Home

User Manual for QTL Cartographer

1. It is 10 39 44 on Wednesday 25 March 1998 The position is from the left telomere on the chromosome window 10 00 Window size for models 5 and 6 background 5 Background parameters in model 6 Model 6 Model number trait 1 Analyzed trait Trait_1 cross B2 Cross Test Site Like Ratio Test Statistics 7 Additive c m position HO H1 R2 0 1 TR2 0 1 Hl a s1 S 1 0 0001 0 411 0 002 0 473 0 027 11531 2 0 0133 0 016 0 000 0 472 0 005 1 542 2 0 0333 0 023 0 000 0 472 0 006 1 547 2 0 0533 0 031 0 000 0 472 0 008 1 554 2 0 0733 0 041 0 000 0 472 0 009 1 563 2 0 0933 0 052 0 000 0 472 0 010 1 572 2 Oah 0 063 0 000 0 472 0 011 1 582 Ze 001333 0 073 0 000 0 472 0 012 1 593 e For a backcross let a be the additive effect We have two hypotheses e Ho no QTL effect at the test position i e a 0 e H There is a QTL effect at the test position i e a 4 0 The first eight columns correspond to 1 Chromosome of test position 2 Left flanking marker of test position 3 Absolute position from left telomere in Morgans 4 Likelihood ratio test statistic for He Itis a x random variable with one degree of freedom for any position meaning that a value of 3 84 or higher is evidence for a QTL The significance level over more positions will be higher due to multiple test ing 2 ri Estimate of a the additive effect under H OOF LT Ay 91 Test statistic S for the normality of the residuals
2. At the moment this number 1 is ignored With the convention that A alleles originated from the P line and 42 from the Pz marker genotypes will be encoded with the following integer values e 2 for A Aj e 1 for A Ao e 0 for 4243 e 12 for A that is individuals with at least one dominant A allele e 10 for A2 that is individuals with at least one dominant Ag allele 41 June 22 2000 CHAPTER 2 SIMULATING REFORMATTING DATA e 1 for unknown genotypes Rcross read something but could not translate it e 2 is also for unknown genotypes In this case no data had been read in The trait values follow the marker genotypes and finally the other categorical or qual itative traits follow at the end The sequence repeats for all individuals Note that there is a permissible range for trait values By default all trait values must be real numbers with absolute value less than one million 10 Any trait value that is less than negative one million is treated as a missing phenotype by the programs Other Traits Other traits can be thought of as qualitative or categorical traits Examples include sex brood plot etc In some cases these factors will have been regressed out that is a regres sion of the quantitative trait of interest on the categorical trait will have been performed and the residuals used as the phenotypes in the analysis Presently they can be input via a file of filetype cross inp
3. Analysis A recent review Doerge Zeng and Weir 1997 summarizes the statistical issues for map ping QTLs It is the best place to start for a general overview of the analytical methods used in QTL Cartographer Figure 3 1 shows a schematic of the analysis procedure There are five programs in this step Qstats does some basic quantitative genetic statistics and summarizes missing data It is a useful program to run at the beginning of your analysis LRmapqtl does single maker analysis using linear regression It also runs very fast and will give some idea of where QTLs are SRmapqtl does stepwise regression either forward backward or for ward with backward elimination The final program is Zmapqtl which implements in terval mapping Lander and Botstein 1989 and composite interval mapping Zeng 1993 Zeng 1994 This program generally requires more computing power JZmapqtl is a mod ule that implements multitrait mapping Jiang and Zeng 1995 The basic requirements for using these three programs is a genetic linkage map and a data file The linkage map should be of filetype format Rmap out and the data file of Rcross out Whether the files are simulated real or bootstrapped data is irrelevant The analysis is the same regardless of the origin of the data 3 1 Qstats Ostats is a good place to start in analyzing your data It computes some basic statistics on the quantitative traits and summarizes missing data Let y1 y2
4. Ge 2 50 CHAPTER 3 ANALYSIS QTL Cartographer which is distributed as a x with two degrees of freedom The critical values for the rejec tion of normality are 5 99 and 9 21 for tests at the 5 and 9 levels respectively An example of the output follows This is for trait 1 called szfreq Sample GPZ6 A se bathe els a oo 119 MA ed eee ed ead ane ee as 0 4349 MED eit eai etat eee 0 2184 MECS 5 Avie rd Male ia Des 0 1195 MAI etat Re aOR 0 0694 Mean Trait Value 0 4349 Variances ass es LL ES han 0 0295 Standard Deviation 0 1718 Coefficient of Variation 0 3951 Average Deviation 0 1398 SKw LN 24 ance ii 0 0010 dae ts Sant 6 1 sek AS a 0 2245 Beery A to see feet gr 0 0022 seu Sort 24 7 Anean iva ehss 0 4491 Ke TW 2A aoe set nana ns 0 1922 RE A a gs date diras 0 5250 SI DI LS DL ad 2 0992 In the above example LW i refers to a page number in Lynch and Walsh 1998 where one can find an explanation of the quantity The value of the test statistic S is 2 0992 thus one would fail to reject the hypothesis that this trait is normally distributed After the basic statistics Ostats draws a histogram of the quantitative trait It is a simple histogram in that the range of the data are divided into 50 equally sized bins and the num ber of data points falling into each bin are counted and plotted A small table following the histogram gives the sample size
5. 2 3 3 Output The flag g can be used to indicate the output format of Rcross As with the input for mats there are three options for output Rcross will write output in a format suitable for MAPMAKER if the g option is used with the integer 2 while a cross inp formatted file will be written with the value 1 The default is what we term the gtlcart cro format and is indicated by using zero with the g option Here is an example of the output of Rcross 1472574604 filetype Rcross out QTL Cartographer V 1 12c March 1997 n 300 is the sample size 40 CHAPTER 2 SIMULATING REFORMATTING DATA QTL Cartographer p 63 is one more than the number of markers Cross B1 is the type of cross traits 1 is the number of traits Names of the traits 1 Trait 1 otraits 0 is the number of other traits s 1 1 PEE rE A LD 2 AnS DD DED AL D ee A l A A DIS D Le DIL D DD DD D T SDE DAD De Di D Gh De PRED PZ 7 035406650635 1 BOD ID D D BADD Desai 3 555115422473 cl 2 1 BBD DE BOB BD NDE De BD 1 TA 202 2 22 2242 De 2 D D D DD BD BED DAD DA DDR D2 DD DAT CT pa 4 165996548162 E The section prior to the s token is self explanatory The area between the s and the e is the data It starts with an identification number 1 2 3 etc and is followed by a 1
6. Human Population Genetics The Pittsburgh Symposium New York pp 209 228 Van Nostrand Reinhold Kosambi D D 1944 The estimation of map distances from recombination values Ann Eugen 12 172 175 Lander E S and D Botstein 1989 Mapping mendelian factors underlying quantitative traits using rflp linkage maps Genetics 121 185 199 Lander E S P Green J Abrahamson A Barlow M Daley S Lincoln and L Newburg 1987 MAPMAKER An interactive computer package for constructing primary ge netic linkage maps of experimental and natural populations Genomics 1 174 181 Lincoln S M Daly and E S Lander 1992 Constructing genetic maps with MAP MAKER EXP 3 0 Technical report Whitehead Institute Technical Report Liu B 1998 Statistical Genomics Linkage Mapping and QTL Analysis Boca Raton FL CRC PRess LLC Lynch M and B Walsh 1998 Genetics and Analysis of Quantitative Traits Sunderland MA Sinauer Associates Inc Meng X and D B Rubin 1993 Maximum likelihood estimation via the ECM algo rithm A general framework Biometrika 80 267 268 Morgan T H 1994 The Theory of Genes New Haven CN Yale University Press Press W H B P Flannery S A Teukolsky and W T Vetterling 1988 Numerical Recipes in C The Art of Scientific Computing Cambridge UK Cambridge University Press Rao D C B J Keats J M Lalouel N E Morton and S Lee 1979 A maximum likelihood
7. N N ER pe O O 63 June 22 2000 CHAPTER3 ANALYSIS Permutation Test output If you chose to do a permutation test Churchill and Doerge 1994 for the purpose of es timating experiment specific threshold values Zmapqtl will create two auxiliary files to store interim comparisonwise and experimentwise test statistics If the filename stem is qtlcart and the model for analysis is 6 then these files will be qtlcart z6c and qtl cart z6e The former file should look something like this Row Chrom Mark Position Original P Val Count perm 899 start 1 1 1 0 00010 0 00000 0 982202 883 2 Al 1 0 02010 0 00000 0 976641 878 whose columns are 1 Integer indicating the row Chromosome of test position Left flanking marker of test position Absolute position of test from left telomere in Morgans Hi Ho oa FF W N Likelihood ratio test statistic of actual data For backcrosses this is itis Es it is Ho while for Fy s 6 Proportion of permuted data sets with an LR greater than or equal to the observed ER 7 Actual count of the number of permuted data sets with an LR greater than or equal to the observed LR In each step of the permutation test this file is rewritten and the number following the perm token incremented This way if the computer crashes during a run Zmapqtl can be restarted from where it left off If you were running Zmapqtl with 1 000 permutations and the pro
8. Yyn be a vector of quantitative trait values For each trait in turn it calculates the sample size n mean y 157 yi variance s a NY yi standard deviation s Vs skewness kurtosis and average deviation 4 1 lyi y The coefficient of variation is the sample standard deviation divided by the sample mean Lynch and Walsh 1998 provide a lucid explanation of some of the statistics calculated by Ostats Let the kth sample moment be M k X yk Clearly M 1 y Using the notation y M k we can estimate the sample variance with _ y 9 3 1 49 June 22 2000 CHAPTER 3 ANALYSIS qtlcart map 2 LRmapqtl 4 Zmapqtl 3 SRmapqtl 5 JZmapqtl 1 Ostats qtlcart Ir qtlcart z qtlcart sr qtlcart zt Figure 3 1 Analysis Schematic An estimate of the skewness is n2 PRU n 1 n 2 y 3 5 29 The standard error of skewness depends on the underlying distribution but can be approx imated by 6 n The coefficient of skewness k3 is Skw y k3 33 where the sample standard deviation s Vs is estimated from 3 1 Kurtosis is esti mated by Kur y 7 Te A sa Sua 39 and the coefficient of kurtosis is Kur y 3s4 a AS s Like skew the standard error of kurtosis is dependent upon the population distribution We give the estimate y24 n A test of normality for the vector y then involves the test statistic nk nki
9. but not automatically analyzed One includes these other traits in the regression model by prepending a plus sign to the other trait name For example Names of the other traits 1 Sex 2 Line would incorporate a Sex effect in the regression model while ignoring the Line effect 2 4 Prune Prune takes a genetic linkage map and a data set as input The user can either eliminate some of the data markers or traits bootstrap permute or simulate missing data Table 2 5 summarizes the command line options for Prune Originally Prune was strictly a command line program In adding the interactive menu it became necessary to add a second level of interaction When Prune is invoked in the interactive mode the user will see a menu in which all the parameters of Table 2 5 can be set The user will then proceed to another interactive menu in which data manipulation can be performed The second menu will list actions that can be taken The user selects an action and provides the proper values at which time the action is taken This second menu is in a loop The user can continue to take actions until the option to quit is chosen At the end the data set is printed out A few actions can be done automatically They are bootstrapping permuting and simulating missing data These are provided so that Prune can be run ina batch file for permutation tests or bootstrap experiments The output files of Prune may include a genetic linkage map and
10. 127 background parameter 58 beta distribution 37 bootstrap 27 44 60 70 BOOTSTRAPS JACKKNIVES AND PER MUTATIONS 132 bug 22 127 categorical trait 42 53 command line 23 composite interval mapping 46 56 65 93 covariate 47 69 cross advanced intercross 14 backcross 14 62 Design III 14 doubled haploid 14 intercross 14 63 recombinant inbred line 14 repeated backcross 14 test cross 14 CROSSES 110 dominance 37 43 ECM algorithm 57 137 June 22 2000 INDEX Eqtl 45 47 64 69 EQTL 131 Eqtl options 71 output 69 experimentwise significance level 65 experimentwise significance threshold 47 filename stem 31 FILENAME STEM 99 ftp server 19 75 81 gamma distribution 37 gamma function 37 genetic linkage map 14 16 33 38 genotype 39 GLOBAL BEHAVIOR 97 GLOBAL COMMAND LINE OPTIONS 97 GNUPLOT 17 33 36 72 73 help file 27 HELP FILE 98 heritability 38 40 HINTS 127 inbred line 14 INPUT FORMAT 104 107 110 113 115 118 120 122 126 132 install Macintosh 21 MS Windows 20 UNIX 20 interactive menu 26 27 42 interval mapping 46 56 65 93 jackknife 60 70 JZmapqtl 65 JZMAPQTL 125 JZmapqtl option 65 least squares 54 linear regression 53 LINPACK 18 138 LOD 72 log file 27 LR 72 LRmapqtl 53 LRMAPOTL 117 LRmapqtl options 54 output 54 Macintosh 16 19 24 44 binhex 21 install 21 StuffitExpander
11. NC 27695 8203 USA Phone 919 515 1934 118 CHAPTER 8 UNIX MAN PAGES QTL Cartographer 8 8 SRMAPOTL NAME SRmapqtl Map quantitative traits on a molecular map SYNOPSIS SRmapatl o output iinput m mapfile t trait M Model F pFin B pFout DESCRIPTION SRmapatl uses stepwise regression to map quantitative trait loci to a map of molecular markers It requires a molecular map that could be a random one produced by Rmap or a real one in the same format as the output of Rmap The sample could be a randomly generated one from Reross or a real one in the same format as the output of Rcross This program should be run before Zmapqtl if you want to use composite interval map ping The results will be used to pick markers background control in composite interval mapping The main result from using this program is to rank the markers in terms of their influence on the trait of interest OPTIONS See OTLcart 1 for more information on the global options h for help A for automatic V for non Verbose W path for a working directory R file to specify a resource file e to specify the log file s to specify a seed for the random number generator and X stem to specify a filename stem The options below are specific to this program If you use this program without specifying any options then you will get into a menu that allows you to set them interactively o This requires a filename for outp
12. The filename stem is an important concept in the usage of this package Beginning with version 1 12 the programs utilize the filename stem gtlcart All files are then named using this stem and filename extensions relevant to the filetype For example if the X option is followed by corn then when new files are created they will have the stem corn followed by a logical extension An example would be corn map for a genetic linkage map With some practice you will be able to know the contents of a file by its extension 99 June 22 2000 CHAPTER 8 UNIX MAN PAGES USING THE INDIVIDUAL PROGRAMS For now it is best to use the individual programs rather than the front end If you have no data then you would use the programs in the following order Rmap to create a random map of markers Rqtl to generate a random genetic model for the map Reross to create a random cross LRmapgqtl to do a simple linear regression of the data on the markers SRmapqftl to do a stepwise linear regression of the data on the markers to rank the markers Zmapqtl to do interval or composite interval mapping Preplot to reformat the output of the analysis for GNUPLOT GNUPLOT to see the results graphically If you have data then you might use the programs in the following order 1 2 7 8 Rmap to reformat the output of MAPMAKER or a standard input file Rcross to reformat your data Qstats to summarize missi
13. These are explained in greater detail in the manual INPUT FORMAT The input format of the molecular map should be the same as that of the output format from the program Rmap The input format of the individual data should be the same as the output format of the program Rcross EXAMPLES JZmapqtl Calculates the likelihood ratio test statistics of the dataset in gtlcart cro using the map in gtlcart map nice JZmapgtl A V i corn cro m corn map M 6 t 3 I 34 amp 126 CHAPTER 8 UNIX MAN PAGES QTL Cartographer Calculates the likelihood ratio test statistics of the dataset in corn cro using the map in corn map Model 6 is used for analysis This file has two traits so specifying trait 3 means that both traits are analyzed Hypothesis 34 means that GxE interactions are also analyzed The program is nice d as a courtesy to other users and run in the background so that the user can logout and relax MODELS Different parameters for the M option allow for the analysis of the data assuming different models See the Zmapqtl man page for explanations of models 3 6 and 7 These are the only models available in JZmapqtl REFERENCES 1 Lander E S and D Botstein 1989 Mapping Mendelian factors underlying quanti tative traits using RFLP linkage maps Genetics 121 185 199 2 Zeng Zhao Bang 1993 Theoretical basis for separation of multiple linked gene ef fects in mapping quantitative trait loci Proc Natl Acad
14. and determines whether the likelihood ratio test statistic is increasing or decreasing Upon a change it picks out the position and estimates of other parameters The user can specify that the peaks of interest need be higher than some Significance threshold to be considered QTLs The default is 3 84 that is any peak that is less than 3 84 is ignored This can be changed with the S option If you have run Zmapqtl and done a permutation test Eqtl automatically reads the output and sets the significance threshold subject to the value of the size set with the a option For a size of a the the 100 1 a 71 June 22 2000 CHAPTER 4 VISUALIZATION OF RESULTS percentile is calculated from the experimentwise test values The final option is a flag to output LOD scores rather than likelihood ratios The default behavior of the QTL Cartographer system is to use a likelihood ratio test statistic LR rather than a LOD score For a hypotheses H let L be the likelihood of the data given the hypothesis For a pair of hypotheses Ho and H1 this would yield Lo and L The LOD score is defined as Lo LOD log O E T The likelihood ratio test statistic LR is LR 2ln r 21n10 02 2 In10 LOD 4 605LOD 1 and thus LR 1 LOD log exp 3 los e LR 0 217LR 4 2 Preplot Preplot reformats the output of the analysis programs so that they may be plotted by GNUPLOT The output files could be imported
