transform This is known as factoring We then get a series of terms like Fr es in this case for a Fourier IC F Gau reputedly used the equivalent of real valued FFTs in 1805 but no surprise there There were independent discoveries of FFT algorithms in modern times preceding that of Cooley and Tukey often referred to as the discoverers It is interesting to note that there are various software patents connected with the FFT and similar transforms which monopolies may restrict your freedom to use these algorithms this is of course impossible Such software monopolies are a serious threat to free software and quasi free software such as CCP4 Oppose software patents and interface copyrights 59 60 10 2 1 Example 10 2 2 Why is it fast The Fast Fourier Transformation transform of points that are successively even odd even etc in the successive subdivi sions of the data In the extreme case when N is some power of two we can divide the data all the way down to transforms of length one A transform of length one is just an identity operation Now the transform consists of terms like ES even _ fa 10 2 The problem now is with book keeping we need to know what n is The answer is as follows Let even 0 odd 1 and then read the resulting patterns backwards i e reverse the bits and you get in binary the value of n It works because the successive subdivision of data into even an
125. ion vector we want to determine For m molecules in the unit cell there will also be atoms at R x p t wihl lt m lt j where R is the rotation matrix and t the translation vector for the appropriate symme try operation The structure factor equation now becomes F h yy fen Rj x p t j i Y F h ech Rip 7 5 J where F h fje Rjxi j The F h s which are independent of p need only be calculated once and the R factor for each p can be calculated without having to loop over all atoms for each value of h In general the R factor varies only slightly as the model is moved throughout the asymmetric unit so a sudden drop of a few percent is likely to be significant Difficul ties may arise mainly from two factors e The R factor is quite sensitive to misorientation of the model a misorientation of a few degrees may be disastrous If no clear solution is obtained initially it may be worthwhile though tedious to repeat the calculation with the orientation slightly altered e The second point to remember is that the R factor varies steeply with the position of the model and if the function is not sampled on a sufficiently fine grid the solution may easily be missed altogether If possible the grid points should be spaced no less than 1 apart Despite all the tricks to speed things up the R factor search is still slower that the Translation Function For example in a P6522 problem the Translation Function
126. ions to parameters to be fitted is too low for conventional least squares minimisation to converge 9 6 Free R factor A free R factor may be calculated by excluding a randomly chosen fraction of reflex ions from the refinement Bringer 1992 a special case of the technique of cross validation A T Br nger 1995 The agreement between their Fp and F is indepen dent of the refinement procedure freerflag may be used to add a column of tags to an MTZ file to label this set of reflexions This is also included in the uniquei fy script which should be run on a dataset as soon as possible see unique documenta tion Note the CCP4 convention for this differs from X PLOR see the freerflag documentation 2mtz and mtz2various may be used respectively to import and export X PLOR and SHELX datasets with a free R flag to CCP4 taking into account the different conventions Since the deviation in Rfree is roughly proportional to its value divided by the square root of the number of reflections a test set of about 1000 reflections should be accept able Both Reryst and Rfree are global measures which cannot detect local errors If atoms are placed in correct positions the R factor will decrease even if they are chem ically inappropriate NCS will reduce the value of Rfree and different types of NCS The CCP4 manual 56 Refinement and validation will have different effects Any pseudo lattice where the NCS does not increase the reciprocal
127. is resolution the least squares minimum is no longer well defined Unless there is non crystallographic symmetry or extremely high solvent content it is therefore foolish to refine against data sets at resolutions below 2 8 A If there is NCS this limit can be relaxed with care always ensuring the number of parameters is less than the observations this absolutely requires the NCS to be imposed Caveat if your NCS is close to pseudo crystallographic symmetry e g P21212 The CCP4 manual 9 7 3 Bad practice 57 but pseudo 1222 or P65 pseudo P6522 then it is less powerful and you will have special problems e Do not try to refine individual isotropic atomic B values till you have enough observations to about 2 5 A or better e The significance of introducing extra parameters should be cross validated using Riree The CCP4 manual Chapter 10 The Fast Fourier Transformation or some of the things you always wanted to know about the FFT but were afraid to ask 10 1 Introduction 10 2 How it works The CCP4 manual Of course the Mock Turtle said advance twice set to partners change lobsters and retire in same order continued the Gryphon Alice s Adventures in Wonderland There must be very few algorithms that have made as great an impact on science as the Fast Fourier Transform FFT It was Lynn Ten Eyck 1973 1977 who introduced the FFT to crystallography and we have been using it ev
128. is section summarises the results of Matthews 1968 for solvent content Matthews number is W cellvolume A E 8 1 Mnasymunmolsasu MZ Where M is the protein molecular weight in Daltons V is the unit cell volume Nasymu is the number of asymmetric units nmolsasy is the number of molecules in the asymmetric unit Z the number of molecules in unit cell nasymunMolSasu Molecular weight is e something you get off a gel e or number of protein residues in molecule x 1 10 very roughly e or number of non hydrogen protein atoms in molecule x 14 roughly The CCP4 manual 8 4 Estimating solvent content 51 Use rwcontents to read your PDB file if you have one it will count the number of atoms of each type The output of truncate provides information on the cell contents Matthews found Vm between 1 66 and 4 0 corresponding to protein contents of 30 75 but proteins with higher solvent contents will give higher values of Vm Solvent content of 90 gt Vy 12 Calculate Vu assuming nmolSasu 1 2 3 etc You may be able to narrow down the number of possibilities for nmolsagu Turning this into the fraction of protein in the asymmetric unit MZAy Av 8 2 VNAD NaDpVm where Vp is the fraction of protein volume in the asymmetric unit Dp is the density of protein 1 35 Matthews 1968 Ay is the average atomic volume in 10 approximately Na is Avogadro s constant This is the sam
129. istribution always contains a file named README When asked hackers invariably relate the README convention to the famous scene in Lewis Carroll s Alice s Adventures In Wonderland in which Alice confronts magic munchies labelled Eat Me and Drink Me Raymond 1993 The distribution is supplied in source form You will need Fortran C and C compil ers See 813 for hints about freely available systems if necessary To build phaser python is required By default the unsupported programs will be built as well as the main set You may want to avoid building them to save space Make sure you have plenty of spare disc space to build in You need 250 MB free for CCP4_MASTER if everything in ccp4 is downloaded off the web site ccp4 phaser The space taken by the full set of binaries will vary con siderably with system and compiler options but is 350 MB on an Iris compiled with the default configuration you can reduce this either by playing with the configuration options or only by compiling a subset of the programs If ccp4mg and coot are also downloaded the space requirements will increase to 1GB Some of the programs use considerable amounts of scratch space in execution you will need 10s of MB free in the directory pointed at by CCP4_SCR see ccp4 setup Unix In some cases you may find a need for 100 MB of scratch space depending on the programs you use and their input Some programs take up large am
130. ivative we obtain a circle with radius Fpp and origin at Fy that should intersect the Fp circle at one of the two previous values giving a unique solution for Op In practice however experimental errors approximations in determining Fy and lack of isomorphism affect the precision of this graphical construction and usually we need more than two derivatives to obtain satisfactory protein phases MIR or Multiple Isomorphous Replacement 6 6 2 Anomalous scattering data In phase determination using anomalous data the out of phase component F acts in the same way as the entire vector Fy in the isomorphous method we can obtain Bay from Fpy 2F in the same way as Fpp from Fp Fy Since Fi and Fy are in general perpendicular one each other they give complementary information and in the absence of errors one could get a unique solution for the phase Op by combining isomorphous and anomalous data Although this phase solution is unique it is es sential that phases are generated from both possible heavy atom solutions x y z and x y z Each solution will give the same phasing statistics one will be approxi mately correct and the other wrong The solutions obtained by combining the Single The CCP4 manual 6 6 3 Multi wavelength Anomalous scattering Data 37 Isomorphous Replacement and Anomalous Scattering data using a single derivative are referred to as SIRAS phases Phasing by SIRAS benefits from the use of synchrotron ra
131. ks If you can t solve the problem contact CCP4 see 812 2 and we will help if we can but no promises One thing you can do is to try building the library and running the test program for it with make testlib to look for clues if you complain to CCP4 we will ask you for the results of this if we suspect a low level library problem In the meantime if there are problems building some of the programs make fails on the srcdir target having done the libdir target successfully you can build as much as possible using make i and then install this much with make instsome in the src and or unsupported src directory The instsome target doesn t insist that everything is properly built first 11 3 8 Saving space and shared libraries The program binaries are big and you may need or want to make them smaller There are various possible ways to do this which are system dependent Increasing the op timisation level of the compiler might make the binaries bigger or smaller However under agressive optimisation the programs may not work at all Using shared libraries will definitely make them smaller On some systems you can use the configure argument with shared lib successfully to build a shared version of the CCP4 library to link against but make sure you have the correct setup to find the shared CCP4 library at runtime namely the location of the shared library should be included in the environment variable LD_LIBRARY_PATH Other system sup
132. l translation functions that make use of known structure information and can be expanded as Fourier series In Machin 1985 Tickle I J 1992 Fast Fourier Translation Functions In Dodson et al 1992 Tickle Ian 1991 Refinement of SIR heavy atom parameters in Patterson vs reciprocal space In Wolf et al 1991 Tronrud D E 1994 Methods of Minimization and their Implications CCP4 Daresbury Study Weekend nos DL SCI R35 ISSN 0144 5677 Warrington WA4 4AD UK Daresbury Laboratory for Daresbury Laboratory Vaguine A A Richelle J amp Wodak S J 1999 Acta Cryst D55 191 205 The CCP4 manual 101 Wang Bi Cheng 1985 Resolution of Phase Ambiguity in Macromolecular Crystal lography In Wyckoff et al 1985 Wolf W Evans P R amp Leslie A G W eds 1991 Isomorphous Replacement and Anomalous Scattering CCP4 Daresbury Study Weekend nos DL SCI R32 ISSN 0144 5677 Warrington WA4 4AD UK Daresbury Laboratory for Daresbury Laboratory Woolfon M M amp Yao Jia Xing 1994 On the Application of One Wavelength Anomalous Scattering IV The Absolute Configuration of the Anomalous Scat terers Acta Cryst D50 7 10 Wyckoff H Hirs C H W amp Timasheff S N eds 1985 Diffraction Methods for Biological Macromolecules Methods in Enzymology vol 115 Academic Press Zhang K Y J amp Main P 1990 Acta Cryst A46 377 381 Zlotnik Fred 1991 The POSIX 1 Standard A Pr