15. come from Zeng 1994 They are in the doc folder for the Macintosh versions in the same place as the binaries on PCs and in the example subdirectory in the UNIX version These are properly formatted and can be analyzed with Ostats LRmapqtl etc Do the following 1 Proceed with the analysis programs as in the previous example Be sure to set the proper filename stem mletest and working subdirectory Run Ostats LRmapqtl SRmapqtl and Zmapqtl Look at the output after each run 2 Start up Preplot Don t change any parameters Go ahead with the program 3 Start up GNUPLOT From the GNUPLOT command line type in load mletest plt This should display graphical results See the first example for the specifics of PCs and Macintoshes 4 Start up Eqtl Go ahead with the analysis Look at the output mletest eqt 5 7 Analyzing real data Create a new working subdirectory called realdat in you qwork subdirectory Copy the realdat inp files into it There should be two files realdatm inp and realdatc inp The former is a genetic linkage map in the standard input format map inp The latter is a file with marker and trait data in the standard input format cross inp This is a real data set kindly provided by Juan Medrano Horvat and Medrano 1995 It has also been used as an example in a review on the statistical issues in QTL mapping Doerge Zeng and Weir 1997 You will now transla
16. folder that the appli cations reside in The analogous lines for the MS Windows version would look like workdir test The working directory stem corn Stem for filenames helpfile qtlcart hlp The help file 26 CHAPTER 1 INTRODUCTION QTL Cartographer The working directory must exist before you run QTL Cartographer The help file is a plain ASCII text file with indicator tokens that allow it to be used by the programs This file is the same for all platforms and updated versions will be placed on the ftp server from time to time The user can place the help file anywhere and indicate its placement with the helpfile line in the resource file Filename stem The filename stem is an important concept in the usage of QTL Cartographer Beginning with version 1 12 the QTL Cartographer programs utilize the filename stem qtlcart All files are then named using this stem and filename extensions relevant to the filetype In the resource file example above the stem entry specifies corn as a stem for filenames This means that when new files are created they will have the stem corn followed by a logical extension An example would be corn map for a genetic linkage map With some practice you will be able to know the contents of a file by its extension You can set the filename stem on the command line with the X option Log File It s often useful to keep a log of the work done using
17. in F gt and other populations in which dominance can be estimated it is possible to test different sets of hypotheses The user can specify which results from the Zmapqtl output file to process The M option tells Eqtl to examine the results from using the specified analysis model An integer value should be given after the M option By default Eqtl looks for the results from Model 3 or interval mapping If you have done composite interval mapping with say model 6 then you should specify M 6 on the command line or in the interactive menu If model 6 was the last model run in Zmapqtl then Eqtl should already be aware of that fact The output file may also contain results from different traits The default trait is 1 but can be changed with the t option Of course some users may choose to have a different output file for each trait in turn and then the z and t options should be used together Remember that at the beginning of each set of results in the Zmapqtl output file the trait is specified Eqtl looks to match this For F design experiments various hypothesis tests can be performed These are explained in the previous chapter Using H with an integer allows you to specify which hypothesis test results to use Presently the choices are 1 2 and 3 for the Hs Ho Hs H and H Ho comparisons respectively Other Options Eqtl essentially finds the peaks in the graph of the results from Zmapqtl It goes along the chromosome
18. minimum first quartile median second quartile and maximum 3 1 1 Command Line Options Table 3 1 summarizes the command line options for Ostats There are very few of them You can specify the data set genetic linkage map file and output file In addition all the global options of Table 1 4 are valid Another function of Ostats is to summarize the missing data for markers traits and in dividuals Following the histogram there will be a table For each trait it will present a summary of missing data for each marker in turn The table will consist of seven columns 51 June 22 2000 CHAPTER3 ANALYSIS Option Default Explanation i qtlcart cro Data Input File 0 qtlcart qst Output File m qtlcartmap Genetic Linkage Map File Table 3 1 Command Line Options for Qstats The first three columns indicate the chromosome marker number and name of the marker if there is a marker name The fourth column specifies what type of marker Qstats thinks it is There are three types that are recognized The first is codominant and is indicated by a co token The other two are dominant markers and Qstats distinguishes between marker systems in which A is dominant to 42 indicated by the token A and those in which 42 is dominant to A a Column 5 has the counts of individuals with data for the marker while column 6 has the counts of individuals with both marker and trait data Column seven is just the ra
19. the DOF degrees of freedom for the numerator of that F statistic is given For forward stepwise or backward elimination SRmapqtl will try to rank all of the markers no matter how small the F statistic is For the forward regression with backward elimination the program proceeds to add variables until the F statistic p value is less than that specified by the F option 0 1 by default Then SRmapqtl rechecks all the variables added and will eliminate any with an F statistic p value less than the value given with the B option In general the FB method is probably the best method for picking background markers to be used with model 6 in Zmapqtl and JZmapqtl To this end SRmapqtl should be run prior to using either module Zmapqtl and JZmapqtl will read the results of SRmapqtl and use the markers that are ranked You can specify an upper bound to the number of background parameters to be used in Zmapqtl JZmapqtl will use all the markers that are listed for all traits in its analysis The FB method thus selects only a subset of significant markers Be aware that SRmapqftl tries to determine how many markers can be analyzed at once The number of parameters has to be smaller than the sample size If you try to use back ward regression and there are more markers than individuals then SRmapqtl will de fault to forward stepwise regression and rank as many markers as possible You should be aware that when dominance can be estimated each marker will
20. while qtlcart z6i will contain the state after each even numbered iteration If individual has no trait data then the th iteration will be skipped For this reason one cannot be sure that the file ending in j is the last iteration for odd sample sizes It is best to look at both files at the conclusion of a jackknife experiment and rename the interim file with the greater number of iterations to qtlcart z6i It this is done then Eqtl will recognize it and calculate the means and sample standard deviations of the test statistic and effects To clarify the interim file names we consider an example using Model 6 in Zmapqtl and the default filename stem qtlcart Table 3 5 lists the interim file names Eqtl automat ically looks for files named qtlcart z6e qtlcart z6a and qtlcart z6i These files will be processed and the appropriate calculations done Eqtl will overwrite the qtlcart z6b and qtlcart z6j files after completing its calculations so if you want to save them do so before running Eqtl If you chose to use another model say model 3 then the 6 in the filenames of Table 3 5 would be a 3 3 4 4 Output Here is a truncated example of the output of Zmapqtl for a backcross 890840384 filetype Zmapqtl out QTL Cartographer V 1 13b March 1998 This output file qtlcart z was created by Zmapqtl 61 June 22 2000 CHAPTER 3 ANALYSIS
21. 21 mailing list 22 MAPMAKER 16 28 34 40 85 89 mapping function 35 86 Fixed 36 Haldane 35 Kosambi 35 marker translation 90 maximum likelihood 57 missing data 51 MODEL 118 MODELS 123 127 MS Windows 16 24 44 GNUPLOT 20 Windows Explorer 20 Note 25 27 31 33 34 36 40 56 OPTIONS 96 102 106 109 112 115 117 119 121 125 129 131 permutation test 46 54 56 60 64 69 71 PERMUTATION TESTS 123 phenotype 39 PREPLOT 129 Preplot automagic 72 options 72 printing 21 72 74 Prune 42 60 PRUNE 112 Prune interactive menu 43 INDEX QTL Cartographer Ostats 49 recombination 39 OSTATS 115 trait 40 Ostats SRmapqtl 47 58 options 51 SRMAPOTL 119 QTL 14 37 SRmapqtl OTLCART 96 output 55 standard deviation 61 Rcross 38 stepwise regression 47 55 RCROSS 109 backward 55 Reross forward 55 input 40 89 forward backward 55 output 40 REFERENCES 100 104 107 111 113 116 token 85 118 120 124 127 130 132 UNIX 16 19 23 resource file 26 75 install 20 RESOURCE FILE 98 pe pa ses ue shell script 44 46 i USING THE INDIVIDUAL PROGRAMS Sep 100 input 85 input format 33 34 variance options 33 35 environmental 38 39 output 36 87 genetic 39 Rqtl 37 verbosity 26 RQTL 106 virtual marker 69 Rqtl input 37 38 87 web site 21 75 output 69 88 window size 58 working directory 24 26 75 sample WORKING DIRECTORY 99 average deviation 49 kurtosis
22. 49 Zmapqtl 47 56 mean 49 ZMAPOTL 121 skewness 49 Zmapqtl standard deviation 49 model 58 variance 49 option 59 simulation output 61 cross 38 virtual marker 58 gametes 39 genetic linkage map 33 genetic model 37 missing data 44 47 QTL 37 random number seed 25 139
23. 84 0 2 stop qtls The format of the information between the start and stop commands is unimportant You just need whitespaces around each piece of information All the marker names and their distances could be on one line Note that in the above example for Trait_1 there are two QTL on chromosome 1 If this file were called qtls inp then Rgtl A V i qtls inp would convert this file to the format required for the other programs in the QTL Cartog rapher system 6 2 2 Rqtl output files Rqtl overwrites any file that has the same name as specified as the output file Be careful not to destroy any important files The output file will contain the genetic model in a format suitable for input into Rcross 88 CHAPTER 6 INPUT FILE FORMATS QTL Cartographer 6 3 Data files These are files that contain marker and trait data The output format of Rcross is rather difficult for the user to read and create manually We have therefore provided ways to translate other formats 6 3 1 MAPMAKER raw files Rcross will convert MAPMAKER raw files for use in the QTL Cartographer system You will first need to use MAPMAKER to create a genetic linkage map Then convert the map into the Rmap out format for use withRcross Then use Reross to convert the MAPMAKER raw data file into the Rcross out format 6 3 2 Rcross input files We have also defined a format for your data It is similar to the input formats for Rmap and
24. BUGS If you use the interactive mode you can print out the results of crosses The analysis of these arbitrary crosses has not been fully integrated into the other programs SEE ALSO Rmap 1 Rqatl 1 Ostats 1 LRmapqtl 1 SRmapqtl 1 Zmapqtl 1 JZmapqtl 1 Eqtl Prune 1 Preplot 1 OTLcart 1 AUTHORS In general it is best to contact us via email basten statgen ncsu edu Christopher J Basten B S Weir and Z B Zeng Department of Statistics North Carolina State University Raleigh NC 27695 8203 USA Phone 919 515 1934 111 June 22 2000 CHAPTER 8 UNIX MAN PAGES 8 5 PRUNE NAME Prune Prune or resample the data set SYNOPSIS Prune o output iinput m mapfile I interactive M Model b simflag DESCRIPTION Prune allows one to eliminate markers or traits It removes the data from the file containing the cross and reconstructs the molecular map It requires a molecular map that could be a random one produced by Rmap or a real one in the same format as the output of Rmap The sample could be a randomly generated one from Rcross or a real one in the same format as the output of Rcross Prune also does bootstraps permutations and simulations of missing or dominant mark ers OPTIONS See OTLcart 1 for more information on the global options h for help A for automatic V for non Verbose W path for a working directory R file to specify a resource file e to specif
25. Sci USA 90 10972 10976 3 Zeng Zhao Bang 1994 Precision mapping of quantitative trait loci Genetics 136 1457 1468 4 Jiang Changjian and Zhao Bang Zeng 1995 Multiple trait analysis of genetic map ping for quantitative trait loci Genetics 140 1111 1127 BUGS Preplot ignores the output at present So far the program only does joint mapping and one form of GxE Tests for close linkage pleiotopic effects and other environmental effects will be added in the future HINTS You can select traits to include in the analysis in three ways a Set the trait to analyze at 0 so that no traits except those beginning with a plus sign are analyzed You would need to edit the cro file first to prepend a to all traits you wanted in the analysis b Set the trait to a value in the range 1 t inclusive where t is the number of traits in the cro file You will then get single trait results c Set the trait to a value greater than t Then all traits will be put in the analysis unless they begin with a minus sign As in a above you would need to edit the cro file to minus out some traits 127 June 22 2000 CHAPTER 8 UNIX MAN PAGES You need to set the hypothesis test for SFx and RFx crosses The default of 10 is ok for crosses in which there are only two marker genotypic classes BCx RIx To test GxE use 14 For SFx and RFx values of 30 31 or 32 are valid and a 34 invokes the GxE test Recall that w
26. The user can specify a type l error rate size and Eqtl will calculate a threshold value relevant to it Once done the threshold value will be remembered and used by subsequent runs of Eqtl or Preplot For bootstrap results from Zmapqtl using interval mapping Eqtl looks for a file qtl cart z3a If found Eqtl will read in the sums and sums of squares of the likelihood ratio additive effect and dominance effect at each position and print the mean and sample stan dard deviations into a summary file qtlcart z3d Eqtl does similar calculations for the jackknife results that would be in qtlcart z3i Table 4 1 shows the command line options specific to Eqtl 4 1 1 Options Files Similar to other programs in the QTL Cartographer system the input and output files can be specified A genetic linkage map and a file containing the results of Zmapqtl must exist and be properly specified to Eqtl 70 CHAPTER 4 VISUALIZATION OF RESULTS QTL Cartographer Option Default Explanation Z qtlcart z Composite Interval Mapping Results 0 qtlcart eqt Output File m qtlcart map Genetic Linkage Map File M 3 Model from Zmapqtl H 1 Hypothesis Test 30 31 32 for F2 S 10 0 Significance threshold a 0 05 Size a L 0 Output LOD scores 0 no 1 yes Table 4 1 Command Line Options for Eqtl Which Results The output file from Zmapqtl may contain the results of analyzing different traits using different models Furthermore
27. a dater ee delete 3 2 1 Simple Linear Regression 322 QUPUE ES A A ARE ntm 3 2 3 Permutation Tests o Ey tsi ted i oaa eei a ede asi TD Hot Output Is A cg SA Zmapdtl 2 sasa ea a ERA AAA 3 4 1 Computational Methodology 342 Mod l s aare e de nue eS 343 Zmapqtl Options 94A n OUTPUT e Y A EO Ue des 3 5 LmMapatlisiosare da er EI 3 5 1 JZmapqtlOptions 352 ROMPE Sener Ste e ons etes 39537 ISAS ETS cer e ner re 4 Visualization of Results AT E ae ae ae e A AA EA ADA Options eeng pu e ee ia delo PrEplOt 2 ir a eos de a ir Bt 4 2 1 Printing Results aia AS GNUPLOT 4525 8 0 AA ta ne a 43 1 Basic GNUPLOT 5 Tutorial Examples 5 1 General tactics andnotes 5 2 Basic Macintosh 5 3 Basic Windows 5 3 1 Navigating disks Gade WS ele ee a 5 4 Basie UniX zoho kod A nee eee de eae oes CONTENTS QTL Cartographer SAL Helpk d seins lah Seite ae St a te ad ae a ee Gat i a a ae 77 5 42 Basic filesystem commands 77 543r Curious at A Be ati eh ee Pe eee A 78 5 44 Other commands 78 5 5 Simulating and Analyzing data 44 444 ua hu dues 79 5 6 Analyzing simulated data oases cas Be dees te dus diva ni as 80 5 7 Analyzing teal data gt Dira e aceon e a M AE a 80 5 8 Analyzing a MAPMAKER data set sss 22 Lou ee ek e
28. action Dominant Markers There are actions to eliminate dominant markers from the data set These options were in cluded at a time when Zmapqtl and LRmapqtl couldn t analyze dominant markers With the addition of subroutines to analyze dominant markers the need for these options has lessened Selecting option 1 or 2 in the second interactive menu eliminates one dominant markers or one type or the other 43 June 22 2000 CHAPTER 2 SIMULATING REFORMATTING DATA Eliminating markers and traits Option 3 of the interactive menu has an action to eliminate a specific marker You should be aware that the order of elimination is important If all the markers to be eliminated are on separate chromosomes the order is unimportant If two markers from the same chromosome are to be eliminated higher numbered marker should be eliminated first The same concept holds for traits with option 4 Eliminate them in the order of highest to lowest You will need to know the marker number and chromosome number rather than the marker name to use this option Culling sparse data Some markers or traits may have been typed for a small proportion of individuals in the dataset Such markers or traits can be eliminated from the data set Option 5 will allow you to specify a trait number and then eliminate individuals with missing data for that trait Choosing option 6 will require a tolerance level for the percentage of missing marker data The M option spec
29. and logon as a different user option and then click OK 5 3 1 Navigating disks I generally use Windows Explorer to navigate the disks You can click on files copy them and paste them in different directories to make copies of files If you are not familiar with Windows Explorer take a few minutes to play with it You can double click on the My Computer icon and icons therein to explore your hard drive Viewing files There are a lot of options for viewing files Generally I recommend using Notepad It is a simple text editor with a fixed width font You can find it under Start Programs Accessories Notepad When you try to open a file be sure to tell Notepad to look for files of type All Files If you don t then Notepad will only show files with a txt extension Windows NT does not like files to be accessed by two programs at once Be sure to clear out Notepad by creating a new file before running any programs that might read or write to a file that you are viewing 76 CHAPTER 5 TUTORIAL EXAMPLES QTL Cartographer Command Prompt Clicking on Start Programs Command Prompt brings up a command line window for DOS commands You can ftp or telnet from this window if you wish to transfer files or logon to an account elsewhere There is a text editor that can be started with the command edit that will allow you to view files Again take care not to open files that are being access
30. are using a PC the program names will all have an exe ending In this exercise you will simulate a genetic linkage map then a model and finally a data set This data will then be analyzed 1 Start up Rmap Select the option to change the filename stem Change the filename stem to sim You can change any parameters that you like We suggest changing the variances of markers per chromosome and intermarker distance to values other than 0 0 In each case a value of 2 or 3 would work well for the purposes of this exercise Don t change the output format If you are on a Macintosh or MS Windows machine be sure to set the proper working subdirectory folder When satisfied with the parameter values select 0 to run the program Look at the output sim map Start up Rqtl You probably don t need to change any parameters You can run this program with the 0 option Look at the output sim qtl Start up Rcross Again you do not need to change any parameters but you could try a different experimental design Select the number associated with the experimental design Change its value from B1 to SF3 or whatever you like from Table 1 1 Run this program with the 0 option Look at the output sim cro From this point on the analyses will utilize this file and the sim map file Start up Ostats and run it without changing any parameters Look at the output sim qst Start up L
31. automatic V for non Verbose W path for a working directory R file to specify a resource file e to specify the log file s to specify a seed for the random number generator and X stem to specify a filename stem The options below are specific to this program If you use this program without specifying any options then you will get into a menu that allows you to set them interactively o This requires a filename for output JZmapqtl will append the file if it exists and create a new file if it does not If not used then JZmapqtl will use gtlcart zj where the j indicates the trait analyzed and the zero th file contains joint mapping i This requires an input filename This file must exist It should be in the same format as the output of Rcross The default file is gtlcart cro m JZmapqtl requires a genetic linkage map This option requires the name of a file con taining the map It should be in the same format that Rmap outputs The default file is gtlcart map t Use this to specify which trait JZmapqtl will analyze If this number is greater than the number of traits then all traits will be analyzed unless the trait name begins with a minus sign If a negative number is given then only traits beginning with a plus sign will be analyzed The default is to analyze trait 1 only 125 June 22 2000 CHAPTER 8 UNIX MAN PAGES E Allows the user to specify the name of the file containing results from Eqtl JZmapqtl reads th
32. best to contact us via email basten statgen ncsu edu Christopher J Basten B S Weir and Z B Zeng Department of Statistics North Carolina State University Raleigh NC 27695 8203 USA Phone 919 515 1934 105 June 22 2000 CHAPTER 8 UNIX MAN PAGES 83 ROTL NAME Rqtl Place a set of estimated or randomly generated QTLs on a molecular map SYNOPSIS Rqtl o output i input m mapfile b beta t Traits q QTLperTrait d dominance 1 betal 2 beta DESCRIPTION Rqtl will translate a genetic model or simulate a random model for use by Rcross to sim ulate a data set It places a specified number of QTLs Quantitative Trait Loci on the molecular map created or translated by Rmap For simulations they are placed randomly on the map and the additive and dominace effects are also determined The molecular map could be a random one produced by Rmap or a real one in the same format as the output of Rmap OPTIONS See OTLcart 1 for more information on the global options h for help A for automatic V for non Verbose W path for a working directory R file to specify a resource file e to specify the log file s to specify a seed for the random number generator and X stem to specify a filename stem The options below are specific to this program If you use this program without specifying any options then you will get into a menu that allows you to set them interactively o
33. convey the meaning of Random Map Since then we have in cluded the ability to translate genetic linkage map information from various formats into that required by the QTL Cartographer system Thus the R can now mean reformat or random If you have no data you can simulate a genetic linkage map Rmap allows the user to specify the number of chromosomes markers per chromosome and average intermarker distance for the simulation You can also specify standard deviations for the latter two quantities This would yield a simulated map that better approximates one that you might actually produce in the lab Finally you can also specify whether you want some genetic 32 CHAPTER 2 SIMULATING REFORMATTING DATA QTL Cartographer material outside the most telomeric markers on the chromosomes Rmap can also read in files in three formats The first format is the same as its output format We will refer to this as Rmap out filetype format This feature is provided so that you can create as set of output files that GNUPLOT can read and display a graphic representation of your markers The second format is that which is produced by MAPMAKER Lander et al 1987 Lincoln et al 1992 We will refer to it as a mapmaker maps filetype format Rmap will read in the MAPMAKER output and reformat into the Rmap out format The third format is defined in Section 6 1 2 and in the file map inp included with the distribution of the program
34. count two towards the total number of parameters and you will need a sample size of at least twice the number of markers to do backward elimination 3 4 Zmapqtl Zmapqtl implements interval and composite interval mapping There are also options to perform a permutation test Churchill and Doerge 1994 Doerge and Churchill 1996 56 CHAPTER 3 ANALYSIS QTL Cartographer 3 4 1 Computational Methodology Composite interval mapping Zeng 1993 Zeng 1994 combines interval mapping with multiple regression The statistical model is defined as Y x b 2 d XB E 3 5 where e Y isa vector of trait values e b and d are the additive and dominance effects of the putative QTL being tested e x and z are indicator variable vectors specifying the probabilities of an individual being in different genotypes for the putative QTL constructed by flanking makers e B is the vector of effects of other selected markers fitted in the model e X is the marker information matrix for those selected markers e E is the error vector Estimates of the parameters are obtained by maximum likelihood through an ECM for Expectation Conditional Maximization algorithm Meng and Rubin 1993 In each E step the probability of an individual being in different genotypes of the putative QTL is updated In the CM step the estimation of parameters b and d is separated from that of B and each group is estimated conditional on the others This procedure is implemente
35. derived from that intercross The first part of the string T XX indicates that phenotyping is done on the XX population and the second part SF or RF indicates the genotyped population XX can be a B Bo SF or D3 for SF lines or B or B for RF lines D3 stands for Design III experiments Cockerham and Zeng 1996 All of the above experimental designs can be simulated and all but the Design III ex 14 CHAPTER 1 INTRODUCTION QTL Cartographer Fi Figure 1 1 Basic Cross y D EI Design Code Example Backcross to P B B1 Backcross j times to P Bij B13 Selfed generation i intercross SF SF3 Randomly mated generation i intercross RF RF2 Doubled Haploid Rlo RIO Recombinant Inbred via selfing RI RO Recombinant Inbred via sib mating Rl RI2 Testcross of SF to P T B SF T B1 SF3 Testcross of SF for j generations T SFj4 SF T SF4 SF3 Testcross of RF to P T B RF T B1 RF3 Design III T D3 SF T D3 SF5 Table 1 1 Summary of Experimental Design Codes periments can be analyzed Table 1 1 lists all the experimental designs and their QTL Cartographer codes The experimental designs of Table 1 1 can be specified in Rcross for simulations or in certain data input files see Section 6 3 2 15 June 22 2000 CHAPTER 1 INTRODUCTION 1 13 Genetic Linkage Maps A known genetic linkage map will be required for the analysis A good genetic linkage map will comprise a set of Mendelian marker loci t
36. from Numerical Recipes in C 18 CHAPTER 1 INTRODUCTION QTL Cartographer 1 4 How to Get and Install OTL Cartographer Point your web browser to http statgen ncsu edu and follow the link to software and from there to QTL Cartographer You can then follow the link to the ftp site and shift click on the files you want to download QTL Cartographer is also downloadable via anonymous ftp at statgen ncsu edu 152 1 95 36 Use ftp as your username and your email address as the password Here is an example username ftp password basten statgen ncsu edu Next change directory into the distribution subdirectory pub qtlcart and view what is available For example ftp gt cd pub gqtlicart ftp gt ls ChangeLog QTLCartMac sea hqx QTLCartWin zip OTLCart tar Z README gnuplot exe gnuplot sit hqx gnuplot tar Z 1 106 12 13g swang Download the appropriate version Presently the following versions are available e QTLCartWin zip is for Microsoft Windows These are 32 bit applications You will need an unzip utility to unpack this file There should be an unzip utility in the Windows system folder e QTLCartMac sea hqx is for Macintoshes e OTLCart tar Z is for UNIX e gnuplot files are the distributions of GNUPLOT for various platforms e The 1 10b 1 12f and 1 13g are directories containing older versions of QTL Cartogra pher e The swang directory contains Shengchu Wang s Window
37. from the program Rmap The input format of the individual data should be the same as the output format of the program Rcross EXAMPLES o SRmapqtl i corn cro m corn map M 2 Does a forward stepwise regression with a backward elimination step for the dataset in corn cro using the genetic linkage map in corn map REFERENCES BUGS Forward and backward regression should probably use the thresholds for adding and deleting markers from the model When that feature is added the F and B options will have more use SEE ALSO Rmap 1 Rqtl 1 Reross 1 Ostats 1 LRmapqtl 1 Zmapqtl 1 JZmapqtl 1 Eqtl 1 Prune 1 Preplot 1 OTLcart 1 AUTHORS In general it is best to contact us via email basten statgen ncsu edu Christopher J Basten B S Weir and Z B Zeng Department of Statistics North Carolina State University Raleigh NC 27695 8203 USA Phone 919 515 1934 120 CHAPTER 8 UNIX MAN PAGES QTL Cartographer 8 9 ZMAPOTL NAME Zmapqtl Composite interval mapping module SYNOPSIS Zmapqtl o output i input m mapfile 1 Irfile S srfile t trait M Model c chrom 1 d walk n nbp 1 w window r perms r boots DESCRIPTION Zmapqtl uses composite interval mapping to map quantitative trait loci to a map of molec ular markers It requires a molecular map that could be a random one produced by Rmap or a real one in the same format as the output of Rmap T
38. includ ing simulating reformatting or analyzing data and visualizing the results of the analyses Presently the mapping programs can handle data from backcrosses intercrosses and re combinant inbreds as well as a few other experimental designs see Table 1 1 All input and output files are plain text and can be viewed or imported into many text editors and graphics packages on various computing platforms The programs were origi nally written for the UNIX operating system and have since been ported to the Macintosh and Microsoft Windows operating systems Present development is on a Macintosh us ing Metroworks Codewarrior Both Macintosh and Windows binaries are created using Metroworks The UNIX distribution is of the source code This project is ongoing and sug gestions are welcome for further improvements and enhancements The source code and compiled binaries are freely available and may be obtained by anyone over the internet 1 1 1 Definition of the Problem Often traits in plants and animals are influenced by many genes rather than a single lo cus Falconer and MacKay 1996 for an excellent general review These traits are termed quantitative traits and the loci that control these traits quantitative trait loci abbreviated 13 June 22 2000 CHAPTER 1 INTRODUCTION henceforth as QTLs An important goal in genetics and breeding is to identify and charac terize QTLs especially those that contribute to variation in quantitativ
39. integer indicating the maximum number of markers on any chromosome end or quit indicate to Rmap that it should stop reading from the file named Should be followed by yes or no indicating whether the marker systems have names start indicates the start of the genetic map stop indicates the end of the genetic map skip tells Rmap to ignore all tokens until an unskip token is encountered unskip see above Chromosome should be followed by an integer indicating the chromosome number The first line should start with a and have some long integer after it After that it should have the token bychromosome The number will be an identifier for the file and should be unique The token bychromosome indicates how the map should be read in Here is an example of a first line 123456789 bychromosome filetype map inp The final pair of tokens indicate what type of file it is Between the start token and the stop token you should have a repeating sequence of a Chromosome token an integer then markers ordered with their names followed by the appropriate distances This example has the markers followed by their positions in centiMorgans All markers should have unique names Start Chromosome 1 Marker1_1 Marker1_2 Marker1_3 34 1 Marker1_4 Marker1_5 10 2 43 3 52 41 Chromosome 2 Marker2_1 0 0 86 CHAPTER 6 INPUT FILE FORMATS QTL Cartographer Marker2_
40. map of chromosome 1 A J Hum Genet 31 680 696 Sturt E 1976 A mapping function for human chromosomes Ann Hum Genet Lond 40 147 147 Williams T and C Kelley 1993 GNUPLOT An Interactive Plotting Program Version 3 5 Zeng Z 1992 Correcting the bias of wright s estimates of the number of genes affect ing a quantitative trait A further improved method Genetics 131 987 1001 Zeng Z 1993 Theoretical basis for separation of multiple linked gene effects in map ping quantitative trait loci Proc Natl Acad Sci USA 90 10972 10976 Zeng Z 1994 Precision mapping of quantitative trait loci Genetics 136 1457 1468 136 Index A 96 a 131 B 120 b 107 113 122 c 103 110 122 126 d 103 107 122 126 E 110 126 e 97 F 120 f 103 g 103 H 110 130 131 h 96 I 110 112 126 1 102 106 109 112 115 117 119 121 125 L 130 132 1 121 129 M 103 113 119 122 126 m 103 106 109 112 115 117 119 121 125 129 131 n 110 122 126 0 102 106 109 112 115 117 119 121 125 129 131 p 103 q 106 109 129 R 96 r 117 122 S 122 126 130 131 s 96 T 129 t 103 106 117 119 121 125 V 96 vd 103 vm 103 W 96 w 122 126 X 97 Z 129 131 T Williams and C Kelley 1993 GNUPLOT An Interactive Plotting Program Version 3 5 100 a 127 additive effect 37 39 b
41. minus sign will be analyzed in succession This only works with models 1 2 3 and 6 One can also limit the analysis to a single chromosome with the c option Background Parameters and Window Sizes For models 5 and 6 one can specify the size of the window ws on either side of the test interval that is blocked from having markers in the background This option is ignored for all models except 5 and 6 The number of background parameters n is only used with model 6 and is explained above Permutations Bootstraps and Jackknives Zmapatl allows for permutation tests and bootstrap or jackknife resamplings The former is a way to determine experimentwise significance levels and comparisonwise probabil ities Churchill and Doerge 1994 Doerge and Churchill 1996 Phenotypes are shuffled against genotypes and the analyses are redone For each test position the comparison wise probability or P value is the proportion of permuted datasets that have test statistics less than the observed data set test statistic It should correspond to the probability of the observed test statistic assuming a x distribution with one degree of freedom For the ex perimentwise significance level the highest test statistic in each permutation is recorded and these are ordered at the end of the permutations The 90 95 97 5 and 99th percentile values are then the experimentwise significance levels at a 0 1 0 05 0 025 and 0 01 re spectively Permutation