133. lacement parameters U values e A bulk solvent correction is calculated within the program see the SOLVENT keyword e For atomic resolution data full anisotropic refinement can be performed with anisotropic displacement parameters being refined for some or all atoms e If good experimental phases are available then they can be included in the maxi mum likelihood target The accuracy of experimental phases as described by the Figure of Merit or the Hendrickson Lattman coefficients is often overestimated and a blurring function is provided to compensate for this Rigid body refinement can be performed and may be useful in the early stages of refinement One or more rigid body domains can be defined via the RIGID BODY keyword TLS refinement in refmac5 TLS refinement in refmac5 results in e TLS parameters for each defined TLS group held in the TLSOUT file and in the header of the XYZOUT file e Residual B factors output in the ATOM lines of the XY ZOUT file These B factors do not include any contribution from the TLS parameters The XYZOUT and TLSOUT files can be passed to the program t1san1 which will analyse the TLS tensors and also derive individual anisotropic displacement parameters from the TLS parameters Automated model building Victor Lamzin s Automated Refinement Procedure program renamed arp_warp due to a clash with a Unix command can be alternated with a refinement program such as refmac5 to automatically bu
134. later cycles this agreement will have in creased through the averaging of the regions and so becomes less informative The solvent flattening and molecular averaging techniques are closely related and are most effective when used together They are both noise filtering techniques and involve modification of electron density They can be used to improve and extend phases The CCP4 manual 50 Phase Improvement 8 3 Using Solomon Several stages are required for phase refinement with Solomon The stages described here assume that the phase probability distributions were determined experimentally i e SIR MIR MAD e Determine phase probability distributions described by Hendrickson Lattman coefficients This is done automatically by mlphare Calculate the map to be flattened Initially you will have a map calculated from experimental phases It is advised that you start from a resolution which has got significant phase information and do not yet extend the resolution Subsequent maps can be calculated from coefficients producedby sigmaa This gives the advantage that missing reflections can be estimated from the corre sponding Fc calcuated from the flattened map The procedure can be improved by substituting in this way for the low resolution reflections that are missing e Have a look at the map and compare it with the original FOM weighted map If there is no non crystallographic symmetry you might get some additional improvement
135. lf the ideal height The second term will contribute only noise since there is no connection between the Fy and the phases used The fact that for acentric reflexions the magnitude of the differences between native and derivative amplitudes is not generally equal to the heavy atom structure factors The CCP4 manual 34 Isomorphous Replacement does not represent the only cause of noise Much more serious sources of noise are measurement and scaling errors as well as lack of isomorphism Since the size of the Fy contribution is two or three percent of Fpy and Fp it may be of the same order as the standard deviation of the observations and since a Patterson function is dominated by its largest terms a few spuriously large differences can disguise the Patterson signal completely An anomalous difference Patterson synthesis is a Fourier summation of the terms Ba 2 ES Foul 2K Fi sin tp n 6 13 The formula sin 6 1 2 cos 20 2 means the summation of AF2 should also give peaks in the Patterson corresponding to vectors between heavy atoms plus noise There is no lack of isomorphism in the anomalous difference but in general the F signal is much weaker than the Fy signal so the random errors inherent in the measurements can swamp it more easily Better Patterson maps can sometimes be obtained using estimates of Fy which take into account both isomorphous and anomalous data i e the summed or combined Fi E P
136. line coordinate transformations 33 33333333 The CCP4 manual 106 Program index reindex 12 Reindex MTZ file restrain 10 53 55 Structure refinement Driessen et al 1989 revise 9 Program to generate normalised anomlous scattering factor magni tudes from MAD data rfcorr 9 Analysis of correlations between cross and self rotation functions rotamer 11 List amino acids whose side chain torsion angles deviate from Richardson s Penultimate Rotamer Library rotaprep Unsupported Produces multirecord MTZ file from foreign formats rotgen 7 Simulate X ray diffraction rotation images rotmat 9 Interconverts CCP4 mer1lot X PLOR rotation angles rsearch 9 12 40 R factor search rsps 8 30 34 Heavy atom positions from derivative difference Patterson maps rstats 8 Scales two data sets rwcontents 11 51 Count atoms by type in PDB file rwdict 26 Unsupported PDB lt gt PROTIN dictionary inter conversion sapi 8 9 Heavy atom site location program sc 11 Program to analyse shape complementarity of molecular interfaces scala 7 23 24 89 Scaling together multiple observations of reflections re placement for ROTAVATA and AGROVATA scaleit 8 29 Various derivative and native scaling scalepack2mtz 7 Converts merged scalepack output into MTZ format sfall 9 30 39 40 61 93 Structure factor calculations using FFT sfcheck 11 Program for assessing agreement between atomic model and X ray
137. ll have to ensure that they have access to the resources relevant application defaults files Where these files are might depend on how you install the programs and the mechanism for picking up the resources is likely to vary from site to site An example is given in ccp4 setup for Unix users The X Windows based programs can now be built on SGi machines running irix PC Linux based machines and DEC machines using the general CCP4 procedure see the with x option below If you are not using one of these systems consult the README files in the x windows directory for instructions 11 6 The full story on configure 811 3 described a simple use of configure This section explains its use in more complicated situations where it really comes into its own The Unix configuration system allows you automatically and consistently to cre ate the system dependent Makefiles etc to operate several systems with shared files in a distributed filesystem and to test new versions without disrupting the pro duction ones It is based on a Bourne shell script some Makefile skeletons and macro processable program files which are manipulated under the script s control The configure script encapsulates most of the information about the system dependen cies in the suite and automates their application The rest of this information resides in the macro processor directives in unix m4 and the ch files The main configuration parameters apart
138. lue for f as the native Negative occupancies are not necessarily incorrect and do not affect the phasing It just means that the average scattering power of the derivative is less than that of the native 6 6 4 Density modification Isomorphous phases are usually improved by use of a density modification procedure 88 before the initial map is calculated 6 6 5 Treatment of errors Errors are now refined as part of the Maximum Likelihood refinement and phasing procedure Up to now we have assumed that the analysis is free of errors In practice however there are errors that can be thought of as deriving from two different sources e Errors in measuring intensities of reflexions will produce errors 5 in the ampli tudes Fp and Fpy e Errors in locating and describing the scattering of the heavy atoms as well as lack of isomorphism will introduce errors in Fy The result is that the two derivative circles will not intersect the native one at precisely the same point The total error E is a measure of the discrepancies between theory and experiment EY 6 e 6 17 where 5 can be estimated by comparing symmetry equivalent reflexions from the same crystal or the same reflexions from different crystals The estimation of E can be made for centric reflexions from the difference between Fpy Fp and Fy in acentric cases this is not possible because Fpy Fp is not a correct estimate of Fy The error
139. lves develop ers of CCP4 code on which policy is discussed problems aired fixes suggested etc To subscribe send a message SUBSCRIBE CCP4 DEVto Majordomo dl ac uk The list itself is ccp4 dev dl ac uk Here is a summary of the relevant network addresses Majordomo O dl ac uk is for un subscribing to the CCP4BB and CCP4 DEV lists ccp4bb dl ac uk is for broadcasting messages to the CCP4BB subscribers ccp4 dev dl ac uk is for broadcasting messages to the ccp4 dev subscribers ccp4 dl ac uk is for contacting the Daresbury CCP4 staff for bug reports etc ccp4a dl ac uk is the address for retrieving files by anonymous ftp 13 4 Other crystallographic software Here is how to obtain some of the other crystallographic software mentioned herein but not distributed by CCP4 Network references are given as Uniform Resource Locators A URL like ftp site directory means anonymous ftp to directory at site Babel ftp joplin biosci arizona edu BIOMOL ftp rugcbc chem rug nl pub lt cfb chem rug nl gt corels ftp sgjsl weizmann ac il pub corels demon ftp lccp ibs fr dist demon F M D Vellieux lt vellieux lccp ibs fr gt dssp Chris Sander lt Sander embl heidelberg de FRODO TOM lt frodo request biochem ualberta ca gt madnes Jim Pflugrath lt pflugrath cshl org gt madsys Bill Weis lt weis cuhhca hhmi columbia edu gt merlot Paula Fitzgerald Merck Sharp and Dohme Research Laborato ries PO Box
140. m positions for proteins hklplot 12 27 Plots a precession picture from a reflection file hklview 7 12 27 Displays zones of reciprocal space as pseudo precession images under X windows icoefl 8 Scaling of multiple Fes to Fops ipdisp 7 Display various types of image file under X Windows libcheck 10 26 54 Checks a monomer s description for refmacS loggraph 11 19 lsqkab 11 26 Optimise fit of atomic coordinates from two data sets Kabsch 1976 adlat 7 Unsupported Reads madnes output file s generates vector list for plotting in rodo to visualize reciprocal lattice akedict Converts PDB file to TNT dictionaries ama2ccp4 10 25 Converts MAMA format masks to CCP4 ones ap2fs 10 Convert a CCP4 map to XtalView fsfour format apdump 9 25 Prints sections of electron density maps apexchange Unsupported Converts binary map file into ASCII one and reverse Obsolete use NA4 format except for archaeological purposes apmask 9 10 25 55 Map and mask manipulation apreplace Unsupported Combines parts of two maps aprot 10 Map skewing interpolating rotating and averaging program apsig 9 10 40 Print statistics on signal noise for translation function map Also sum product ratio of two maps apslicer 11 27 Section viewer for CCP4 map files 3 3 8 9 8 8 3333 3 The CCP4 manual Program index 105 3 aptona4 20 Converts binary map file to NA4 format and reverse atthews_coef 12 Misha Isupov