42. or Fetch will unbinhex the files for you although they may require a helper application Once the file OTLCart sea hqx has been unpacked you will have a folder called bin with the programs in it You can simply double click on any of them to start them up You will first be presented with a console window All you need to do is click on OK to get to the interactive menu for setting options Note that when you double click on a QTL Cartographer program you will get a command line interface window You can simply click OK here to get the menu You can also enter command line options in that box if you like 15 Getting Help One of the best places to get help is from the UNIX man pages These should be installed with the UNIX distribution of the program and are described in the APPENDIX Since Macintosh and MS Windows users won t have the man pages we have attached them to this document as an APPENDIX and included them in the doc subdirectory This docu ment as well as the man pages are available via the World Wide Web by pointing your web browser to 21 June 22 2000 CHAPTER 1 INTRODUCTION http statgen ncsu edu and following the link to QTL Cartographer which is halfway down the page 1 5 1 Mailing List The address for the mailing list server is MajorDomo statgen ncsu edu Please join the mail ing list for QTL Cartographer It will be a forum for problems you may have in using the programs and we will post annou
43. qtls inp that is explained in Section 6 2 1 The given set of QTLs might be made up by the user or a set of estimates from a previous analysis of a data set Table 2 3 presents the command line options for Rqtl The default values from the table tell Rqtl to simulate nine QTLs for one trait For simulations the user can specify the average number of QTLs per trait the number of traits and parameters for dominance and additive effects We use the convention that Q alleles are from for P lines and Q from P lines Dominance can take on the values 1 2 3 or 4 1 means no dominance while 2 means Q is dominant and 3 means Q is dominant A value of 4 means that dominance for each QTL will be random in magnitude and sign The degree of dominance will be a Beta random variable d with shape parameters 61 2 The density function for d is d 1 1 1 d 82 1 pra as Pra 20 leads B B1 82 i Ha 0 otherwise 28 where _ T 81 1 062 BW Ba a 2 6 and T x is the gamma function T x ye De Y dy 0 Option Default Explanation i None Input File 0 qtlcart qtl Output File m qtlcartimap Genetic Linkage Map File 1 Number of Traits q 9 Number of QTL per Trait b 2 0 Additive effect parameter beta 1 2 0 Dominance effect parameter 3 2 2 0 Dominance effect parameter 2 d 1 Dominance Table 2 3 Command Line Options for Rqtl The additive effects of the QTLs are independent identically distributed ra
44. random variable with mean d and standard deviation oq Finally the amount of telomeric or tail DNA is simulated as a normal random variable with mean t and standard deviation o Setting a standard deviation equal to zero means that the quantity in question is not a random variable but set equal to its mean value The parameters c m d t on and og can be set using the command line options of Table 2 1 or in the interactive menu Note that if an input file is specified all these parameters are ignored and Rmap attempts to translate the input file An alternate method of simulating the genetic linkage map can be invoked by changing the simulation mode parameter from 0 to 1 using the M command line option In this version the length of the chromosomes will be normally distributed with mean d and standard deviation og The number of markers on a chromosome will still be normally distributed with mean m and standard deviation om but will be placed on the chromosome following a uniform distribution You should set the values of d and oq to appropriate 33 June 22 2000 CHAPTER 2 SIMULATING REFORMATTING DATA levels as they are for chromosome length rather than intermarker distance in this mode For example if you want roughly the same results from this mode as that in the original then set d 16 x 10 160 in this mode 2 12 Using MAPMAKER EXP files QTL Cartographer has the added capability of reading map files generated by MAP MAKE
45. sample could be a randomly generated one from Reross or a real one in the same format as the output of Rcross OPTIONS See OTLcart 1 for more information on the global options h for help A for automatic V for non Verbose W path for a working directory R file to specify a resource file e to specify the log file s to specify a seed for the random number generator and X stem to specify a filename stem The options below are specific to this program If you use this program without specifying any options then you will get into a menu that allows you to set them interactively o This requires a filename for output LRmapqtl will append the file if it exists and create a new file if it does not If not used then LRmapqftl will use gtlcart Ir i This requires an input filename This file must exist It should be in the same format as the output of Rcross The default file is gtlcart cro m LRmapqtl requires a genetic linkage map This option requires the name of a file containing the map It should be in the same format that Rmap outputs The default file is gtlcart map r LRmapqtl will do a permutation test a la Churchill and Doerge 1994 This option specifies the number of permutions to do It is zero by default which means no permuation test is done If used you must specify a positive integer Usually 1 000 is sufficient t Use this to specify which trait LRmapqtl will analyze If this number is greater than the num
46. the programs The e option can be used to specify the log or error file Each time a program in the QTL Cartographer system runs a summary of all the parameters and options is written to the log file The file also keeps track of when the program was run and may contain other diagnostic information The log file is appended to with each run rather than overwritten Remember that the log file is appended to during each invocation of any of the programs This is something to keep in mind if you do a bootstrap in a batch file After a thousand replications the log file will tend to grow large The batch file examples included with the QTL Cartographer system see 2 4 2 take this into account by saving a copy of the log file before running the bootstrap and deleting the large and unnecessary log file at the end Interactive Mode The default behavior for the QTL Cartographer programs is to present the user with a menu of numbered options This menu is in a loop so the user can pick options and change them one at a time when satisfied that the proper options have been set selecting 0 will tell the program to continue There will always be an option to quit without doing anything This will be the last numbered option When 0 is chosen the programs will present a summary of the options and continue At termination the options will be written to the resource file so that the options and parameter values are remembered There are a f
47. the user can eliminate individuals markers or traits from the data set In addition Prune allows one to bootstrap or permute the data as well as to simulate missing markers Regardless of whether the data are simulated or real the important output files from this step are the genetic linkage map and the data set We will refer to these files as qtlcart map and qtlcart cro although you can name them anything you like In fact we generally decide on a filename stem and use filename extensions to indicate what is in the various files If we were working on a corn data set we might have files corn map and corn cro for the genetic linkage map and marker trait data set respectively The naming scheme would be consistent throughout the analysis One note on the behavior of Rmap Rqtl and Reross If you choose to translate a data file then the parameters for simulations are unnecessary and they disappear from the interactive menu If you specify no input file for any of these programs by entering a period all by itself for the input filename then the simulation parameters will reappear for the user to change 31 June 22 2000 CHAPTER 2 SIMULATING REFORMATTING DATA Mapinp 1 Rmap gt qtlcart map crue 2 Rcross qtlcart cro Figure 2 1 Reformatting Data Figure 2 2 Simulating Data 2 1 Rmap Originally the program Rmap was designed to simulate a genetic linkage map The R in Rmap was meant to
48. used with the b option b Prune will read in the map and data file and do one of four things depending on the value given to this option 1 a bootstrap resampling of the data where sampling of individuals is done with replacement to create a sample of the same size as the orig inal 2 A permutation of the traits A new dataset is then printed 3 A simulation of missing markers 4 A simulation of dominant markers A new dataset is printed with the percent of missing marker data specified by the M option A value of zero means that this option is ignored INPUT FORMAT The input format of the molecular map should be the same as that of the output format from the program Rmap The input format of the individual data should be the same as the output format of the program Rcross EXAMPLES o Prune m example map i example cross o exout Puts the user into an interactive menu for eliminating traits markers etc Prune m example map i example cross o exout b 1 The b option creates a new sample from the old The new sample is created by resampling the original sample with replacement Phenotypes and genotypes are kept together The new sample will have the same sample size as the old one It will be written to exout crb No new map will be written REFERENCES BUGS You can eliminate multiple markers in the interactive loop You should be aware that the order marker elimination is important If all the markers to be eliminat
49. 1997 Statistical issues in the search for genes affecting quantitative traits in experimental populations Stat Sci 0 000 000 Dongarra J J C B Moler J R Bunch and G W Stewart 1979 LINPACK Users Guide Philadelphia PA SIAM Falconer D S and T F C MacKay 1996 Introduction to Quantitative Genetics Essex UK Longman Group Limited Felsenstein J 1979 A mathematically tractable family of genetic mapping functions with different amounts of interference Genetics 91 769 775 Fisch R D M Ragot and G Gay 1996 A generalization of the mixture model in the mapping of quantitative trait loci for progeny from a bi parental cross of inbred lines Genetics 143 571 577 Haldane J B S 1919 The combination of linkage values and the calculation of dis tances between the loci of linked factors J Genet 8 299 309 Horvat S and J F Medrano 1995 Interval mapping of high growth hg a major locus that increases weight gain in mice Genetics 139 1737 1748 135 June 22 2000 BIBLIOGRAPHY Jiang C and Z Zeng 1995 Multiple trait analysis of genetic mapping for quantitative trait loci Genetics 140 1111 1127 Jiang C and Z Zeng 1997 Mapping quantitative trait loci with dominant and missing markers in various crosses from two inbred lines Genetica 101 47 58 Karlin S 1984 Theoretical aspects of genetic map functions in recombination pro cesses In A Chakravarti Ed
50. 2 Sa Marker2_3 HERS eb Marker2_4 24 8 stop You can annotate the input file as much as you want Just don t put in any extra material before the stop token Everything after the end token is ignored Before the start to ken only the type function units chromosomes and maximum tokens are processed The token following each of these is read and the information used in the program The format of the information between the start and stop commands is unimportant You just need whitespaces around each piece of information All the marker names and their distances could be on one line If this file were called map inp then Rmap A V i map inp would convert this file to the format required for the other programs in the QTL Cartog rapher system If named had a value of no above then the format of the distances would be start Chromosome 1 Chromosome 2 stop 10 2 SAT 43 3 9261 13 7 19 1 24 8 0 0 0 0 6 1 3 Rmap output files Rmap overwrites any file that has the same name as specified as the output file Be careful not to destroy any important files The output file will contain the values of the parameters used the names of chromosomes and markers if a translation was made and the linkage map 6 2 OTL information You can specify a genetic model and use it for simulation by translating it with Rqtl This would be useful if you want to do some what if
51. 3 4 LINPACK Copyright Information 13 5 Numerical Recipes in C Information How to Get and Install QTL Cartographer 141 MS Windows ef 42 UNIX Eat Be aie he An Bin Osa Ra AM AN er ate Ca TE 143 Macintosh 2 33 2 jae ale Gea ea ML Cate ras sas Getting Helpis ato sl big PA Bre cent A Se ah Poems as dal NAMI EASE 2 e e E Beene 12 Bug Repos inss e mii eae A Le Sele ed 153 Contacts exo aaa oe a opte a E et a es General Usage of the Programs icy ccs te he A ae das 1 6 1 Options for all programs di hy eee ad a 1 6 2 Filenaming Conventions rs rd eed a ke RY ee a Simulating Reformatting Data Rap ef at Y eee SA A CES Zila Simu latinga Map in ne eee A Be oe Boeke 2 1 2 Using MAPMARER EXP files pi ed es seul 2 13 QTL Cartographer user input format 5 11 13 13 13 14 16 16 16 16 17 17 18 18 19 20 20 21 21 22 22 23 23 24 28 June 22 2000 CONTENTS 2 14 Command Line Options 200 se a elem DAA id Ed 237 ROSS A AN Na 2 3 1 Simulating Data i ea 2 3 2 Translating Data 233 OUR 42 AAA ve SA aA 24 PUDO ua A mn A ea ME re 2 4 1 Pruning Datasets ices eke Lo e 2 4 2 Recreating Datasets here seat 3 Analysis JL Ostia LR aoe en ee EA SA Re aF 3 1 1 Command Line Options 3 1 2 Segregation e Some oes ne nee 3 2 LRMApPOE sa ete
52. 5 Simulates a random map where the number of markers on each of 23 chromosomes has a normal distribution with mean 16 and standard deviation 3 The intermaker distance is normally distributed with mean 10 cM and standard deviation 1 There will be some genetic material outside the flanking markers on each chromosome with a mean length of 5 cM and standard deviation 0 5 o Rmap o Map out i map mps Opens the file map mps tries to determine its format and translates it if possible The output will be written to the file Map out The extension mps should be used with MAP MAKER output files and the string filetype mapmaker mps should be put somewhere in the first twenty lines of the file REFERENCES 1 Lander E S P Green J Abrahamson A Barlow M Daley S Lincoln and L New burg 1987 MAPMAKER An interactive computer package for constructing pri mary genetic linkage maps of experimental and natural populations Genomics 1 174 181 2 T Williams and C Kelley 1993 GNUPLOT An Interactive Plotting Program Ver sion 3 5 BUGS Note that if MAPMAKER outputs an intermarker distance of 0 00 cM then Rmap will translate it to 0 0001 cM In fact all intermarker distances of 0 0 will be reset to 0 0001 cM SEE ALSO Ratl 1 Reross 1 Ostats 1 LRmapqtl 1 SRmapqtl 1 Zmapqtl 1 JZmapqtl 1 Eqtl Prune 1 Preplot 1 OTLcart 1 104 CHAPTER 8 UNIX MAN PAGES QTL Cartographer AUTHORS In general it is
53. 64 characters in length They should have a repeating order that is the same as the trait data and the missingtrait command is recognized missingtrait start otraits Sex MFMMMPFEME E Broca id LL kd 22 22 2 stop otraits Data by individuals Another way to organize the data is by individuals The program expects that the markers are ordered from marker 1 on chromosome 1 marker 2 on chromosome 1 to the last marker on the last chromosome Since the individuals are named they can be in any order start individuals markers Ind_1 2 2 DAD 2 2 2 Ind 2 222 De iD EInd 3 2 2 2 22 2 Ends 2 122 2 27 Ind_5 2 D D 2 9 Ind_6 1 Ind_7 2 Ind_8 2 91 June 22 2000 CHAPTER 6 INPUT FILE FORMATS Ind_9 22 Ind_10 2 2 stop individual ls markers The traits are done similarly All the traits have to be in these types of blocks but you can have more than one block Each column is for a different trait After the start token put individuals followed by traits then the number of traits 2 in this case then the names of the traits then indicate whether the individuals are named Here they are named but if they weren t put an notnamed token where the named token presently is Other traits follow a similar pattern and an example is given below start individuals tra
54. ANALYSIS QTL Cartographer 3 3 SRmapqtl SRmapqtl uses the technique of stepwise regression to search for QTLs For forward and backward regression it simply ranks the markers for their effect on the quantitative trait In forward stepwise regression FS each marker in turn is tested for its effect on the quan titative trait using linear regression That marker with the largest partial F statistic is as signed rank 1 and included in all subsequent analyses Step two tests all the remaining markers and assigns rank 2 to the marker with the largest partial F statistic This is re peated until all the markers have been ranked Option Default Explanation i qtlcart cro Input File 0 qtlcart sr Output File e qtlcart log Error File m qtlcart map Genetic Linkage Map File S 860437285 Random Number Seed M 0 FS BE or FB 0 1 2 t 1 Trait to analyze F 0 1 Size p Fin B 0 1 Size p Fout Table 3 3 Command Line Options for SRmapqtl Backward elimination regression BE starts with all markers in the model In the first step each marker in turn is removed and a partial F statistic is calculated That marker with the smallest partial F statistic is given the lowest rank and removed from subsequent analyses This is repeated until all the markers have been ranked The above methods seek only to rank the markers They make no effort to determine whether adding or deleting a marker makes a significant difference f
55. BER The first two words of a MAPMAKER QTL raw file should be data type Older versions of Reross cannot process comments at the beginning of a raw file In fact it depends on those first two words to recognize the file as a MAPMAKER QTL raw file Beginning with version 1 12 comments will be allowed in the beginning of a mapmaker raw file if you include the filetype mapmaker raw indicator within the first 100 lines of your file It is usually best to put this on the first line Reross will recognize the command and translate the file appropriately You might want to get into the habit of putting the filetype token with an appropriate identifier in your input files as it will become more important in future releases of QTL Cartographer There are two other things to keep in mind when using MAPMAKER QTI files The first is that marker and trait names are truncated to eight characters in the output Versions of QTL Cartographer prior to 1 12 will be tripped up by this Secondly MAPMAKER EXP has been known to translate underscores _ as minus signs in its output so you might want to avoid them The other format is one designed for the QTL Cartographer system It is defined in the file cross inp included with the distribution and outlined in Section 6 3 2 Finally Reross can read files in it s own output format filetype Rcross out for translation to map maker raw or cross inp filetype formats
56. Ge A 59 Examples of Interim Files for Model 6 61 Command Line Options for JZmapqtl 65 Command Line Options for Egtl tods dd a dc ets 71 Command Line Options for Preplot 73 Filename extensions for Preplot output 73 Timings for Interval Mapping es 10 amare in ne Ee ce mes 93 Timings for Composite Interval Mapping 94 11 June 22 2000 LIST OF TABLES 12 Chapter 1 Introduction 1 1 General Overview QTL Cartographer is a suite of programs for mapping quantitative trait loci QTLs onto a genetic linkage map The general experimental paradigm begins with a pair of inbred parental lines that differ in the trait of interest and in the set of marker genotypes The programs use linear regression interval mapping Lander and Botstein 1989 or composite interval mapping Zeng 1993 Zeng 1994 methods to dissect the underlying genetics of the quantitative traits Mapping is done onto a set of linked genetic markers with known recombination frequencies Genetic linkage maps and data files can be imported from MAPMAKER Lander et al 1987 The mapping program uses a dynamic algorithm that allows a host of statistical models to be fitted and compared including various gene ac tions additive and dominance OTL environment interactions and close linkage This package consists of several programs written in C to perform various tasks
57. IX MAN PAGES QTL Cartographer SEE ALSO Rmap 1 Rqtl 1 Reross 1 Ostats 1 LRmapqtl 1 SRmapqtl 1 JZmapqtl 1 Zmapqtl 1 Prune 1 Preplot 1 OTLcart 1 AUTHORS In general it is best to contact us via email basten statgen ncsu edu Christopher J Basten B S Weir and Z B Zeng Department of Statistics North Carolina State University Raleigh NC 27695 8203 USA Phone 919 515 1934 133 June 22 2000 CHAPTER 8 UNIX MAN PAGES 134 Bibliography Basten C J B S Weir and Z B Zeng 1994 Zmap a QTL cartographer In C Smith J S Gavora B B J Chesnais W Fairfull J P Gibson B W Kennedy and E B Burnside Eds Proceedings of the 5th World Congress on Genetics Applied to Livestock Production Computing Strategies and Software Volume 22 Guelph Ontario Canada pp 65 66 Organizing Committee 5th World Congress on Genetics Applied to Live stock Production Carter T C and D S Falconer 1951 Stocks for detecting linkage in the mouse and the theory of their design J Genet 50 307 323 Churchill G A and R W Doerge 1994 Empirical threshold values for quantitative trait mapping Genetics 138 963 971 Cockerham C C and Z Zeng 1996 Design III with marker loci Genetics 143 1437 1456 Doerge R W and G A Churchill 1996 Permutation tests for multiple loci affecting a quantitative character Genetics 142 285 294 Doerge R W Z Zeng and B S Weir
58. June 22 2000 CHAPTER 1 INTRODUCTION Gershon Elber and many others For more information on GNUPLOT see the documentation that comes with the program 1 3 4 LINPACK Copyright Information We have translated some of the FORTRAN procedures of LINPACK Dongarra et al 1979 into C We have used all of the basic linear algebra subroutines BLAS as well as subrou tines to do the QR factorization of the matrix X of the linear system g X b These include the subroutines SORST STRSL SPODI SORSL and SORDC Not all of the optimizations have been translated These subroutines are used quite extensively in the analysis modules The original FORTRAN subroutines are Copyright C 1979 by the So ciety for Industrial and Applied Mathematics 1 3 5 Numerical Recipes in C Information We have made extensive use of the ideas from Numerical Recipes in C Press Flannery Teukolsky and Vetterling 1988 The source file Utilities c contains subroutines for allo cating memory that are derived from the functions for creating arbitrary offset vectors and matrices in Appendix D We have also used modified versions of the subroutines listed in Table 1 2 The original subroutines are Copyright C 1987 1988 Numerical Recipes Software subroutine section indexx 8 3 moment 13 1 sort 8 2 gammin 6 1 gammp 62 gasdev 7 2 gct 6 2 gser 6 2 poidev 7 3 betai 6 3 beta 6 1 betacf 6 3 Table 1 2 Subroutines
59. QTL Cartographer Version 1 14 Christopher J Basten Bruce S Weir Zhao Bang Zeng June 22 2000 OTL Cartographer OTL Cartographer A Reference Manual and Tutorial for QTL Mapping Christopher J Basten Bruce S Weir and Zhao Bang Zeng Program in Statistical Genetics Department of Statistics North Carolina State University QTL Cartographer Copyright 2000 by Christopher J Basten Bruce S Weir and Zhao Bang Zeng Program in Statistical Genetics Department of Statistics North Carolina State University Raleigh NC 27695 8203 All rights reserved Reproductions for personal use are allowed Anyone wishing to reproduce this book in whole or in part by any means for profit must first obtain permission from the authors Printed in the United States of America Typeset in TEX2e on a Macintosh G3 using Textures version 1 8 from Blue Sky Research Inc Contents List of Figures List of Tables 1 Introduction 1 1 1 2 1 3 1 4 1 5 1 6 2 1 General Overview esihe gege eg d e La 1 1 1 Definition of the Problem T 1 2 Experimental Westen ie Dace typed Gad AN 1 1 3 Genetic Linkage Maps ef at PSO uM GOAN Sw ee Programming Philosophy cita is a a A Len rent Copyright Information and Acknowledgments 13 1 QTL Cartographer Copyright Information 1 3 2 Citing QTL Cartographer cima du 4 a e e a ea 133 lt Gnuplot Copyright Information lila Dd 1
60. QTLcart would be specified by helpfile HardDrive QTLcart qtlcart hlp Be aware that UNIX systems are senstive to the case of the filenames and directories whereas Macintoshes and PCs running MS Windows are not If the program can t find the helpfile then you will be prompted for its location WORKING DIRECTORY You can specify a working directory or folder with the W option This directory folder must exist prior to running any of the programs The directory can be relative or complete and should have the standard directory delimiter appended to it For example W home user qtlcart work would use home user qtlcart work as the working directory All input and ouput files would have to be in this directory For a Windows system the line might be W c gtlcart work whereas a Macintosh would require W HardDrive qtlcart work The equivalent line in the resource file would have workdir instead of just W In UNIX you can set a path variable pointing to the programs and simply set your cur rent working directory to the working directory For Mac you double click the icons and should use a working directory variable Relative paths are also possible For example if the programs reside in a bin folder in the gtlcart folder on a Macintosh then you can have a data folder in the gtlcart folder and use W data as the working directory The two colons mean go up one level and then go into the data folder FILENAME STEM
61. R EXP Lander et al 1987 Lincoln et al 1992 Genetic marker order and chro mosome assignment may be accomplished using MAPMAKER EXP Once map order is established chromosomes may be saved to external files using the following MAP MAKER EXP commands in MAPMAKER EXP make chromosome cl seg M1 M2 M5 M4 M3 attach cl framework cl A chromosome cl is defined and the marker order for example M1 M2 M5 M4 M3 assigned The attach and framework commands tell MAPMAKER EXP to save this marker order on chromosome cl See Section 5 8 1 for a more detailed example of using MAPMAKER EXP to create the genetic linkage map After all chromosomes are defined and marker order assigned exit MAPMAKER EXP You will find files in your directory with the extensions data maps traits xmaps The raw file contains the original genotype and phenotype information The maps file contains the saved marker order per assigned chromosome as well as the estimated recombination fractions between each marker in the established order On MS DOS machines the extension may be map rather than maps It would be a good idea to rename this file with a mps ending so as not to confuse QTL Cartogra pher with its own genetic linkage map file The map order chromosome and recombination fraction estimate information may be used in QTL Cartographer by specifying maps as the input file
62. R files 6 1 2 Rmap input files The general method of inputting data for this format is by tokens Tokens are just collec tions of characters surrounded by whitespace spaces carriage returns tabs line feeds The maximum length of any token must be less than 64 and this may be increased in the future The following file also has commands embedded into it Rmap recognizes any token that begins with a minus sign as an embedded command Some commands require that the following token be a number or piece of information The following table gives a list of tokens that the program recognizes their purpose and what the next token should be e type Defines what the distances will be The token following this command must be either positions or intervals The latter indicates that the numbers are for the interval distance after a marker while the former indicates a position from the left telomere 85 June 22 2000 CHAPTER 6 INPUT FILE FORMATS function Defines a mapping function It can take on the values 1 2 or 3 for the Haldane Kosambi and Complete interference functions respectively Units Indicates the units of the distances Valid tokens following this command are cM M or r for centiMorgans Morgans or recombination probabilities Case is important chromosomes Indicates the number of chromosomes The following token must be an integer equal to the number of chromosomes in the map maximum Should be followed by an
63. RF crosses The default of 1 is fine for crosses in which there are only two marker genotypic classes backcrosses and recombi nant inbreds For SF and RF values of 30 31 or 32 are valid Recall that we have the following hypotheses 1 Ho a d 0 2 H a 0 d 0 3 H a 0 d 0 4 H3 a40 d40 66 CHAPTER3 ANALYSIS QTL Cartographer For 30 we test H3 Ho For 31 we test Hs Ho H3 H and H Ho For 32 we test H Ho Hs H and Hz Ho 30 is probably fine for initial scans Also if you do only have two genotypic classes then 10 is the same as 1 for the hypothesis test Model 6 For Model 6 be sure to run SRmapqtl first Once done JZmapqtl will use all markers that are significant for any of the traits in the analysis We need to work out a better way to select the cofactors Now it uses any markers that are significant for any trait Also be sure to use FB regression or else you will end up using all markers as cofactors G x E Analysis One special case of G x E analysis has been incorporated into JZmapqtl namely the situ ation where a set of genotypes is raised in more than one environment The value of the trait in each environment is treated as a separate trait for the common genotype For this type of data use hypothesis 14 or 34 to invoke the G x E analysis Hypothesis 14 is for data with two marker genotypes while 34 is for three marker genotypes There will be an extra column in the output that give
64. Rmapqtl and run it without changing any parameters Look at the output sim lr Start up SRmapqtl You might want to change the analysis model from its default value of 0 forward stepwise regression to 2 forward regression with backward elimination Run it and look at the output sim sr Start up Zmapqtl You won t need to change any parameters Tell it to go ahead with the analysis and look at the output sim z Start up Zmapqtl again This time choose Model for Analysis and change it to 6 Tell it to go ahead with the analysis and look at the output which will be appended to what you did in the first run Start up Preplot Don t change any parameters Go ahead with the program 79 June 22 2000 CHAPTER 5 TUTORIAL EXAMPLES 10 If you are on a Macintosh move the GNUPLOT binary into the working subdirec tory and double click it If on PC or UNIX machine start up GNUPLOT From the GNUPLOT command line type in load sim plt If you are on a PC you may need to go through the file menu and search for the sim plt file This should display graphical results Press returns when requested 11 Start up Eqtl Go ahead with the analysis Look at the output sim eqt 5 6 Analyzing simulated data Create a working subdirectory call it mletest and copy the simulated data sets into it The simulated datasets called mletest map and mletest cro
65. Rqtl Input is token based and the data file has embedded commands to indicate to Rcross what it is reading The first line of the data file should contain a pound symbol and a long integer for example 123456787 filetype cross inp The number will be an identifier for the file and should be unique to this file In addition the filetype cross inp token helps Rcross determine what type of file it is reading Here is a list of embedded commands e skip indicates that Rcross should skip all tokens until an unskip token is read e unskip see above e Cross Should be followed by the type of cross See Table 1 1 for valid tokens e traits should be followed by the number of traits that have numerical values e otraits should be followed by the number of other traits that is those with character or string values Examples would include sex or brood e SampleSize would be followed by the sample size e case should be followed by yes or no depending on whether the names of marker systems are case sensitive With no all names of individuals markers and traits are converted to lower case to make comparisons e TranslationTable will allow one to define a table to translate marker values After this command a small table of six rows and three columns must follow The first two columns should match exactly the example given below and the third column can be whatever your data set is encoded as e missingtrait followed by a tok
66. This requires a filename for output Rqtl will overwrite the file if it exists and create a new file if it does not If not used then Rqtl will use gtlcart qtl i This requires an input filename This file must exist Rqtl will attempt to identify the format of the file and translate it to another format This file should contain a genetic model defining a set of QTL and including their positions and effects See the file gtls inp for the format m This requires a filename that must exist Rqtl will read the genetic linkage map from this file t This allows the user to specify the number of traits to simulate Itis 1 by default q This requires an integer argument It allows the user to specify the number of QTL that affect the trait If one trait is simulated then exactly this number of QTL will be created If more than one trait are simulated then the number of QTL per trait will vary but have mean value specified here The default is 9 106 CHAPTER 8 UNIX MAN PAGES QTL Cartographer d You can specify the type of dominance at the trait loci If we assume inbred parental lines with line one marker trait alleles all Q and line two trait alleles all q then use a 1 for no dominance a 2 for complete dominance of Q over q a 3 for complete domi nance of q over Q and a 4 for dominance that is random in direction and magnitude for each locus It is 1 by default that is no dominance b Specifies the parameter needed to determine
67. W path for a working directory R file to specify a resource file e to specify the log file s to specify a seed for the random number generator and X stem to specify a filename stem The options below are specific to this program If you use this program without specifying any options then you will get into a menu that allows you to set them interactively o This should be used with a filename indicating where the output will be written Rmap will overwrite the file if it exists and create a new file if it does not If not used then Rmap will use gtlcart map i You can use this option to specify an input filename This file must exist and have one of three formats Rmap out map inp or mapmaker mps Rmap will attempt to identify the format of the file and translate it to another format If you specify an input file then the simulation parameters will be ignored 102 CHAPTER 8 UNIX MAN PAGES QTL Cartographer g Requires an integer to indicate the output format You can use a 1 for the default ouput format a 2 for GNUPLOT output or a 3 for both If you use a 2 or a 3 then you can use GNUPLOT to see a primitive looking linkage map f Requires an integer option to specify the mapping function Rmap can use the Haldane Kosambi fixed or a number of other functions The default is to use the Haldane function which is specified with a 1 Using a 2 invokes the Kosambi mapping func tion 3 means that a fixed function is used an
68. X MAN PAGES EXAMPLES o Qstats i corn cro m corn map Calculates basic statistics on the dataset in corn cro using the genetic linkage map in corn map The program will display and interactive menu for setting options and print out messages to the screen while running These can be turned off with A and V respectively If the dataset in corn cro has more than one trait then all traits will be analyzed REFERENCES 1 M Lynch and B Walsh 1998 Genetics and Analysis of Quantitative Traits Sinauer Associates Sunderland MA BUGS Are there any other statistics that we can do Your suggestions are welcome SEE ALSO Rmap 1 Rqatl 1 Reross 1 LRmapqtl 1 SRmapqtl 1 Zmapqtl 1 JZmapqtl 1 Eqtl 1 Prune 1 Preplot 1 OTLcart 1 AUTHORS In general it is best to contact us via email basten statgen ncsu edu Christopher J Basten B S Weir and Z B Zeng Department of Statistics North Carolina State University Raleigh NC 27695 8203 USA Phone 919 515 1934 116 CHAPTER 8 UNIX MAN PAGES QTL Cartographer 8 7 LRMAPOTL NAME LRmapqtl Single marker QTL analysis SYNOPSIS LRmapqtl o output iinput m mapfile rreps t trait DESCRIPTION LRmapqtl uses simple linear regression to map quantitative trait loci to a map of molecular markers It requires a molecular map that could be a random one produced by Rmap or a real one in the same format as the output of Rmap The