141. mber 50 maximum likelihood 6 33 54 maximum likelihood phasing 30 35 Mc Laughlin Sandra vii 87 merging 5 7 23 metalanguage documentation 15 Miller indices 19 MIR 29 49 missing data 89 molecular dynamics 6 molecular envelope 6 molecular replacement 6 8 39 48 49 molecular weight 50 51 MS Windows 12 MTZ 17 87 header 17 MTZ files 19 column labels 90 column types 89 multi record 19 MTZ format 19 multi record files 7 19 23 24 88 NA4 format 87 Naismith Jim 1 NaN 89 NCS averaging 9 Netlib 83 network news 77 noise filtering 49 non crystallographic symmetry 5 6 39 43 48 refinement 49 numerical code 83 Onesti Silvia vii operating system dependencies 85 operating systems 73 85 orthogonal coordinates 26 outliers 23 Program index output verbosity 21 partiality 90 partially recorded reflexions 7 Pascal 81 patch files 73 patents 59 Patterson 30 Patterson functions 32 Patterson map 5 6 8 Patterson search 8 Patterson vectors 8 PC 13 peak search 8 phase combination 8 phase determination 36 phase extension 48 49 phase improvement 9 48 49 phase refinement maximum likelihood 35 phased translation function 40 Photon Factory 76 pitfall 16 PLOT84 84 93 PLOT84 93 plotting 6 12 25 27 portability code 81 82 85 data files 19 porting 73 POSIX 81 85 postrefinement 7 PostScript 12 15 66 pre processor 83 pre
142. mer the solution should correspond to the appropriate point group symmetry A stereographic projection of e g ak 180 in polar convention section of the rotation function can be plotted It should contain all the peaks corresponding to two fold axes If for example we have a tetramer in the asymmetric unit point group 222 the 180 section should contain a set of three orthogonal peaks 90 apart 7 3 8 3 Concluding remarks In a typical case several calculations of the rotation function are needed at various resolution limits and integration radii to find the correct solution The correct solution does not always correspond to the highest peak in the rotation function but it should be the most persistent feature as the different parameters are varied It is worth trying to solve the rotation problem as accurately as possible as all the methods for finding the translational parameters are quite sensitive to mis orientation errors The CCP4 manual 44 Molecular Replacement 7 4 The translation search 7 4 1 R factor search Having found the orientation of the molecule s in the asymmetric unit we now have to determine the translational parameters needed to position the molecule correctly relative to the symmetry axes We might expect this to be easier than solving the rotation problem since the methods we use for this task involve correlating the complete set of structure factors or Patterson functions In some cases however it tu
143. n and the string is a suitable message Programs should terminate successfully with the call CALL CCPERR 0 Normal termination Keywords wherever possible use the standard input keywords Subroutines for han dling the input of the most common keywords are provided in the parser sub routine library or probably better the higher level keyparse library Logical names Where possible the logical names for file connexion should have stan dard values as follows see also environ def HKLIN Input reflexion file HKLOUT Output reflexion file XYZIN Input coordinate file XYZOUT Output coordinate file MAPIN Input map file MAPOUT Output map file The DATA and PRINTER files for parameter input and log output respectively are assumed to be preconnected although it is good practice to call CCPDPN for them anyhow If several files of one type are required then their file names should be derived by adding numbers to the names given above e g HKLIN1 HKLIN2 File opening There should be no OPEN statements in Fortran code Instead callCCPDPN for general files and appropriate library routines for the standard data formats This avoids problems with such things as VMS READONLY Version printing A call to CCPRCS will be inserted to output version information if it is not present already MTZ history Programs which operate on MTZ files should use the subroutine LWHIST to write a line of informational text to the history header
144. n the build process There are a number of configuration parameters that you don t usually want to change but whose defaults can be over ridden by the values of environment variables when configure is run CC the C compiler default is system s native compiler another possibility might be gcc COPTIM C compiler optimisation flags default is the highest optimisation which is considered safe or unagressive XCFLAGS any extra flags you need to give to the C compiler apart from those for optimisation these are system dependent FC the Fortran compiler default is system s native compiler there isn t currently any support for other than native compilers FOPTIM Fortran compiler optimisation flags The default varies but will have no debugging extras In some cases notably IRIX the default is used because there are so many problems encountered with the optimiser You may want to try higher optimisation levels and see if there s a significant performance improve ment and that the code still works XFFLAGS any extra flags necessary for the Fortran compiler apart from those for optimisation you might want to change this for instance to add s to make smaller binaries at the expense of debugging ability CXX the C compiler CXXOPTIM C compiler optimisation flags XCXXFLAGS any extra flags you need to give the C compiler XLDFLAGS any extra flags needed for 1d typically extra libraries F_LIBS the additional libraries
145. nagement system which allows work to be divided into separate projects and which automatically keeps a database of the jobs that have been run in that project The database retains details of input and output files as well as automatically storing the parameters which makes it easier to keep track of progress The graphical user interface offers some other useful features for example it has an extensive system of help files in HTML format and is linked to a number of util ities for viewing different types of files for example loggraph sketcher and mapslicer 7N B The mechanism for looking for files in CLIBD can cause confusion if you do want to use your own version of a file from the current directory In this case either give it an extension different to the list here or use an explicit path e g SYMOP symop lib Unix 8Though purists will object that this is not a true database The CCP4 manual 22 General CCP4 usage To start up the interface within a Unix environment use the command ccp4i at the command prompt Within Microsoft Windows environments click on the CCP 4 icon or select the appropriate option from the Start menu The CCP4 manual Chapter 4 Data processing and reduction 4 1 Data processing 4 2 Data reduction The CCP4 manual About a thousand and four said Bruno You mean about a thousand Sylvie corrected him There s no good saying and four you can t be sure ab
146. nd plotting 2 2 22 oo mn nennen Interpretation and manipulation of coordinates 5 3 1 Converting between fractional and orthogonal coordinates PIOUS ns teats rates doses oe Be he Oe BREA Se es tes de fh 6 Isomorphous Replacement 6 1 6 2 6 3 6 4 6 5 6 6 7 7 1 7 2 7 3 7 4 8 8 1 8 2 8 3 8 4 MIR structure solutions steps and programs Combination of phase information 2 2 222 nenn BackStound roe ss se ed DE a ee Determination of heavy atom positions 2 o nenn 6 4 1 Scaling between native and derivative data 6 42 Estimationof FH 6 4 3 Determination of the heavy atom positions Refinement of heavy atom parameters 6 5 1 Maximum Likelihood Phase refinement 6 5 2 Vector space refinement o a 6 5 3 Fp refinement soe m peie a a A A e Phase determination ooa e 6 6 1 Isomorphous replacement data 6 6 2 Anomalous scattering data 6 6 3 Multi wavelength Anomalous scattering Data 6 6 4 Density modification a a 6 6 5 Treatment of errors ooo e 6 6 6 Hendrickson Lattman coefficients 6 6 7 Single Isomorphous Replacement Molecular Replacement Molecular replacement steps and programs 7 1 1 Detection of non crystallographic symmetry using self rotation FUNCHONE e
147. ns and translations generates symmetry equivalent coordinates truncates B values outside specified range renumbers residues changes chain ids Inputs outputs X PLOR format O format etc finds range of a PDB file 1sqkab least squares fits subsets of coordinates together using the Kabsch al gorithm and outputs rotated coordinates baverage averages and analyses B factors residue by residue watertidy part of a procedure to move and rename water molecules to sen sible values relative to the protein molecule sortwater similar to watertidy rwdict converts a PDB file into PROTIN dictionary format chemnotes commercial will build you a substrate and output a PDB file e Coordinate checking act Various analyses distang calculates distances and angles between near atoms using symmetry Used to check bad contacts and find hydrogen bonds etc contact similar procheck checks file for standard naming conventions and sorts the atoms within a residue into the standard order Checks structural geometry should be used at all stages of refinement libcheck refmac5 which are part of the refinement package also check for bad distances planarity chirality torsion angles They can be used for this purpose without the refinement option Converting between fractional and orthogonal coordinates This is a recurring problem especially when using coordinates extracted from direct methods programs or from mlphare To remind you If you have a P
148. nships with other molecules from which a portion of the structure can be deduced This information can be applied through non crystallographic symmetry averaging Bricogne 1974 and molecular replacement respectively Molecular averaging can be used in the case of non crystallographic symmetry NCS NCS related regions of the map are averaged together thus reducing the signal to noise ratio in the density The NCS relationships and masks must be determined before this calculation can be performed Atomicity In the high resolution limit the electron density map for a structure will show resolved atomic features Atomicity can be enhanced by modifying the map to satisfy Sayre s equation Zhang amp Main 1990 Cowtan amp Main 1993 Sayre s equation is a technique from small molecule direct methods which ap plies the constraint of atomicity to the density Protein structures do not normally diffract to atomic resolution however in the case of a good starting map and ob served magnitudes at close to 2 0 A Sayre s equation can be a powerful tool for phase improvement Note that all of these features can be distinguished in a map without knowing the de sired structure Techniques have been developed by which a poor density map can be altered in a sensible manner in order to bring it into line with each of these constraints these are set out in more detail in the references given above All these techniques depend on knowledge of the m
149. nt not taken to convergence is difficult to detect but happens there are many examples of structures refined first against one data set to a certain resolution then re refined against a different higher resolution data set The final structure gives a lower R factor against the original data than the one refined against that data This is may be due to poor weighting of prior and experimental information 9 7 3 Bad practice 9 7 3 1 Not using all the available data e Do not use a low resolution cut off e g many structures are still reported as being refined with data in the resolution range 5 2 A The data within the 5A shell contain a wealth of important information on your structure e Make sure you have not lost all the strong often low resolution terms through detector saturation especially important with image plates at synchrotrons Make a second data collection pass or even a third to avoid this The big terms domi nate all steps in your structure analysis If you can possibly avoid it do not leave a large wedge of data uncollected Offset your crystal by up to 15 to avoid a blind region Make sure you cover the appropriate rotation range and start at an appropriate orientation 9 7 3 2 Attempting refinements when the observation to parameter ratio is too low e At about 2 8 A for a protein crystal with about 50 solvent the number of obser vations is equal to the number of positional xyz atomic parameters Even at th