69. a data file If markers had been eliminated then the linkage map is regenerated to take this into account The new output files will have the extensions mpb and crb and filename stems specified by the o option 42 CHAPTER 2 SIMULATING REFORMATTING DATA QTL Cartographer Option Default Explanation 0 qtlcart Output Filename Stem e qtlcart log Error File m qtlcartmap Genetic Linkage Map File i qtlcart cro Data File S 860436420 Random Number Seed I 1 Interactive mode 0 1 no yes b 0 B 1 P 2 M 3 or D 4 M 0 100000 Percent missing data to simulate Table 2 5 Command Line Options for Prune 2 4 1 Pruning Datasets The pruning of datasets occurs in an interactive menu After setting parameters in the first menu continue on to the second interactive menu where actions can be taken The second interactive menu looks like this You can loop through items 1 6 but 7 8 and 9 terminate Z O D OH I oO O amp amp N ma ea i Action Eliminate A marker systems P1 dominant Eliminate a marker systems P2 dominant Eliminate marker m on chromosome c Eliminate trait t Eliminate individuals with missing phenotypes for trait t liminate individuals with more than m missing markers Bootstrap the data Permute the traits in the data Simulate m missing markers Write modified dataset and exit Exit without writing anything Help Pick a number to do an
70. a likelihood ratio for a G x E effect versus no effect When running Eqtl subsequent to doing a G x E analysis be sure to specify the same hypothesis test 67 June 22 2000 CHAPTER 3 ANALYSIS 68 Chapter 4 Visualization of Results The final step in analyzing your data will be to summarize your results either graphically or as a compact set of estimates for QTL positions and effects We have provided some utilities that read the output of the analysis programs and reformat it for use in graphics packages The freeware program GNUPLOT is recommended as a graphics engine but the results could be plotted in any plotting package on any machine All of the results from the analysis programs are simple text files and all the reformatted files are also simple text Figure 4 1 is a schematic of the programs and files that are involved in this step Eqtl is a utility that quickly picks out the possible QTLs from the results of Zmapqtl Preplot can read the output of Rqtl LRmapqtl and Zmapqtl and produce simple files containing two columns of text corresponding to the values for the abscissa and ordinate of a plot These files in turn can be plotted by GNUPLOT or imported into various plotting packages on various platforms 41 Eqtl Zmapqtl outputs a great deal of information Often the experimenter will want a quick summary of the positions and effects of the QTLs The program Eqtl scans the output of Zmapqtl and reformats it Part of the
71. ahead with the analysis Look at the output sample cro 3 Proceed with the analysis programs as in the previous examples Run Qstats LRmapqtl SRmapqtl and Zmapqtl Look at the output after each run 82 CHAPTER 5 TUTORIAL EXAMPLES QTL Cartographer 4 Start up Preplot Don t change any parameters Go ahead with the program 5 Start up GNUPLOT From the GNUPLOT command line type in load sample plt This should display graphical results 6 Start up Eqtl Go ahead with the analysis Look at the output sample eqt 83 June 22 2000 CHAPTER 5 TUTORIAL EXAMPLES 84 Chapter 6 Input File Formats All of the input and output files in the QTL Cartographer system are plain text and can thus be viewed by virtually any text editor or word processor on any platform The input files for many of the programs will have embedded commands that start with a minus sign Care should be taken not to have stray tokens such as Chromosome in input files Also the case of commands is generally very important When in doubt use the exact case that is specified here 6 1 Genetic Linkage Maps 6 1 1 MAPMAKER output files Rmap can translate the output of MAPMAKER into the format required by the QTL Car tographer system Use the maps file that is the output of MAPMAKER as the input to Rmap and it will be translated automatically An alternate format has been designed for those who don t have the MAPMAKE
72. ains the man pages is defined in the environmental variable MANPATH then you can get the online help with a command such as man Rmap We provide html versions of the man pages on the web server for Macintosh and Windows users If you have World Wide Web access first point your browser to our home page http statgen ncsu edu Then go down about halfway until you get to the QTL Cartographer link Follow it to the online man pages You can also access the rest of the QTL Cartographer manual The man ual is written in TEX2e and has been translated into HTML by the program html2latex The complete set of man pages are reprinted here for your benefit Here follow the BIEX formatted versions of the man pages Since the documentation will change regularly it is a good idea to check the Web site for the current online manual The Web pages will always be updated with the manual updates 95 June 22 2000 CHAPTER 8 UNIX MAN PAGES 8 1 OTLCART NAME QTLcart A rudimentary front end for the QTL Cartographer system SYNOPSIS OTLcart h V A sseed W workdir X stem e logfile R resource DESCRIPTION OTLcart does not actually exist It is intended to be the front end to a set of programs collectively known as QTL Cartographer This man page explains the options that are valid in all the programs of the OTL Cartographer suite It also outlines how to get started using the programs OPTIONS The fo
73. alues for quan titative trait mapping that we have implemented in this program Basically it does a per mutation of the trait values and the genotypes and redoes the analysis Over the number of replicates two types of thresholds are defined experimentwise and comparison wise We calculate the experimentwise thresholds but only give p values for the compar isonwise values to save on storage space The p values give the proportion of permuted replicates that have loglikelihood ratios larger than the observed ratios 123 June 22 2000 CHAPTER 8 UNIX MAN PAGES If you choose to do permutation tests you need to run Zmapqtl with the model of choice prior to doing the permutation test Also if the program terminates prematurely you can restart it from where it left off to complete the permutation test REFERENCES 1 Churchill G A and R W Doerge 1994 Empirical threshold values for quantitative trait mapping Genetics 138 963 971 2 Lander E S and D Botstein 1989 Mapping Mendelian factors underlying quanti tative traits using RFLP linkage maps Genetics 121 185 199 3 Zeng Zhao Bang 1993 Theoretical basis for separation of multiple linked gene ef fects in mapping quantitative trait loci Proc Natl Acad Sci USA 90 10972 10976 4 Zeng Zhao Bang 1994 Precision mapping of quantitative trait loci Genetics 136 1457 1468 BUGS It is likely that we will abandon the internal permutation te
74. ams 22 CHAPTER 1 INTRODUCTION QTL Cartographer a Which program is giving you trouble and what parameter values were used b Are the input files simulated or real c Would it be possible to send me the input files the log file and the resource file qtlcart rc d When the program crashed did it give any diagnostics e When did you download the programs f What is the version number This is valid for programs downloaded after 1 January 1996 and supersedes the previous question When reporting a problem try to include the answers to all of the questions above Some of them may not be relevant for you particular case and can be ignored Email is generally the best way to report problems as the messages stay on a queue until they are dealt with One of the most difficult steps in using the QTL Cartographer system is to reformat datasets Question 2 c above asks whether you would be willing to send us your data in order to diagnose a problem We would like to emphasize that if you send us your data files they will be kept in the strictest confidence Data files sent to us are stored on a machine which cannot be accessed by the network We will also delete your data files upon your request 1 5 3 Contacts For any other problems with QTL Cartographer contact Christopher J Basten via any of the methods listed in Table 1 3 In general email is the best method for indicating a problem Chris may not always get back t
75. an download them using some web browsers The entire manual as well as the man pages have been been translated into html 75 June 22 2000 CHAPTER5 TUTORIAL EXAMPLES 5 2 Basic Macintosh The MacOS is so easy to use that little instruction is necessary I would recommend getting a copy of BBEdit Lite for viewing and editing text files It is freeware and can open large files as long as you have the memory BBEdit Lite can also view and convert text files with DOS UNIX or Macintosh line endings Other free programs such as Fetch to download files Telnet 2 7 to access UNIX servers and Acrobat Reader to view and print documents are also useful 5 3 Basic Windows This is a quick summary of some basic commands and techniques for working in the Win dows NT environment Other versions of Windows from 95 up should be similar Logging in Using control alt delete will bring up the login screen Click in the login box and type your login name Press tab to get to the password box and type your pass word and a return If you then see a timer which looks like a little clock you ll know you have succeeded Just wait while the windowing system starts up Logging out You may want to empty the trash before you log out Right click on the recycle bin and select empty to do so When you want to log out simply click the left mouse button on the Start icon and select shutdown This will bring up a menu Select the Close all programs
76. and line options More recently we have added an interactive menu that allows the user to set parameters Once inside any of the programs all the parameters of the program are displayed with their current values The user chooses whichever parameter he or she wishes to change by selecting a number The menu is in a loop Choosing 0 will end the loop and proceed with the current parameter values The menu is also where one can get online help Online help will be a numbered option in the list of parameters Choose it and specify the location of the help file if the program couldn t find it When the programs begin to run they will print out their parameter values to a log file gtlcart log by default Here is an example of the Qstats menu 97 June 22 2000 CHAPTER 8 UNIX MAN PAGES No Options Values O Continue with these parameters 1 Data Input File qtlcart cro 2 Output File atlcart qst 3 Error File atlcart log 4 Genetic Linkage Map File qtlcart map 5 Random Number Seed 961681144 6 Specify Resource File atlcart rc 7 Change Filename stem qtlcart 8 Help 9 Change Working Directory 10 Quit Quit but update the Resource File Please enter a number This menu is in a loop To change a parameter select its number and press return You will be prompted for a new value or filename You can clear out a filename or working directory by inputting a single period When satisfied
77. are created at the best estimates for the positions of the QTLs Zmapqtl Model Six Model 6 requires two additional parameters One is the number of markers to control for the genetic background np and the other is a window size ws When invoked the program will read in the results of a prior run of SRmapqtl to pick the most important markers to control for the genetic background Then when testing at any point on the genome it will use up to n of these markers If SRmapqtl didn t rank as many markers 58 CHAPTER3 ANALYSIS QTL Cartographer as specified with n then n is reset to the number of markers ranked The window size will block out a region of the genome on either side of the markers flanking the test site Since these flanking regions are tightly linked to the testing site if we were to use them as background markers we would then be eliminating the signal from the test site itself Note that if ws 0 0 and n equals the total number of markers then Model 6 reduces to Model 1 If ws is large say the size of the largest chromosome and np equals the number of markers then Model 2 is the result If n is zero then Model 3 is the result In the future we will recommend that people use model 3 or model 6 for analysis The default values of 5 for ny and 10 for w should be good starting points for Model 6 Increasing np will allow better resolution for mapping linked QTLs 3 43 Zmapqtl Options Table 3 4 shows the comma
78. atio test statistics of the dataset in corn cro using the map in corn map Model 6 is used for analysis and a permutation test with 500 replications is performed The program is nice d as a courtesy to other users and run in the background so that the user can logout and relax MODELS Different parameters for the M option allow for the analysis of the data assuming different models Models 1 3 were described in Zeng 1993 1994 1 Fit all the background markers 2 Fit all unlinked background markers 3 Fit only the mean Lander and Botstein 1989 method 4 Fit a subset of the other markers namely those unlinked markers with the highest correlation with the trait on each chromosome 5 This model uses a pair of markers from each other chromosome and all linked mark ers that fall outside a window around the flanking markers This window extends to 10 cM beyond the markers immediately flanking the test position The window size can be changed with the w option 6 This model uses a specified number of markers that fall outside a window around the flanking markers This window extends to 10 cM beyond the markers immediately flanking the test position The number of markers are set by the n option You need to run SRmapqtl to rank the markers before using model 6 The default is to fit only the mean that is to use interval mapping PERMUTATION TESTS Churchill and Doerge 1994 describe a method to calculate the threshold v
79. atistics are redone and printed out See the section Prune as to how to do bootstrapping 60 CHAPTER3 ANALYSIS QTL Cartographer Interim file Created during Contains qtlcart z6e permutation test Experimentwise state qtlcart z6c permutation test Comparisonwise state qtlcart z6a bootstrap resampling Iteration i bootstrap qtlcart z6b bootstrap resampling Iteration i 1 bootstrap qtlcart z6i jackknife resampling Iteration i jackknife qtlcart z6j jackknife resampling Iteration 1 jackknife Table 3 5 Examples of Interim Files for Model 6 Jackknife resampling is performed by calculating n the sample size new estimates of the parameters The th estimate is calculated by deleting individual i from the dataset The standard deviation over these n new estimates provides an estimate of the standard de viation for the test statistic and additive and dominance effects You invoke the Jackknife by setting the number of bootstraps to 2 Zmapqtl uses two interim files to perform the jackknife If you are using Model 6 in Zmapqtl and your filename stem is qtlcart then these files will be called qtlcart z6i and qtlcart z6j These files contain the sum and sum of squares up to the previous and current iteration as Zmapqtl runs Initially the qtlcart z6i file contains columns of zeros This is the sum before any iterations are performed Subse quently qtlcart z6j will contain the interim state after each odd numbered iteration
80. ber of traits then all traits will be analyzed The default is to analyze trait 1 only 117 June 22 2000 CHAPTER 8 UNIX MAN PAGES MODEL The basic linear model is Trait Mean Slope x Marker Error The marker value will be in the range 1 1 inclusive Two hypotheses are compared The null hypothesis is that the Slope is zero The alternate is that the Slope is non zero A p value for the likelihood ratio of these to hypotheses is calculated for each marker trait combination LRmapqtl outputs a table with parameter estimates F statistics Likelihood ratios and p values INPUT FORMAT The input format of the molecular map should be the same as that of the output format from the program Rmap The input format of the individual data should be the same as the output format of the program Rcross EXAMPLES 5 LRmapqtl i corn cro m corn map Calculates the regression coefficients for each marker on the dataset in corn cro using the genetic linkage map in corn map REFERENCES 1 Churchill G A and R W Doerge 1994 Empirical threshold values for quantitative trait mapping Genetics 138 963 971 BUGS SEE ALSO Rmap 1 Rqtl 1 Rcross 1 Ostats 1 SRmapqtl 1 Zmapqtl 1 JZmapqtl 1 Eqtl 1 Prune 1 Preplot 1 OTLcart 1 AUTHORS In general it is best to contact us via email basten statgen ncsu edu Christopher J Basten B S Weir and Z B Zeng Department of Statistics North Carolina State University Raleigh
81. bility I 0 Interactive flag 8 0 Output format E 1 000000 Environmental Variance used if gt 0 Table 2 4 Command Line Options for Rcross 2 3 1 Simulating Data Rcross will simulate a dataset using the genetic linkage map prepared by Rmap and the genetic model prepared by Rqtl The user can specify the sample size type of cross and heritability or environmental variance An interactive mode allows the user to generate arbitrary crosses Rcross can automatically generate backcrosses intercrosses or any of the other experimental designs defined in Section 1 1 2 Below we describe how each indi vidual is created The process is repeated as many times as are necessary to get the sample size specified 38 CHAPTER 2 SIMULATING REFORMATTING DATA QTL Cartographer Generation of Individuals For generating backcrosses or intercross samples the parental lines are known Individ uals in the F are all heterozygous and all pairs of loci are in coupling Samples derived from F gt and later crosses need to take into account the different possible parents This section explains how individuals are simulated in a general way We assume that there is one or two parental samples that will be used to create the next generation Refer to these as lines 1 and 2 We assume monoecious diploid individuals To generate a new individual one parent is selected from line 1 and one from line 2 If line 1 and line 2 are the same sample for exampl