150. oduction distang checks for bad symmetry contacts There are various other programs for checking packing including ones from merlot and pacman from Uppsala graphics rigid body refinement e g amore Castellano et al n d CNS It is often possible to determine the structure of a molecule using a similar or homolo gous structure that is already known One such case may be when we try to determine the crystal structure of a known molecule which has crystallised in a different space group Another area in which molecular replacement MR is particularly useful is site directed mutagenesis where large numbers of mutants are produced and studied Even a minor change in sequence may change the way a protein crystallises The most economical way to solve such structures is usually MR Another of its uses involves different proteins belonging to the same structural family As the number of known structures increases it is becoming apparent that the number of distinct protein folds is limited and that even functionally unrelated proteins may share a common fold It is therefore increasingly likely that a new protein has a structure related to one that is already known MR then becomes a possible way of solving it Thus although it is already an important technique MR is bound to be used to an even larger extent in the future In MR a preliminary model of the crystal structure is obtained by first orienting rotating and then positioning translating
151. of a user defined solvent mask output of the internally calculated solvent mask averaging of multiple non crystallographic symmetry related domains with different matrices and automatic refinement of non crystallographic symmetry matrices The output of the program can be examined using xloggraph The most useful indicator of whether the method has been successful is the free R factor This is analogous to the free R factor used in coordinate re finement in that it is calculated using a set of reflections which are excluded from the initial map calculation However once phase relationships between structure factors are introduced it becomes impossible to completely isolate a set of reflection since the absence of a set of reflections leads to a systematically error in the phasing of the rest of the reflections Thus it is necessary to change the free R set from cycle to cycle or if time is not a factor to run each cycle twice with and without a free R set The density modification free R is a property of both the density modification tech nique and the initial data set so it can not be used to compare different density mod ification techniques During a typical calculation the density modification free R will drop from 0 60 to 0 45 In the case of averaging calculations the correlation between regions related by non crystallographic symmetry provides a good indication of whether the mask and matrices have been correctly determined In
152. ogical name SYMOP is associated with a file that is input only in and has a default extension of 1ib other file types are out output only and inout for read write files After this the pre processor then reads in the file default def which speci fies some common logical names defined for all CCP4 processes A typical line here may be SYMOP syminfo 1ib and means that the logical name SYMOP will be as sociated with the file syminfo lib You can edit default def to use different files by default e g a new atomic scattering dictionary but usually the defaults are overridden by command line options 83 7 1 Finally the command line arguments are parsed and the logical name file name pairs are matched up and assigned An alternative to specifying logical names on the command line is to define a Unix environment variable corresponding to the logical name with the file name as its value although this is not the recommended way of operating Command line values override any defined in the environment A logical name on the command line that has not been defined in environ def causes the name to be added to its internal list and a warning to be issued Within this procedure are several devices to make running the program easier Firstly files without extensions have them appended using the default value where de The CCP4 manual 3 7 1 3 7 1 3 8 Command line Command line 21 fined in environ def Secondly
153. ogrammer s Guide Ben jamin Cummings The CCP4 manual Program index CCP4 programs The CCP4 manual And thick and fast they came at last And more and more and more Through the Looking Glass abs 8 9 Determine the absolute configuration hand of heavy atom substructure absurd 23 acorn 9 ab initio procedure for the determination of protein structure at atomic resolution act 10 11 26 Analyse coordinates almn 9 39 40 61 Crowther s rotation function using FFT amore 8 9 39 40 Molecular replacement Navaza 1994 angles 10 Unsupported Bonds and dihedral angles from coordinate files anisoanl 10 Analyses of anisotropic displacement parameters areaimol 10 Solvent accessible area or area differences arp_waters 10 54 Automated Refinement Procedure version 5 0 asc2p84 19 Unsupported Converts ASCII file to PLOT84 type meta file astexviewer 11 27 Java program for interactive 3 d display of molecular structures and electron density maps axissearch 12 51 Unsupported Changes axis and cell baverage 11 26 Averages B over main and side chain atoms beast 8 Brute force molecular replacement with Ensemble Average Statistics Maximum likelihood based molecular repaclement Obseleted by phaser bones2pdb 10 25 Convert bones output to a PDB file BP3 9 multivariate likelihood substructure refinement and phasing of S MIR AS and or S MAD bplot 11 Plot average B factors along the chain bulk 8
154. olecular boundary dm can cal culate this automatically by Wang s method Wang 1985 Leslie 1988 if the fraction of the unit cell occupied by solvent is known The phase improvement calculation is cyclic and involves modification of the phases in order to obtain best agreement with the known constraints on the electron density and the observed structure factor data This process is shown diagrammatically in fig 8 1 Fobs h stow e Pito Phase Density Recombination Modification 541 h w h F mod h FF moa e PA p 1 X Figure 8 1 Phase improvement schematic The CCP4 manual 8 2 Using dm 8 2 Using dm 49 dm provides an automatic procedure for using the techniques described above A single run will read an initial reflection file apply multiple cycles of density modification and produce an updated reflection file The initial reflection file must contain at least the observed structure factor magni tudes and their deviations and the estimated phases and their weights In the case of an MIR dataset the phases and weights will come direct from the MIR phasing typically from the program mlphare It is also possible to use MAD data although typically the phase extension scheme will be different since there is usually some phase infor mation at all resolutions This sort of calculation has proven particularly effective Density modification can also be applied in a molecular replacement calculation In this
155. olumn types and new labels if necessary moves all data to the same asymmet ric unit necessary if you are bringing data together from several sources and outputs all columns It can apply a scale factor to all columns in a file exclud ing phases It can also be used to extend your data to cover more of reciprocal space sortmtz Sorts merges MTZ files mtzutils has many purposes It can e Select columns from one or two input files e Merge interleave two sorted files e Change header or labels e Change symmetry etc 5 2 Map file utilities and plotting mama2ccp4 Convert MAMA format masks to CCP4 format mapmask General map and mask manipulation program O displays maps amongst other things overlapmap does map manipulation npo can contour map sections amongst other things mapdump prints sections of maps and is useful to extract the RMS density deviation for assigning plotting contour levels xdlmapman Interconverts formats and does manipulations 5 3 Interpretation and manipulation of coordinates Most CCP4 programs use the Brookhaven format for reading and writing coordinates They require the CRYST1 and SCALEi cards to be present and use these to convert to fractional values when required See also 816 2 e Manipulation bones2pdb Convert MAMA format masks to PDB format coordconv interconverts formats pdbset The CCP4 manual 25 26 5 3 1 Manipulating and Displaying Data Files applies specified rotatio
156. olution shells of constant volume and scal ing the data so that the mean intensity does not decrease with resolution This is like calculating the E values used in direct methods The scaling is usually done by apply ing an exponential temperature factor to the structure factors The normalised data are what we would get if atoms were like geometric points rather than spheres of electron density Normalising increases the contrast of the Patterson map and the peaks should be better resolved If we decide to leave weak terms out of the calculation it is important that we do it after normalising Thus we ensure that the reciprocal space is sampled evenly and not dominated by a few large low resolution terms Normalising can introduce series termination ripples in the Patterson but this has not been found to be a serious prob lem The procedure can be recommended generally for both rotation and translation searches A Bayesian justification is given by Bricogne in Dodson et al 1992 N B Normalising is related to but distinct from sharpening where high resolution intensities are scaled up by applying an exponential B factor The CCP4 manual 7 3 7 Origin Removal 43 7 3 7 Origin Removal Patterson maps always have a strong origin peak It arises from the null vectors i e from an atom to itself Normally the origin peaks just add a constant term to a rotation function There is however one case where the origin peak can cause problems
157. on of Fy Centric reflexions The first step is to scale the native and derivative data sets together In comparing different sets of data the normal procedure is to use a Wilson plot to put both of them onto the same absolute scale but this is never very reliable with data to a resolution below 2 5A A sort of relative Wilson plot comparing F and F2 can also be used Dodson 1976 This will not be accurate since no allowance can be made for the additional contribution from Y des and this quantity depends on the degree of heavy atom substitution which is difficult to estimate at the beginning However the various Patterson functions are not very sensitive to errors in scale of up to 10 and this estimate is usually good enough for preliminary calculations Kraut suggested a relatively simple way of calculating the derivative scale factor by equating the Patterson origins Y A Y Ken Y FR 6 2 where K is the scale factor to be estimated Kraut et al 1962 Tickle 1991 see also the fhscal documentation This method often gives excellent results Once the heavy atom positions have been determined the relative scale and tem perature factor can be refined along with the other heavy atom parameters positional coordinates and occupancies It is necessary to estimate the amplitude of the heavy atom contribution Fy Once this is done we can sum the contributions to give a Patterson map where we should see heavy atom heavy
158. on the same origin 6 2 Combination of partial structure and isomorphous phase information Nat MIR map gt O build model MIR phases obtained using MLPHARE Preliminary refinement gt SFALL gt gt SIGMAA gt gt FFT SFALL PROTIN PROLSQ calculate Fc alpha c XPLOR combine alpha c s new map with Hendrickson TNT Lattman phase probabilities calculated using MLPHARE Figure 6 2 Combination of partial structure and isomorphous phase information 6 3 Background The basic technique of solving the phase problem in protein crystallography is isomor phous replacement Molecular replacement is usually limited to particular cases where a homologous structure is already known The size of the problem and the intrinsic limitation in resolution make the use of direct methods almost impossible Two crystals are said to be isomorphous if they have essentially the same structure but are composed of chemically different atoms In protein crystallography what is called isomorphous replacement should properly be named isomorphous addition since it is usually the binding of a heavy atom in a position previously occupied by disordered solvent The electron density of the isomorphous derivative should differ from the electron density of the native crystal just for the peaks due to one or a few specifically bound heavy atoms the unit cell dimensions and the protein molecule must be essentially the
159. ore information e mail biosci help net bio net or preferably see http www bio net The crystallography entry in the World Wide Web virtual library is at http www unige ch crystal crystal index html WWW access to the BMCD crystallisation database is through http ibm4 carb nist gov 4400 bmcd bmcd html 13 6 Sources of general free software 13 6 1 Free compilers Sources of free software such as the GNU project distribution include the following anonymous ftp sites at the time of writing Asia ftp cs titech ac jp utsun s u tokyo ac jp ftpsync prep cair kaist ac kr pub gnu Australia archie oz au gnu archie oz or archie oz au for ACSnet Europe src doc ic ac uk gnu ftp informatik tu muenchen de ftp informatik rwth aachen de pub gnu nic funet fi pub gnu ugle unit no isy liu se ftp stacken kth se ftp win tue nl ftp denet dk ftp eunet ch nic switch ch mirror gnu archive eu net United States wuarchive wustl edu ftp cs widener edu uxc cso uiuc edu col hp com gatekeeper dec com pub GNU ftp uu net systems gnu These are probably mainly but not solely appropriate for Unix stuff Even if you acquire the GNU software by ftp please consider making a donation to the Free Software Foundation e g by buying a tape or CD from them CCP4 devel opment depends on GNU software Their address is Free Software Foundation 59 Temple Place Suite 330 Boston MA 02111 1307 USA 1 617 542 5942 voice 1 617 542 26
160. otein Struc tures CCP4 Daresbury Study Weekend nos DL SCI R16 ISSN 0144 5677 Warrington WA4 4AD UK Daresbury Laboratory for Daresbury Laboratory Matthews B W 1966 The Determination of the Position of Anomalously Scattering Heavy Atom Groups in Protein Crystals Acta Cryst 20 230 Matthews B W 1968 The Solvent Content of Protein Crystals J Mol Biol 33 491 497 Mc Ree Duncan E 1993 Practical protein crystallography Academic Press Mukherjee A K Helliwell J R amp Main P 1989 The use of MULTAN to Locate the Positions of Anomalous Scatterers Acta Cryst A45 715 718 Murshudov G N Vagin A A amp Dodson E J 1997 Refinement of Macromolecular Structures by the Maximum Likelihood Method Acta Cryst D53 240 253 Navaza J 1993 On the Computation of the Fast Rotation Function Acta Cryst D49 588 591 Navaza J 1994 AMoRe an Automated Package for Molecular Replacement Acta Cryst A50 157 163 Otwinowski Zbyszek 1991 Maximum likelihood refinement of heavy atom parame ters In Wolf et al 1991 Otwinowski Zbyszek 1993 Oscillation data reduction program In Sawyer et al 1993 Perutz M F 1956 Isomorphous Replacement and Phase Determination in Non centrosymmetric Space Groups Acta Cryst 9 867 Perutz Max 1992 Protein structure new approaches to disease and therapy New York W H Freeman and Co The CCP4 manual 100 Podjarny A D Bhat T N amp Z
161. ounts of memory which can cause problems Personal memory useage quotas can be exceeded or for small systems like PCs hard limits 11 2 Directory structure The CCP4 manual The directories under the ccp4 directory which holds the CCP4 suite are as follows bin is the default destination for installed program binaries ccp4i graphical user interface for the suite doc plain text documentation files generated from the ht m1 files etc Unix shell scripts examples example scripts DCL procedures toxd data and model coordinates for a small protein alpha dendrotoxin from green mamba venom toxd used in the examples rnase data and model coordinates for a second protein ribonuclease from Streptomyces aureofaciens used in the examples unix non runnable scripts which can t be run for want of data runnable scripts which can be run with the data in toxd tutorial contains procedural scripts in five main areas namely MR MIR refinement MAD and density modification html library and program documentation in HTML form include setup scripts and the def files needed for program startup 83 7 1ib The default destination for installation of the binary library files ccif source code for the CCIF library data machine independent library files such as the symmetry operators src source code for the CCP4 library 65 66 Installation mmdb source code for the MMDB library clipper source code for pirate and the clipper libraries ssm so
162. ource tree The script dupt ree contains further details With a configured system you can now actually build the programs Still in the CCP4 directory just type make and leave it to build When you re satisfied with the build see 11 3 6 for information on testing you can use make install to copy things to the appointed places You can follow this by make realclean to tidy up by deleting the files you ve copied from where they were built unless you want to do tests before overwriting an existing installation for instance Alternatively make clean empty targets will leave dummy copies of the binaries so that if the sources get updated make will only rebuild the relevant ones The suite should now be functional and you can start doing some useful work with it Although the program documenation is in plain text and HTML form when building the Suite man pages are also generated If you want to run tests before installing the suite you have to put the compiled pro grams on your path do with csh set path builddir src path where builddir is the directory in which you ran configure CCP4 in the case described above You ll find a small example dataset in CCP4 examples toxd and a second one in CCP4 examples rnase and some runnable scripts in CCP4 examples unix runnable Some log files generated at Daresbury are distributed for comparison 3You could use make install to start with if you re confident and th
163. out the four Sylvie and Bruno Concluded This is a vital part of your experiment Don t be afraid to repeat it until you are satisfied with the results Extensive x Loggraph facilities are available from the programs to monitor the results More information on data collection processing and reduction can be found in the CCP4 Study Weekend proceedings Helliwell et al 1987 Sawyer et al 1993 By data processing we mean that you e index the image correctly e assign profiles e refine orientation etc to get the best set of hkl I OI and fractionality where the hkl are the measured indices Relevant programs are provided for example in the mosf1m system see also the tutorial information included with it Data reduction is the next stage We need to use the processed set of reflexions and to produce the optimal set of hkl F and oF where the hkl lie in the chosen unique part of the reciprocal lattice The steps involved are 1 Apply geometric corrections which are independent of data collection device These are Lorentz and polarisation corrections Try to apply absorption corrections often not easy but important 2 Find symmetry equivalents of measured indices Steps 1 and 2 may be done by the data processing program or may be in a separate program e g absurd for madnes 3 If you use a data processing program which doesn t output MTZ files e g denzo or XDS use combat to produce a multi record MTZ file f
164. pecified Otherwise if a file isn t given for the logical name the program will attempt to open a file in the current directory with a name corresponding to the logical name Some common logical names with default extensions are HKLIN input MTZ reflection file mtz HKLOUT output MTZ reflection file mt z MAPIN input map file map MAPOUT output map file map XYZIN input coordinate file brk XYZOUT output coordinate file brk In some cases more than one file corresponding to the same type of entry in SCINCL default de 1s required e g several input reflexion files In this case they are typically specified as HKLIN n for instance where n is a sequence number The rules for match ing logical names with the default specifications given above then apply to the pre fix of the name which matches the entries in the files SCINCL environ def and SCINCL default def In Unix the environment variable CCP4_OPEN can be used to determine whether logical names opened as new over write existing files with the name specified Over writing will occur if the value of the variable is UNKNOWN otherwise the program will abort For the fine print of the logical name mechanism see 83 7 3 2 2 Keyworded input Most programs take keyworded input to set their parameters This is read on the standard input in the form of records with a leading keyword followed possibly by arguments which might be numbers or strings or k
165. porting the dlopen mechanism at least should be able to use the library shared too See 11 6 for ways to change the configuration parameters or use parameters to make e g make FOPTIM O In extremis you can probably save some space by removing all debugging information from the binaries using the strip program or installing using install s This will prevent diagnosing some problems though 11 3 9 Installing updates If you subsequently get updates as a complete distribution copy it into the right places in CCP4 as per 11 3 2 and use make install again you probably don t need to re edit ccp4 setup but it s worth looking at the new version to make sure Re run configure although this may not be strictly necessary If bug fix updates arrive as a patch file you should copy this into CCP4_MASTER and use the patch program to apply them A helpful tutorial on using pat ch is in the documentation for version 2 0 and later of the GNU diff implementation See also http www ccp4 ac uk problems patches html To build individual programs the interesting ones you know have changed run make name in CCP4 src where name is the program name e g fft When you have installed the changes to your satisfaction a make clean will remove the orig files left behind by patch If you kept the directory intact after running mak mpty targets a simple make will just rebuild the necessary Before installing a new major version it is possi
166. r code is a call to the routine CCPFYP as the first executable statement in your program This routine performs all of the pre processing It should be called before any other CCP4 library routines 14 6 Library modules Here is a brief list of the modules in the CCP4 library Documentation on them is available mostly as doc files in the doc directory ecplib contains various utility routines which are potentially machine dependent It is built on VMS specific code in vms for and vmsdiskio for and Unix specific code in unix m4 and library c derived from the literate program file library nw diskio contains routines for random access to stream mode files but most of the relevant code is actually in library c The VMS version is in vmsdiskio for fftlib crystallographic FFT routines keyparse a higher level interface to the parser routines maplib for handling CCP4 style map data modlib fairly random collection of mainly numerical analysis routines mtzlib reflexion file handling parser processing free format input containing keywords it actually lexes more than parses The CCP4 manual 84 Writing and contributing programs plot84lib low level graphics with PLOT84 metafiles plotsubs higher level interface to PLOT84 or possibly other routines rwbrook for handling coordinate PDB Brookhaven files symlib useful routines for handling symmetry operations The CCP4 manual Chapter 15 Porting CCP4 CCP