82. cess stopped at 899 as above then restarting Zmapqtl with 1 000 permutations will begin with permutation 900 and continue to 1 000 The second file qtlcart z6e will contain two columns of numbers the permutation and the maximal likelihood ratio over the genome in that permutation Each permutation will add a line to the output When enough permutations have been done Eqtl can be run to summarize the experimentwise levels A small table will be written to the log file that looks like start Performed 899 permutations of the phenotypes and genotypes Here are the Experimentwise significance levels for different alpha Permutation significance level for alpha 0 1 11 6858 Permutation significance level for alpha 0 05 13 3108 64 CHAPTER3 ANALYSIS QTL Cartographer Permutation significance level for alpha 0 025 14 6669 Permutation significance level for alpha 0 01 16 8008 end of shuffling results For each shuffle the largest likelihood ratio test statistic over all test positions is saved in the file At the end of the shuffling these maximum values are sorted and the 1 a x 999 th largest is the experimentwise significance level for a test of size a The number of permutations can be changed from 899 to any integer from 0 to 10 000 This upper bound could be made higher by changing the appropriate definition in the Main h source file and recompiling In general we fin
83. chromosome 1 19 attach cl Attach the sequence to chromosome 1 20 framework cl Create the framework puts in distances for chromosome 2 21 make chromosome c2 Create chromosome 2 22 sequence 4118129610 Specify the sequence of markers on chromosome 2 23 attach c2 Attach the sequence to chromosome 2 24 framework c2 Create the framework puts in distances for chromosome 2 25 quit Exit the program The map will be in sample maps On a UNIX machine you will now have a file called sample maps On a PC it will be called sample map It will be one of these two on a Macintosh Rename this output file to sample mps and use it along with the sample raw file for the next part 5 8 2 Using the MAPMAKER files Create a new working subdirectory called mm in you qwork subdirectory Copy the sam ple files into it There should be two files sample mps and sample raw The former is a genetic linkage map created by MAPMAKER EXP The latter is MAPMAKER QTL raw file You will now translate the data files into the QTL Cartographer format and then analyze the data 1 Start up Rmap Select the option to change the filename stem Change the filename stem to sample and set the proper working subdirectory Then select the input file option and change it to sample mps Then go ahead with the analysis Look at the output sample map 2 Start up Rcross Select the input file option and change it to sample raw Then go
84. d for numerical consideration As x and z are separated from X X is unchanged in each iteration and its costly recalculation is avoided For an F gt population the hypotheses for testing are Ho b 0 and d 0 and Hz b 0 or d 0 This is performed through a likelihood ratio test procedure In addition it is possible to test hypotheses on b and d individually For a backcross data set dominance cannot be estimated and d is dropped from Equation 3 5 The trait will have a variance s Under the null hypothesis Hy Y XB E the sample variance of the residuals will be sj For a given alternative model say H Y x h 2 d XB E the variance of the residuals would be s With this in mind we can calculate the propor tion of variance explained by a QTL at the test site The quantity is usually called r and estimated by nr 2 _ 50 1 re 2 S An alternative estimate would use the total variance Denote it by 2 s s Tt s2 57 June 22 2000 CHAPTER 3 ANALYSIS r is the proportion of the variance explained by the QTL conditioned on the background markers and any explanatory variables r is the proportion of the total variance explained by the QTL and the the background markers and any explanatory variables Generally Eten re r 3 4 2 Models When we speak of models for analysis we mean to specify the markers used as cofactors in composite interval mapping There are presently six models fo
85. d that 1000 permutations is a sufficient number In a test values of 1000 and 17 000 were used with little difference in the ultimate comparisonwise and experimentwise values 3 5 JZmapatl JZmapqtl implements interval and composite interval mapping for multiple traits Jiang and Zeng 1995 It is very similar to Zmapqtl except that it can jointly analyze more than one trait It is best used after Zmapqtl when one suspects that two traits are correlated 3 5 1 JZmapqtl Options Table 3 6 shows the command line options specific to JZmapqtl Most are the same as those for Zmapqtl One thing to note is that there is no facility for permuation tests or bootstraps at this time Option Default Explanation i qtlcart cro Input File 0 qtlcart z Output File e qtlcart log Error File m qtlcart map Genetic Linkage Map File S qtlcart sr SRmapqtl results Model 6 E qtlcart eqt Eqtl results Model 7 S 893339277 Random Number Seed M 3 Model 3 6 7 3 gt IM t 1 Trait to analyze c 0 Chromosome to analyze 0 gt all d 2 000000 Walking speed in cM n 5 Number of Background Parameters Model 6 W 10 000000 Window Size in cM Model 6 I 1 Hypothesis test Table 3 6 Command Line Options for JZmapqtl 65 June 22 2000 CHAPTER 3 ANALYSIS 3 5 2 Output JZmapqtl will create a number of different output files depending on the number of traits in the joint analysis There will be one file per trait that has es
86. d thus the distance in Morgans is the recombination fraction The type of mapping function used would then be recorded in the ouput and all following analyses will use this function One must edit the map file to change this if not using Rmap p Requires a real number Some map functions need an extra parameter and this allows the user to specify it See the manual for details c This allows you to specify the number of chromosomes if you are simulating a genetic linkage map It is 4 by default If you are translating a file then this will be ignored as will the remaining options m This allows you to specify the average number of markers per chromosome in a simu lation The default is 16 vm This allows you to specify the standard deviation in the number of markers per chro mosome The number of markers per chromosome will have a normal distribution with mean given in the previous option and the standard deviation specified here If zero then each chromosome will have the same number of markers d Rmap uses the value given after this option as the average intermarker distance in centiMorgans for a simulation It is 10 centiMorgans by default vd The intermarker distance will have a normal distribution with mean set by the pre vious option and standard deviation specified with this option It is 0 0 by default which means that the intermarker distances between consecutive markers will all be the same Set it to a positive value
87. do a bootstrap experiment on a data set one might use the sequence of commands in the following shell script written for the C shell on a UNIX workstation 44 CHAPTER 2 SIMULATING REFORMATTING DATA QTL Cartographer bin csh Bootstrap csh Usage Bootstrap csh stem bootstraps email where stem is the filename stem permutations is the number of permutations and email is the user s email address Note This only works if you have set and used a filename stem and make sure that you don t use temp as your stem if 1 h then echo Usage Bootstrap csh stem model bootstraps email echo Where echo stem filename stem echo model Zmapatl analysis model echo bootstraps number of bootstraps echo email user s email address echo echo Now exiting exit endif set tlog temp log usr bin rm f tlog echo Bootstrap experiment started gt tlog usr bin date gt gt Stlog echo Stem 1 gt gt Stlog echo Model 2 gt gt tlog echo Reps 3 gt gt Stlog echo Email 4 gt gt Stlog set bindir usr local bin mv 1 log 1 logsave set i l while Si lt 3 Sbindir Prune A V i l cro b 1 gt gt amp Stlog nice Sbindir Zmapqtl A V M 2 i l crb b 1 m l map gt gt amp Stlog usr bin mv 1 z 2 b 1 z 2 a i end mv l logsave 1 log echo Bootstrap experiment ended gt gt Stlog usr bin dat
88. e crossing two F lines to form an F3 then selfing is a possibility Once the parents have been selected gametes are produced one from each parent The first step in producing gametes is to simulate recombination We assume that the number of crossovers on each chromosome is distributed as a Poisson random variable with mean equal to the length of the chromosome in Morgans A separate random integer is generated for each chromosome subject to the Poisson and this indicates the number of crossovers on that chromosome These crossovers are placed on the chromosome subject to a uniform distribution Once the crossovers are in place gametes are generated Starting with the first chromo some one of the two homologs is chosen at random This chromosome is followed until a crossover is encountered at which point the other homolog is used At the end of the first chromosome a homolog from the second chromosome is chosen at random and the process continues At the end a gamete is created which contains the markers and QTLs The gametes from each parent are then combined to form a new individual Phenotypic values can then be generated Phenotypic Values Phenotypic values are calculated from the genotypic values for each individual for each trait Each individual s phenotypic value is calculated from its genotypic value with an environmental effect determined by the heritability h The individual s genotypic value is based on the alleles it inhe
89. e gt gt Stlog usr ucb mail 4 lt tlog Note that the work is done in the while end loop For each repetition a bootstrapped data set is created with Prune This data will be placed in the file ending with crb Zmapqtl then analyzes this bootstrapped data and updates a file with the sum and sum of squares of the test statistic and estimates of effects You will need to have run Zmapqtl on the original data before doing the bootstrap When this is finished you can run Eqtl to get the mean and variance of the likelihood ratio additive effect and dominance effect at each test site 45 June 22 2000 CHAPTER 2 SIMULATING REFORMATTING DATA An alternate method of performing the bootstrap is similar to the above except that you omit the b 1 flag In this mode the results will be appended to the z file and you would need a script to calculate the means and variances at the end of the run Permutation Tests Zmapqtl can perform permutation tests using interval mapping but if you want to do a proper permutation test using composite interval mapping and reselecting your back ground markers during each permutation you will need to do it in a batch file similar to the one for bootstrapping Prune can create a single permuted dataset by using the b option with a value of 2 Permutation tests are then done as were the bootstrap in the previous example Here is a UNIX shell script example for a permutation test bin csh Per
90. e 22 2000 CHAPTER 8 UNIX MAN PAGES 8 2 RMAP NAME Rmap Simulate or reformat a map of molecular markers SYNOPSIS Rmap o output iinput g gmode f mapfunc p mapparam c chroms m MarkersPerChrom vm sd MPC d InterMarkerDist vd sdIMD t Tails M Mode DESCRIPTION Rmap creates a random map of molecular markers The user specifies the number of chro mosomes the number of markers per chromosome and the average intermarker distance If one specifies standard deviations for the number of markers and the average intermarker distances they will vary subject to the normal distribution The output gives a table of markers by chromosomes with the distances between consecutive markers in centiMor gans in the table If you specify an input file Rmap will open it determine if it is in the same format as Rmap outputs and process it based in the value given to g If the input file is the output of MAPMAKER then the map will be reformatted from MAPMAKER into the Rmap output format Finally there is a standard input format that Rmap can translate and is defined in the file map inp that comes with the distribution of the programs Note that if the user specifies an input file no simulations will be done and the latter half of the command line options are ignored OPTIONS See QTLcart 1 for more information on the global options h for help A for automatic V for non Verbose
91. e an F1 generation All crosses are then derived from these lines Backcrossing to P1 is encoded by B1 and to P2 by B2 Selfed intercrosses of generation i are encoded by SFi Randomly mated intercrosses of generation i are encoded by RFi Recombinant inbreds created by selfing have the code RI1 while those by sib mating are RI2 Doubled haploids have the code RIO A test cross of an SFi line to a Pj line is encoded by T Bj SFi The QTL Cartographer manual explains some other crosses that are possible Note that the UNIX shell may interpret and so they should either be quoted or the cross entered into the interactive menu EXAMPLES Rceross A V c SF2 n 1000 Does a selfed F2 cross with 1000 offspring using the linkage map in gtlcart map and the model in gtlcart qtl The command line options A and V turn off the interactive menu and the verbosity mode respectively 110 CHAPTER 8 UNIX MAN PAGES QTL Cartographer o Rcross i Cross raw Reads from the file cross raw tries to determine its format and translates it if possible The file cross raw could be a MAPMAKER QTL formatted file a cross inp formatted file or one that is already in the Rcross out format REFERENCES 1 Lander E S P Green J Abrahamson A Barlow M Daley S Lincoln and L New burg 1987 MAPMAKER An interactive computer package for constructing pri mary genetic linkage maps of experimental and natural populations Genomics 1 174 181
92. e gcc you might try cc Finally you will want to set the install directory By default it is BINDIR usr local bin but you can change it to whatever you wish Note that to install the programs in the install subdirectory you will need write per missions for that subdirectory 3 Change into the root directory of the distribution and make the programs o make install 4 The binaries will be in the BINDIR subdirectory Make sure that this subdirectory is in your path variable and then rehash Presently we use gcc version 2 95 2 on our Sun Workstations under Solaris 2 7 If you have troubles compiling you may need to update your operating system or compiler If you would like a hardcopy of the man pages you can either cd to the doc subdirectory and send the postscript files to a postscript printer or cd to the man subdirectory and is sue the make hardcopy command The second method requires that the program a2ps be installed on your system and that the default printer be able to handle postscript Alter natively you can print the files ending in pdf using Adobe Acrobat Reader which is freely available from Adobe http www adobe com 1 4 3 Macintosh You will need a Macintosh with a power pc chip Download the file OTLCartMac sea hqx Use StuffitExpander or BinHex4 to unbinhex the self extracting archive Double click the OTLCart sea file to unpack the binaries and supplemental files Some programs such as Netscape
93. e have the following hypotheses 1 HO a d 0 2 Hl a 0 d 0 3 H2 a 0 d 0 4 H3 a 0 d 0 For 30 we test H3 H0 For 31 we test H3 H0 H3 H1 and H1 H0 For 32 we test H3 H0 H3 H2 and H2 H0 30 is probably fine for initial scans Hypothesis 34 does a test for H3 H0 as well as the GxE For Model 6 be sure to run SRmapqtl first Once done JZmapqtl will use all markers that are significant for any of the traits in the analysis We need to work out a better way to select the cofactors Presently we use any markers that are significant for any trait Also be sure to use FB regression Model 2 in SRmapqtl or else you will end up using all markers as cofactors SEE ALSO Rmap 1 Rqtl 1 Rcross 1 Ostats 1 LRmapqtl SRmapqtl 1 Zmapqtl 1 Eqtl 1 Prune 1 Preplot 1 OTLcart 1 AUTHORS In general it is best to contact us via email basten statgen ncsu edu Christopher J Basten B S Weir and Z B Zeng Department of Statistics North Carolina State University Raleigh NC 27695 8203 USA Phone 919 515 1934 128 CHAPTER 8 UNIX MAN PAGES QTL Cartographer 8 11 PREPLOT NAME Preplot Process results of LRmapqtl and Zmapqtl for input to gnuplot SYNOPSIS Preplot o output m mapfile llrfile z zfile q gtlfile S threshold T terminal H hypo L lod DESCRIPTION Preplot reformats the output of LRmapqtl and Zmapqtl so that it can be plotted by GNU PLOT It require
94. e traits both within and between populations or species The recent advances in molecular biology have al lowed the construction of genetic linkage maps based on molecular markers Such genetic linkage maps can span the genome at regular intervals The experimenter can then look for correlations between these mapped markers and the trait of interest in controlled breeding experiments to gain insight into the regions of the genome that control the trait 1 1 2 Experimental Design The paradigm for the programs in the QTL Cartographer package is that of highly inbred lines with very little genetic variation within lines but variation between lines We shall refer to these inbred lines as parental lines and denote them by the symbols P and P As a general rule the P lines will correspond to the high lines with respect to the trait of interest that is they will have mean values larger than the P or low lines These parental lines can be crossed to produce Fi lines which are heterozygous for both markers and QTLs One can then cross the F populations with either parental line to produce backcrosses The symbols B and B will refer to backcrosses involving the P and P lines respectively Alternatively the F lines can be intercrossed to produce F lines In each of these cases the resultant lines will have variation in both the trait of interest and the underlying quantitative trait loci and marker genotypes These crosses are illust