167. r the reciprocal space treatment in mlphare It is not known to have been tried with anomalous MAD data but should work See Tickle 1991 In this scheme the function to be minimised is a difference between the observed and calculated Fu It was first proposed Rossmann 1960 as a procedure to refine heavy atoms parameters for two derivatives A modified version of Fyre refinement has been proposed Terwilliger amp Eisenberg 1983 This was implemented in the program heavy but has been superseded by the maximum likelihood technique of mlphare 6 6 Phase determination 6 6 1 Isomorphous replacement data From the vector triangle Fpy Fp Fy knowledge of the amplitudes Fpy Fp Fu the phase Oy and using the cosine law p can be derived Op Oy cos Fpy Fp Fy 2FpFu onto 6 16 To obtain an unambiguous estimate of amp p we need to repeat the phase determination using another vector Fy not collinear with the first one since the two solutions are symmetric with respect to Oy this can be done by using another derivative with heavy atoms substituted at different sites Harker 1956 gave an elegant geometrical solution to the problem of phase deter mination he showed that if a phase circle of radius Fp is drawn from the origin of the Argand plane and another circle of radius Fpy is drawn from the end of the vector Fy their points of intersection will give the two solutions for dp With a second der
168. ram some of its features are now in npo pltdev xplot84driver xccpjiffy2idraw Convert PLOT84 metafiles to X Windows Tektronix HPGL or PostScript format xccpjiffy2idrawcon verts the result to PostScript which can be edited with idraw prepi a molecular graphics program which can be used to interac tively manipulate many different types of molecular representations http bonsai lif icnet uk people suhail prepi html proplot Assorted pretty plots from the procheck suite q v includes Ramachan dran plot RasMol X Windows and MS Windows visualiser for PDB files raster3D a set of tools for generating high quality raster images of proteins or other molecules including rendering pictures composed in molscript ribbon John Priestle s package Priestle 1988 for Jane Richardson type cartoon drawings or graphics displays using frodo Associated programs ribrot postplot pltout pdbrot splitd Aggregated setor hardware lighted 3 dimensional solid model representations of macro molecules topdraw Sketchpad for creating protein topology diagrams xloggraph Plots graphs of tables from many CCP4 programs log files under X Windows axissearch Changes axis and cell See also tracer Babel Interconverts many coordinate formats cad Combine assorted data and sort a number of reflection files with various possi ble operations on the data items Apart from manipulating the values data may be converted from one area
169. replaced by NaN or Rop As a safety feature only columns which are explicitly specified in the input to mt zmn f are converted Old style MTZ files may still be used with all CCP4 programs and old style checks on missing data remain in place occurring after the check for a NaN or Rop However new data sets completed with CETC uniqueify and combined with cad should automatically include the necessary NaN and Rop entries The electron density map format was devised by Phil Evans It can also be used for envelopes and images Maps are stored in a randomly accessible binary file as a 3 dimensional array preceded by a header which contains all the necessary information about the map The header is organised as 56 words followed by space for ten 80 char acter text labels as in fig 16 2 Maps are structured as a number of sections each con taining a fixed number of rows and each row contains a fixed number of columns The standard format adopted for coordinate data is that used in the Brookhaven Protein Data Bank The programs of the suite will handle either complete files or files con taining only a subset of the types of record which may be present in a complete file In particular the records containing the coordinate data ATOM or HETATM records are of interest These PDB file records have the structure shown in fig 16 3 The Protein Data Bank provides a full description of the complete format see 13 4 A typical Fortran format
170. rier based on the phases of the first derivative will give the sites of the subsequent derivatives on the same relative origin and hand Matthews 1966 Various correlation functions working in Patterson space have also been used Perutz 1956 proposed two correlation functions later improved and modified Blow 1958 Rossmann 1961 suggested a method based on the calculation of difference Pattersons between pairs of derivatives with coefficients Feu Fou This synthesis will give positive peaks for vectors between atoms in the same derivative and negative peaks for the correlation vector between atoms in different isomorphs More elaborate correlation functions have been proposed Kartha amp Partharasarathy 1965 6 5 Refinement of heavy atom parameters After rough heavy atom parameters have been found they must be refined to their best values before being used for phase determination An important aim of the refinement is to discover all the possible minor sites The techniques are similar to the ones used to refine small molecule structures but there are more difficulties due to the fact that especially for acentric reflexions the estimate of Fy can be relatively poor and errors due to lack of isomorphism and scaling cause the data to be less accurate Usually since the ratio observations parameters to be refined is high it is wise to reject any suspect reflexions as long as those remaining are uniformly distributed through s
171. rns out to be the more difficult part of the problem The most likely reason for this is the errors introduced by the inaccurate orientation of the model To get the translation search working well it is important that the rotational parameters are optimised It is worth running translation searches for slightly different rotations The two main approaches to solving the translation problem i e the R factor search and translation functions are described next This is the conceptually simplest used method for determining the translational pa rameters It involves testing the agreement between the data and the structure factors calculated from a model placed at various positions in the unit cell It is apparent that this is a computationally large task It would seem that the calculation should involve looping over all atoms for each calculated structure factor for each position in the unit cell The problem however is not quite so bad and we shall now see that there are some simplifying features Starting with model coordinates of the molecule in the correct orientation but in an arbitrary position we can calculate the structure factors in a Pl cell with the same dimensions as in the crystal as Eh Dres 14 where the x are the coordinates of the i atom f is the atomic scattering factor and h is the reciprocal lattice vector The position of the i atom with respect to the crystallo graphic origin will be x p where p is the translat
172. roject Number 4 1994 given as COLLABORATIVE COMPUTATIONAL PROJECT NUMBER 4 1994 The CCP4 Suite Programs for Protein Crystallography Acta Cryst D50 760 763 Such citations may be valuable to us in the future in maintaining funding In addition authors of specific programs should be referenced where applicable see the program s documentation and or output The CCP4 manual Part II Tutorial material Chapter 2 Overview of the crystallographic process and CCP4 programs What is it you want to buy the Sheep said at last looking up for a moment from her knitting I don t quite know yet Alice said very gently I should like to look all round me first if I might Through the Looking Glass 2 1 Steps in structure solution The CCP4 manual The basis of any crystallographic project is well diffracting crystals which allow the diffraction pattern to be measured accurately This can be achieved in a variety of ways depending on the type of structure we are attempting to solve the availability of equipment and on the personal preferences and prejudices of the investigator The basic choice in data collection is between diffractometers and area detectors including photographic film and image plates Area detectors are usually more suit able for proteins Most machines provide their own software for data collecting The raw data are integrated and suitable corrections are made The relevant so
173. rom a file of un merged reflexions or 2mt z for programs which produce a plain text file of merged reflexions Some systems can output both merged and un merged data so be careful You may have to extend combat for formats it doesn t already support 4 Sort merge the data so all equivalent sets of indices are adjacent using sortmt z 5 Use reliable equivalent measurements from different sets to calculate relative scale factors Check the scales make physical sense 6 Merge equivalent reflexions applying scale factors 7 Examine the data carefully for outliers Try to correct or at least understand as far as possible 8 Analyse for better estimates of standard deviations testing data and processing guestimates against observation scatter Correct guestimates as far as possible Steps 5 to 8 are done with scala which will need to be run several times 9 Analyse the final data for sensible crystallographic behaviour Using t runcate cumulative intensity distribution Wilson plot e g with wilson 23 24 Data processing and reduction 10 Run uniquify script to complete your dataset and add freeR column This should be done for your native dataset See program unique for more information To merge data from different sources you need to produce multi record files one from each source as step 3 above and go to step 4 iff the files at this stage have the same number and type of columns sortmtz will merge an arbitrary
174. roximate for mula is AFmo Foy Fou gt 2k Ey sin pH ag 6 6 The CCP4 manual 6 4 3 Determination of the heavy atom positions 33 where k is defined as Fy Fyj and can be obtained from the International Tables although it is more usual to substitute an empirical value derived from the data see below Although both isomorphous and anomalous data are inaccurate they give comple mentary information and used together can give a better estimate for Fy The problem arises of how isomorphous and anomalous data may be put together to give an accu rate estimate of Fy The simplest way is to sum the contributions with the appropriate scaling AF2 K 4 AF gt Fy cos Opu 04 F sin apy On Fy 6 7 A more precise equivalent expression can be used to combine the isomorphous and anomalous data Matthews 1966 Singh amp Ramaseshan 1966 FR Fe FR 2 RAR k 4 2 an 6 8 where F4 Ur F and A Er re and we assume 1 k lt 1 As the isomorphous differences are only small only the solution with the negative sign is normally considered yielding Fyre Heavy atom Lower Estimate 1 2 Fae Fy t Fe 2 Far k 4 Ar 6 9 k the ratio of the real and imaginary parts of the heavy atom structure factor can be approximated from the data Matthews 1966 k 2 Fon Fol Fon val 2AFeo AFano 6 10 The anomalous differences are usually overestimated and thi
175. rs syminfo lib Contains symmetry operators for the 230 spacegroups listed in and following the con ventions of Hahn 1992 IT and non standard settings used by various CCP4 pro grams Each space group has a header line comprising space separated values of Spacegroup number Number of lines of symmetry equivalents positions in IT n Number of lines of primitive equivalents p Spacegroup short name subscripts are typed as is and a prefix represents an overbar e g P21 m P2 m P 1 PI Point group name the IT name is prefixed by PG contrary to the spacegroup name convention an overbar is represented by a trailing bar e g PG4bar3m e Crystal system e Possible comments about non standard settings etc The CCP4 manual 94 File formats Following the header are n lines of symmetry equivalents of which the first p are the primitive ones Layout is such that e The symmetry operator lines are limited to 80 characters e The elements of operator triplets are separated by commas and triplets are sep arated by or newline the translations may be before or after the coordinate e g 1 2 X or X 1 2 e The header lines should start in the first column and the other lines be indented for ease of locating the headers The CCP4 manual Part V Appendices References The CCP4 manual A T Briinger 1995 The free R Value A More Objective Statistic for Crystallography Methods in Enzym
176. s causes an overesti mation of Fyre for acentric reflexions French Dodson et al 1975 suggested a for mula for the bias for Fygi and proposed the use of Faki bias Fux1 to calculate Pattersons Isomorphous and anomalous data as well as any prior information can be com bined correctly in a maximum likelihood approach Bricogne 1991 mlphare pro vides a rough approximation to the correct approach 6 4 3 Determination of the heavy atom positions The estimated Fys are used to find the heavy atom positions These coordinates will then be used to calculate the vector Fy required to find the protein phases 6 4 3 1 Patterson methods The most general method that does not require any previous information is the use of Patterson techniques An isomorphous difference Patterson synthesis is a Fourier summation of the terms Fpu Fp gt Fi cos Opy On 6 11 In the case of centric reflexions Fpy Fp really corresponds to Fy and a Patterson with Fey Fp coefficients is a good estimate of a FR Patterson the only noise is due to the few cases of crossing over and to measurement errors However erroneously large differences and the lack of centric data can reduce the quality of the Patterson For acentric reflexions we can use cos 1 2cos 20 2 6 12 to show that the summation has two terms Fy 2 and Fy cos 2 py 01 The first term will generate vectors between the heavy atoms at ha
177. somorphous or anomalous differences Mukherjee et al 1989 6 4 3 4 Difference Fouriers When a previous estimate of the protein phases exist for example when we are dealing with the n derivative or looking for further sites in the current derivative a Fourier synthesis with coefficients m Fpy Fp e where m is the weight associated with this phase is the simplest way to locate the heavy atoms This method usually works even when the protein phases are very inaccurate The biggest problem with this method is the fact that these Fourier syntheses will often have ghost peaks or halos at or near the heavy atom sites used for phasing These can be erroneously interpreted as minor sites It is sensible to check any sites against the derivative Pattersons 6 4 3 5 Choice of origin and hand Any Patterson function will be equally well satisfied by coordinates x y z or x y Z This is true for direct methods techniques as well In other words the hand of the solution will not be determined In addition in many spacegroups there is an arbitrary origin choice and it is essential that all derivatives use the same origin and are on the same hand The first choice is in fact arbitrary but once fixed for example assigning The CCP4 manual 6 5 Refinement of heavy atom parameters 35 the coordinates to an heavy atom in one isomorph must be used consistently for all subsequent derivatives The use of a difference Fou
178. space sampling cannot easily be utilised for refinement 9 7 Validation gross and overall errors 9 7 1 Validation An lot of validation requires common sense For example a good structure will not have most of its torsion angles in strange parts of the Ramachandran plot Other obvious but very useful checks include Are there unacceptable symmetry contacts between adjacent molecules e Are the B factors sensible e g higher at the surface than in the core not wildly divergent between adjacent atoms Does the chemistry make sense e g do the H bondable groups actually make H bonds Are there charged groups buried in hydrophobic environments e Do the maps show the expected features e g do omit maps reveal the missing atoms do difference maps show substrate atoms e Is there a suspicious divergence from NCS Most programs flag many of the above 9 7 2 Errors Serious mistracing is rare but can happen The existing tools Rfree stereochemical checks as applied in procheck and what if easily detect such errors if ap plied sensibly Actually the Ramachandran plot alone is a powerful tool for cross validation to identify such gross errors since the dihedral angles are not usually used as restraints in refinement programs Local errors such as loops out of register can easily be overlooked or ignored They can be identified by the real space R factor and B values as a function of residue number Overall imprecision and refineme
179. t require N to be a power of 2 but in general N has to be a nice number No prime factors greater than 19 are permitted Additional requirements have to be satisfied for specific space groups In crystallographic applications the FFT is additionally optimised to take advan tage of crystallographic symmetry The application of the FFT in crystallography is explained in full by Ten Eyck 1973 1977 1985 Spacegroup specific transforms use symmetry to speed up the calculation but are more likely to have bugs especially in uncommon spacegroups Use of Pl is thus recommended if speed isn t a problem if it is test the higher symmetry version against Pl The CCP4 suite contains several programs making use of the FFT e g fft based on Ten Eyck s original calculates Fouriers difference Fouriers Pattersons and difference Pattersons from reflection data sfall calculates structure factors and X ray gradi ents for refinement using inverse and forward FFTs almn calculates rotation function overlap values using FFT techniques etc fftbig is an in core version of f ft which runs in P1 only It may be faster you have enough real memory available otherwise it will cause page thrashing This has replaced the original fft program The CCP4 manual Part III The propagation care and feeding of a CCP4 installation Chapter 11 Installation 11 1 First off README file n By convention the top level directory of a UNIX source d
180. tage of loose organisation is that it is very easy to add new programs or to modify existing ones without upsetting other parts of the suite This is the approach successfully taken by Unix Converting a program to use the standard CCP4 file for mats is generally straightforward and the philosophy of the collection has been to be inclusive so that several programs may be available to do the same task The compo nents of the whole system are then a collection of programs using a standard subroutine library to access standard format files Most of the programs are written in standard FORTRAN77 To use the programs the user must assign input and output files including library and scratch files where necessary though defaults are usually defined and run the programs Often an output file becomes the input to the next step and system parameter substitution may be used to create filenames in a systematic way Most crystallographic calculations involve a series of steps in which no decisions need be made until the end 2 Introduction and a command file provides an easy way of chaining calculations A graphical user interface is now also provided as a way of facilitating running of the programs Standard file formats are defined for the principal sorts of data used in crystallog raphy reflection data electron density maps and atom coordinates In defining these formats a number of trade offs have to be made between efficiency in space and ac c
181. te files pdbset 11 25 39 40 Various useful manipulations on coordinate files peakmax 10 11 29 30 Search for peaks in the electron density map phaser 65 phaser 1 3 Maximum likelihood molecular replacement phaser MR 8 phistats 12 Analysis of agreement between phase sets and checking it against weighting factors pltdev 12 27 93 Convert PLOT84 metafiles to printer speak pmf PMF pattersons search polarrfn 9 39 Fast rotation function in polar coordinates polypose 11 Superposition of multi domain structures postref 7 Unsupported Post refinement of film data procheck 11 12 15 26 56 Checking stereochemistry Laskowski et al 1993 Comprises nb proclean anglen secstr mplot tplot pplot probplot procheck_comp Version of procheck for related structures Uses rmsdev in addition professs 8 Determination of NCS operators from heavy atoms proplot 12 Plotting and listing for procheck Uses pplot tplot and probplot rantan 9 Direct Method module for the determination of heavy atom positions in a macro molecule structure or to determine a small molecule structure RasMol 12 rdent 10 Create dictionary entries for rest rain from PDB file rebatch 7 Alters batch numbers in an unmerged MTZ file saves running mosf1m again refmac 53 55 refmac5 10 26 53 54 Refine or idealize structures using intensity or ampli tude based least squares or loglikelihood residuals reform Unsupported Various on
182. tion dmmulti 9 A multi crystal version of DM dtrek2mtz 7 Converts d trek scalemerge output into MTZ format dtrek2scala 7 Convert integrated intensity and header files from d trek into multi record MTZ files suitable for scala dyndom 11 Determines dynamic domains when two conformations are available ecalc 8 39 Calculates normalised structure amplitudes extend 55 Unsupported Extends the map to cover the specified volume of the unit cell extends 10 f2mtz 7 12 23 25 55 Converts a formatted reflection file to MTZ file fffear 10 Fast Fourier Feature Recognition for density fitting ffjoin Joining model fragments from FFFEAR fft 9 29 30 40 54 55 61 68 In memory crystallographic fast Fourier trans formation Pl only fftbig 6l fhscal 8 29 32 Scaling derivative to native by Kraut s method findncs 8 Detect NCS operations automatically from heavy atom sites freerflag 12 55 Tag reflexions in MTZ file with free R flags fsearch 8 6 d molecular replacement envelope search gensym 11 Generate sites by symmetry geomcalc 11 Does various geometry calculations on a molecule getax 9 Real space searching for rotation axis of a D n or C n multimer havecs 8 27 29 Unsupported Generates Patterson vectors from atom coor dinates more or less replaced by vectors and gensym hbond Unsupported Calculates possible main chain H bonds helixang Unsupported Distances and angles between helices hgen 11 Generate hydrogen ato
183. tion of Phase Probability Distributions for Simplified Combination of Independent Phase Informaton Acta Cryst B26 136 143 Jones Geraint 1989 Deriving the fast Fourier algorithm by calculation Technical report TR 4 89 Oxford University Programming Research Group Jones T A Zou J Y Cowan S W amp Kjeldgaard M 1991 Improved Methods for Building Protein Models in Electron Density Maps and the Location of Errors in these Models Acta Cryst A47 110 119 Kabsch W 1976 A Solution for the Best Rotation to Relate Two Sets of Vectors Acta Cryst A32 922 923 Kabsch W 1988 Evaluation of Single Crystal X ray Diffraction Data from a Position Sensitive Detector J Appl Cryst 21 916 924 Kabsch W amp Sander C 1983 Dictionary of protein secondary structure pat tern recognition of hydrogen bonded and geometrical features Biopolymers 22 2577 2637 The CCP4 manual 99 Kartha G amp Partharasarathy R 1965 Combination of Multiple Isomorphous Re placement and Anomalous Dispersion Data for Protein Structure Determination I Determination of heavy Atom Positions in Protein Derivatives Acta Cryst 18 745 Kendrew J C Bodo G Dintzis H M Parrish R G Wyckoff H amp Phillips D C 1958 A Three dimensional Model of the Myoglobin Molecule Obtained by X ray Analysis Nature London 181 662 666 Kleywegt G J amp Jones T A 1996a Acta Cryst D52 826 828 Kleywegt G J
184. tions which are made include e Hierarchical filing system with file names of the form foo bar and possibly rather long e ASCII character set e 32 bit words 8 bit bytes not middle endian It is definitely assumed that the system has environment variables symbols or logi cal names in VMS DCL that can be read and set from programs although the library could implement its own dictionary mechanism instead and that programs can inter rogate their command line 15 3 Unix flavour dependencies The CCP4 manual Most of the flavour and compiler specific stuff should be confined to the low level routines defined in library cl binsortint c unix m4 and the configuration variables set in the configure script For a new port the first thing to do try is to run configure with argument generic to generate Makefi les etc with all the defaults and see what you need to change One thing you will need to find out is what your compiler s conventions are for calling C from Fortran and make the appropriate edits to library c and binsortint c If the number representation of your ar chitecture isn t picked up correctly by the library c code you may need to alter that If you re lucky this will be the same as one of the existing conventions and you won t have to do much editing You may be able to guess other configuration param eters from existing ports or by trying first with configure generic Edit a new This is not meant to
185. tored in multi record files in which each reflexion occupies sev eral records which are distinguished by different batch numbers symmetry numbers etc In this case the fourth column is the m isym number recording the symmetry and whether the reflexion is fully recorded and the fifth column is the batch num ber Multi record files have fixed column names which are understood by the relevant programs The files also contain various other information in a header block This includes cell parameters symmetry column limits sort order title and history information For more details see 16 1 See also 16 for the standard coordinate and map file formats 3 6 Transporting data files Reflexion map and PLOT84 files are binary and are not generally portable between machines with different representations of integer and or real numbers However re flexion and map but not PLOT84 files written using the current CCP4 release are readable on all machines from the set we recognise They are written using the ma chine s native number formats and can be read back on the same architecture without overhead On a different architecture the necessary conversions between foreign and native formats will be done with some overhead which is not likely to be noticeable on modern machines This facility relies on a machine stamp in the file header to identify the architec ture on which the file was written If this doesn t exist the library
186. urce code for superpose and ssm libaries fftw source code for the MIT fftw libaries man contains soft links to enable man command manual IAIEX source for this manual and corresponding PostScript file src program source unsupported source for the unsupported stuff x windows X Windows programs xloggraph etc The ccp4 directory itself contains top level configuration files etc 11 3 Building under Unix 11 3 1 Preliminaries First decide where you want to put the sources for the software this will be a direc tory we ll refer to as CCP4_MASTER which will contain the ccp4 directory A conventional possibility for CCP4_MASTER is public xtal but somewhere in usr local is more in line with Unix conventions 11 3 2 Unpacking the files If you got the distribution by ftp or via the web site as gzip d or compressed tar files you can do the following gunzip file or uncompress file tar xvf file where file is replaced by the name of each tar gz or tar Z file in turn This may be performed automatically by the supplied install sh If you got the distribution on cd rom cd into CCP4_MASTER and read the cd rom using tar tar xvf device where device is the appropriate cd drive e g dev cdromand the v option lists the file names as they come off for comfort only An additional option like o is necessary on some systems e g IRIX to prevent the files original ownership being retained If you don
187. ver all space R C f P x U x P2 Cx dV 7 2 11 space where U x is defined equal to one inside the sphere of integration and zero elsewhere In reciprocal space this becomes R C Y Iops h Fu2 Ch 7 3 h where the ops h are the observed intensities and Fiy2 Ch the Fourier transform of U x P2 Cx is a continuous function defined over all reciprocal space This formula tion first used by Lattman 1970 is conceptually clear and flexible in application but unfortunately slow to compute The usual way to compute the rotation function is by the fast Fourier method Crowther 1972 It uses the idea that since we are rotating spherical volumes it should be more appropriate to expand the Patterson density in a sphere in terms of Bessel spherical harmonics The following are some guidelines and practical points to consider when calculat ing and interpreting the rotation function Firstly one should try to use the best search model available Refined coordinates should be used whenever possible The importance of this may not seem obvious especially as the data we use in the calculations do not usually extend beyond 3 0 3 5 In practice however the difference may be between a clear solution and no solution at all In general one should also remove from the model any atoms that are known to differ from the target structure For example side chains that are known from sequence alignment to be different should be trimmed
188. w in map Number of first section in map Number of intervals along x Number of intervals along y Number of intervals along Z Cell dimensions A Cell angles degrees Which axis corresponds to cols 1 2 3 for X Y Z Which axis corresponds to rows 1 2 3 for X Y Z Which axis corresponds to sects 1 2 3 for X Y Z Minimum density value Maximum density value Mean density value Average Space group number Number of bytes used for storing symmetry operators Flag for skew transformation 0 none 1 if foll Skew matrix S in order 11 S12 13 S21 etc if LSKFLGSO Skew translation t if LSKFLG0 Skew transformation is from standard orthogonal coordinate frame as used for atoms to orthog onal map frame as Xy map S Xo atoms t Some of these are used by the MSUBSX routines in mapbrick mapcont and frodo All set to zero by default Character string MAP to identify file type machine type stamp as for MTZ files Rms deviation of map from mean density Number of labels being used 10 80 character text labels i e A4 format Symmetry records if any follow stored as text as in International Tables operators separated by x and grouped into lines of 80 characters i e symmetry operators do not cross the ends of the 80 character lines and the lines do not terminate in a Binary data follow as a byte stream Figure 16 2 Map file format The CCP4 manual 16 4 PLO
189. wick M 1987 Improving Crystallographic Macro molecular Images the Real space Approach Ann Rev Biophys Biophys Chem 16 351 373 Press W H Flannery B P Teukolsky S A amp Vetterling W T 1986 Numerical Recipes The Art of Scientific Computing Cambridge Scientific Press Priestle John P 1988 RIBBON a stereo cartoon drawing program for proteins J Appl Cryst 21 572 576 Ralph A C amp Woolfson M M 1991 On the Application of One Wavelength Anomalous Scattering III The Wilson and MPS Method Acta Cryst A47 533 537 Raymond Eric ed 1993 The New Hacker s Dictionary Second edn MIT Press Read R J 1986 Improved Fourier Coefficients for Maps Using Phases from Partial Structures with Errors Acta Cryst A42 140 149 Read R J 1991 Dealing with imperfect isomorphism in multiple isomorphous re placement In Wolf et al 1991 Rhodes Gale 1993 Crystallography Made Crystal Clear A Guide for Users of Macromolecular Models San Diego London Academic Press Rossmann M G 1960 The Accurate Determination of the Position and Shape of Heavy Atom Replacement Groups in Proteins Acta Cryst 13 221 Rossmann M G 1961 The Position of Anomalous Scatterers in Protein Crystals Acta Cryst 14 383 388 Rossmann M G Arnold E amp Vriend G 1986 Comparison of Vector Search and Feedback Methods for Finding Heavy Atom Sites in Isomorphous Derivatives Acta Cryst A42 325 334
190. with gt location of different ranges of non crystallographic POLARRFN integration sphere symmetry axes and resolution limits 2 Crowther s cross rotation function Model brk AMORE gt a f y rotation angles co ordinates of model structure also for all symmetry equivalents gt ECALC Native Fs mtz 3 Translation function Model brk Rotated using gt AMORE gt a y rotation angles co ordinates of rotation matrix TFFC model structure from 2 x y z translation vector Native Fs mtz gt ECALC Figure 7 1 Molecular Replacement The main steps in obtaining initial phases by Molecular Replacement are shown in fig 7 1 The principal program is amore Navaza 1994 which is a state of the art molecular replacement program incorporating the rotation and translation steps plus rigid body fitting within one program 7 1 1 Detection of non crystallographic symmetry using self rotation function amore or polarrfn yields the rotation function from native Fs Run trials with dif ferent ranges of the integration sphere and resolution limits to locate non crystallographic symmetry axes amore s rotation function is superior to almn s Navaza 1993 Driessen amp Tickle 1994 7 1 2 Cross rotation function pdbset finds the dimensions of a model molecule from the input PDB file sfall calculates structure factors in a Pl cell for the model if necessary not neces sarily for
191. xtensions 21 library 16 file formats 1 2 73 files connexion 15 109 110 input and output 15 pre release 2 2 20 film 5 7 Fortran 2 65 81 82 85 FORTRAN77 1 81 85 Fortran90 81 non portable features 82 standard 81 fractional coordinates 26 fractional sites 8 free R factor 55 density modification 49 free software 2 59 77 Free Software Foundation 77 ftp 66 73 75 76 83 sites 77 funding 1 C F Gau 59 geometric corrections 23 GNU 68 GNU project 68 77 Graphical user interface 2 graphical user interface 21 graphics editing 12 graphics metafile 93 heavy atom positions 5 31 Hendrickson Lattman coefficients 38 90 Henrick Kim vii histogram matching 9 47 49 104 homologous protein 5 Howard Maeri 1 HPGL 12 image plate 5 7 initialisation program 81 input keyworded 15 IRIX 66 70 isomorphous differences 8 Jane Richardson type cartoons 12 Kabsch algorithm 26 Keegan Ronan 1 keyworded input 15 16 keywords 82 83 kluges 1 113 koan 113 Kraut scaling 32 Laue method 7 Leslie Andrew vii library routines 81 Linux 2 The CCP4 manual Program index literate programming 83 log file 15 19 74 logical names 15 20 82 85 common 16 library 16 login procedure 15 Love Dave vii 113 Lp correction 23 machine stamp 19 87 MAD 9 36 37 49 map files 9 19 maps calculation 9 extension 9 masks 92 Matthews nu
192. ymmetric unit Selects peaks from peakmax and puts them close to the respective Display PLOT84 plots under X Windows Unsupported Sets part of the map to zero diff 68 dssp 11 76 dupt ree 67 dvips 113 2c 70 77 frodo 12 30 92 ftnchek 82 83 ftp 75 977 78 gcc 70 77 81 gunzip 69 heavy 36 html 65 idraw 12 idxref 88 install 70 ksh 18 66 1d 70 in 70 lynx 82 m4 70 madnes 7 23 76 madsys 9 37 76 makeindex 113 make 67 68 man 66 67 82 109 mapbrick 92 mapcont 92 merlot 40 76 molscript 11 12 27 76 mosflm 7 23 24 76 multan 30 noweb 85 ortep2 12 oscgen 89 p2c 81 pacman 40 patch 68 patsee 30 pdbrot 12 108 pltout 12 postplot 12 prepi 12 ranlib 70 raster3D 11 12 76 ribbon 12 27 113 ribrot 12 setor 12 sftools 77 sh 18 19 66 The CCP4 manual Program index splitd 12 strip 68 tar 66 67 tcsh 19 66 weis 76 what if 56 wulff ps 12 xentronics 7 xfig 113 Index The CCP4 manual Yes this index is incomplete and untidy man 67 ab initio phases 8 absorption corrections 23 abstraction 81 Adams Douglas 59 aggregated software 7 12 66 Alice in Wonderland vii 77 analysis of coordinates 6 anomalous differences 90 anomalous occupancy 36 anomalous scattering 8 103 anonymous ftp see ftp archaeology 104 area detector 5 7 asymmetric unit changing 8 atomicity 48 averag

The CCP4 Suite

Contents

Download Pdf Manuals

Related Search

Related Contents