95. ed are on separate chromosomes the order is unimportant If two markers from the same chromosome are to be eliminated order should be to eliminate the highest numbered marker The same concept holds for traits eliminate them in the order of highest to lowest Do not try to eliminate any markers or traits AND do a bootstrap permutation or simula tion of missing markers in the same run SEE ALSO Rmap 1 Rqtl 1 Reross 1 Ostats 1 LRmapqtl 1 SRmapqtl 1 JZmapqtl 1 Eqtl 1 Preplot 1 OTLcart 1 113 June 22 2000 CHAPTER 8 UNIX MAN PAGES AUTHORS In general it is best to contact us via email basten statgen ncsu edu Christopher J Basten B S Weir and Z B Zeng Department of Statistics North Carolina State University Raleigh NC 27695 8203 USA Phone 919 515 1934 114 CHAPTER 8 UNIX MAN PAGES QTL Cartographer 8 6 OSTATS NAME Qstats Calculate basic statistics for a QTL dataset SYNOPSIS Ostats o output iinput m mapfile DESCRIPTION Ostats does some basic statistics on a dataset of quantitative traits It plots a histogram and calculates the sample size mean variance standard deviation skewness kurtosis and average deviation for a quantitative trait The program also summarizes missing marker and trait data as will as determining the marker types dominant or codominant Finally Qstats will test whether markers are segregating at random It requires a molecular ma
96. ed by other programs Transferring Files You can start a Command Prompt and from there ftp files to your home account You will need the IP number or hostname and domain name to do this Simply start up the Command Prompt type in the drive from which you want to transfer files and cd to the directory where the files are Then ftp to your home machine and put the files there Use quit to kill ftp and exit to get back to Windows Here is an example c gt k k gt cd modules k module5 gt ftp mymachine somedomain net ftp gt prompt ftp gt mput ftp gt quit k module5 gt exit 54 Basic Unix This is meant to be a quick summary of some basic Unix commands One thing to keep in mind is that Unix is case sensitive Feel free to practice any of the following commands but be careful with rm and mv 5 4 1 Help The man command is one of the most important for the novice and experienced user If you would like to know what it does type man man at the prompt in a command window You can use it to get information on most of the commands below 5 4 2 Basic filesystem commands Here is a list of basic commands for seeing copying and moving the files in your direc tory creating new subdirectories and navigating Go ahead and experiment with these commands e ls is a command to list the files in the present working directory You can give it options for example Is 1 will give listings with more information ab
97. en in the other options 107 June 22 2000 CHAPTER 8 UNIX MAN PAGES SEE ALSO Rmap 1 Rcross 1 Ostats 1 LRmapqtl 1 SRmapqtl 1 Zmapqtl 1 JZmapqtl 1 Eqtl 1 Prune 1 Preplot 1 OTLcart 1 AUTHORS In general it is best to contact us via email basten statgen ncsu edu Christopher J Basten B S Weir and Z B Zeng Department of Statistics North Carolina State University Raleigh NC 27695 8203 USA Phone 919 515 1934 108 CHAPTER 8 UNIX MAN PAGES QTL Cartographer 8 4 RCROSS NAME Rcross Simulate or reformat a data set SYNOPSIS Rcross o output iinput m mapfile q modelfile r repetitions c Cross n SampleSize H heredity E Ve I Interactive DESCRIPTION Rcross performs a random cross or reformats a data set Cross types include F1 backcrosses to the P1 or P2 F2 crosses produced by selfing or random mating recombinant inbred lines as well as a few others It simulates marker and trait data The markers simulated come from a molecular map that could be a random one produced by Rmap or a real one in the same format as the output of Rmap The QTL model could be a random set produced by Rqtl or an estimated set in the same format as the output of Rqtl Rcross can also translate files from three different formats If the user chooses to translate a file then the simulation options are ignored OPTIONS See OTLcart 1 for more information on the
98. en indicates that when reading trait data the given token indicates missing phenotypic data 89 June 22 2000 CHAPTER 6 INPUT FILE FORMATS Translation Table This is an example of a translation table for marker information TranslationTable AA 2 2 Aa I 1 aa 0 0 A 12 12 a 10 10 ST Note a few things in the above translation table There are six rows and three columns There must be a token in all 18 positions of the table The first column is the genotype The program assumes that the A allele is diagnostic for the High Parental 1 line and the a allele is diagnostic for the Low parental 2 line These were previously denoted by A and Ag They aren t here because the above text comes from an ascii file A minus sign means the allele is unknown Thus dominant markers can be encoded The middle column is how the output of these genotypes will be encoded while the right 3rd column is how you will code the input of this file The above TranslationTable maps 2 to 2 1 to 1 0 to 0 etc Just about any set of tokens can be used for the third column but DO NOT change the first two columns If you encoded your P1 homozygotes as BB heterozygotes as Bb etc your translation table might appear as TranslationTable AA 2 BB Aa 1 Bb aa 0 bb A 12 B a 10 p 1 Anything in the following data file that is not recognized doesn t match something in column 3 will become unkno
99. er will also be used as a unique identifer on the first line of the output file 96 CHAPTER 8 UNIX MAN PAGES QTL Cartographer This can be a useful option It is recorded in the log file when any program is run It is possible to recreate exactly what was done using the log file e This requires a filename for the log file It will be appended to if it exists and created if not The default is gtlcart log X Give a filename stem All output will start with this stem and have extensions indicat ing what is in them EXAMPLES For all the following examples assume that QTLCart is just a wildcard for any of the programs in the suite QTLcart R resource file OTLcart will read option values from the file resource file The other programs do this and except for Preplot will regenerate the file upon exit QTLcart X corn Will set the filename stem to corn The output files will then have names beginning with corn and logical extensions For example the map file will be placed in corn map and the file containing the data from a cross will be in corn cro Filenaming conventions follow the old DOS 8 3 due to historical reasons GLOBAL COMMAND LINE OPTIONS All the parameters for OTLcart are also parameters for the other programs in the QTL Cartographer system GLOBAL BEHAVIOR All the programs in the QTL Cartographer suite behave in the same general way They were originally UNIX programs and can be run as such using comm
100. es with MAPMA KER EXP Each number is a command in a sequence to be done in MAPMAKER EXP Anything in side of square braces are comments and should not be typed into MAPMAKER EXP Start up MAPMAKER EXP in an appropriate subdirectory and proceed with these commands 1 prepare data sample raw Input the data from the raw file 2 photo sample tutorial Save what you do in a log file 3 sequence 1234567891011 12 Start with all markers 4 group Group them into linkage groups 5 sequence 12357 Use randomly ordered group 1 makers 6 compare Compare all orders For each in turn calculate the Likelihood 7 sequence 13257 Decide that this is the best order and specify it 8 map Print the map to the screen This attaches distances as well 9 sequence 4 6 89 10 11 12 Now use the rest of the markers 10 list loci Summarize the number of informative progeny 11 lod table Show pairwise distances and linkage LOD scores 81 June 22 2000 CHAPTER 5 TUTORIAL EXAMPLES 12 sequence 89101112 Use a randomly ordered subset of markers from group 2 13 compare Compare all orders For each in turn calculate the Likelihood 14 sequence orderl Use the best order from the compare command 15 try 46 Try all possible positions of markers 4 and 6 Also try unlinked idea 16 sequence 4118129610 This is the best sequence 17 make chromosome cl Create chromosome 1 18 sequence 13257 Specify the sequence of markers on
101. ew options that can only be set or changed in this interactive menu One of these is the aforementioned filename stem which will be explained in greater detail in the next section 27 June 22 2000 CHAPTER 1 INTRODUCTION One can also access the online help in the menu There will be an option to choose help and if the program cannot find the help file it will ask the user for the full path and filename of the help file The help file is an ASCII text file with tags indicating topics and subtopics There are summaries of all the programs and their options in the help file A feature that is not apparent from the interactive menu is that of rewriting the resource file without doing any calculations There is a quit command which is the penultimate numbered command If you choose the quit command you will exit without rewriting the resource file It is possible to change parameters in the menu and save them without run ning the program Simply select the last value The program will overwrite the resource file and exit without doing anything else This is a feature for all the programs of the suite 1 6 2 Filenaming Conventions The QTL Cartographer system reads and creates many files and each has a default name For example the default output file for Rmap is qtlcart map We find it convenient to specify a filename stem and allow for the filename extension to indicate which program created it and what it contains Suppose we were working
102. experiments 6 2 1 Rqtl input files The input format is similar to that for Rmap The input is token based The first line should start with a pound symbol and have a long integer after it The number will be an identifier for the file and should be unique Finally a filetype should be speciefied on the first line 12345789 filetype atls inp These commands are recognized e Units Indicates the units of the distances Valid tokens following this command are cM M or r for centiMorgans Morgans or recombination probabilities 87 June 22 2000 CHAPTER 6 INPUT FILE FORMATS e named indicates whether traits will have names Valid tokens following this com mand are yes and no e skip Begin skipping tokens until an unskip token is encountered e unskip see above e start start data segment e stop stop reading data e end quit close file After the start token there should be the token qtls and a number to indicate the number of traits to be modeled After this there should be a repeating sequence of a trait name number of loci for that trait then the chromosome position additive and dominance ef fects for each locus This example has the loci followed by their positions in centiMorgans from the telomere Please give all traits unique names start qtls 3 Trait_1 4 1 Oecd 075 0 10 891 OND 0 0 3 68 4 0 22 0 0 4 43 2 0 95 0 0 Trait_2 2 2 93 4 0 42 0 0 4 33 2 0 90 0 0 Trait_3 1 1 33 4 0
103. for Rmap The raw file is the input for the Rcross utility 2 13 QTL Cartographer user input format The third format is one defined for the QTL Cartographer system It is similar to the MAP MAKER output format but has commands embedded in the file to allow the program to read in the data more easily There is an example and further explanation of this format in Section 6 1 2 It can be annotated quite freely the example file map inp is self document ing 34 CHAPTER 2 SIMULATING REFORMATTING DATA QTL Cartographer 2 14 Command Line Options Table 2 1 summarizes the command line options for Rmap Most of these were explained in 2 1 1 The default options in Table 2 1 would produce a genetic linkage map on four chromosomes with 16 markers each The markers would be equally spaced at 10 centi morgan intervals and would span the genome Option Default Explanation i Input File 0 qtlcart map Output File f 1 Map Function p 0 0 Map function parameter g 1 Output Flag c 4 Chromosomes m 16 Markers per Chromosome vm 0 0 Standard deviation of Markers per Chromosome d 10 0 Intermarker Distance cM vd 0 0 Standard deviation of Intermarker Distance t 0 0 Tails Flanking DNA in cM M 0 Simulation Mode 0 1 Table 2 1 Command Line Options for Rmap Map Function A map function is a mathematical relationship between recombination probabilities and map distances measured in centimorgans or Morga
104. global options h for help A for automatic V for non Verbose W path for a working directory R file to specify a resource file e to specify the log file s to specify a seed for the random number generator and X stem to specify a filename stem The options below are specific to this program If you use this program without specifying any options then you will get into a menu that allows you to set them interactively o This requires a filename for output Reross will overwrite the file if it exists and create a new file if it does not If not used then Reross will use gtlcart cro This output is in a format suitable for any of the mapping programs i This requires an input filename This file must exist Reross will attempt to identify the format of the file and translate it to another format Specifying a file with this option turns off the simulation parameters below m Rcross requires a genetic linkage map This option require the name of a file containing the map It should be in the same format that Rmap outputs The default file is gtlcart map q Rcross needs a genetic model to simulate a data set It will read from the file specified by this option The file specified should contain a genetic model in the same format as the output of Rqtl The default file is qtlcart qtl 109 June 22 2000 CHAPTER 8 UNIX MAN PAGES H Allows the user to specify the heritability for the trait If used it requires a value in the ra
105. h World Congress on Genetics Applied to Livestock Production Guelph Ontario Canada e Basten C J B S Weir and Z B Zeng 2000 QTL Cartographer Version 1 14 Depart ment of Statistics North Carolina State University Raleigh NC 1 3 3 Gnuplot Copyright Information We suggest that you download and make use of the fine plotting package GNUPLOT Williams and Kelley 1993 which we use as the graphics engine to display the results of analyses GNUPLOT is freely available for UNIX Macintosh and MS Windows machines It is quite easy to use produces nice results and all the input files are plain text We reprint the copyright information for GNUPLOT verbatim GNUPLOT copyright information Copyright C 1986 1993 Thomas Williams Colin Kelley Permission to use copy and distribute this software and its documentation for any purpose with or without fee is hereby granted provided that the above copyright notice appear in all copies and that both that copyright notice and this permission notice appear in supporting documentation Permission to modify the software is granted but not the right to distribute the modified code Modifications are to be distributed as patches to released version This software is provided as is without express or implied warranty AUTHORS Original Software Thomas Williams Colin Kelley Gnuplot 2 0 additions Russell Lang Dave Kotz John Campbell Gnuplot 3 0 additions 17
106. hat are evenly spaced and span the genome Average intermarker distances of 5 to 10 centimorgans would be optimal We have provided ways to simulate linkage maps as well as to convert linkage map informa tion into a format suitable for QTL Cartographer Presently the user has two options for genetic linkage map input The first is a format designed for the QTL Cartographer system that allows for free annotation of the data file An example is given in 6 1 2 A second option allows the user to import the results of a MAPMAKER session This is covered in more detail in 2 1 and 5 8 1 12 Programming Philosophy These programs were originally developed on a UNIX workstation Consequently the programming philosophy is heavily influenced by the UNIX operating system All the programs have command line options which mimic those of regular UNIX commands We have added interactive menus so as to make the programs more user friendly on Mac intoshes and PCs running Microsoft Windows Y There are a number of different programs in the package rather than one program that does everything In this way each program does a small job and the user can combine the programs as a group to do a complete analysis The user can examine the input and output files for each step and have a better idea of what the programs are doing All input and output files are plain ASCII text They can be transferred to any platform and viewed or edited there We have also been
107. he output terminal Valid options can be found in the GNU PLOT manual The default is x11 on UNIX mac for Macintosh and windows for MS Windows 129 June 22 2000 CHAPTER 8 UNIX MAN PAGES S When given an argument Preplot will use this significance threshold It is 3 84 by default H Preplot will get results for this hypothesis test from the Zmapqtl outputfile Test 1 is the default which is the only value for a backcross L If given an argument of 1 Preplot will output LOD scores instead of the LR test statis tics EXAMPLES Preplot L 1 Preplot will automagically reformat your results to be plotted by GNUPLOT converting the likelihood ratio test statistics into LOD scores along the way REFERENCES 1 T Williams and C Kelley 1993 GNUPLOT An Interactive Plotting Program Ver sion 3 5 BUGS Preplot ignores JZmapqtl output SEE ALSO Rmap 1 Rqtl 1 Reross 1 Ostats 1 LRmapqtl 1 SRmapqtl 1 JZmapqtl 1 Eqtl 1 Prune 1 Zmapqtl 1 OTLcart 1 AUTHORS In general it is best to contact us via email basten statgen ncsu edu Christopher J Basten B S Weir and Z B Zeng Department of Statistics North Carolina State University Raleigh NC 27695 8203 USA Phone 919 515 1934 130 CHAPTER 8 UNIX MAN PAGES QTL Cartographer 8 12 EOTL NAME Eqtl Summarize the output of Zmapqtl SYNOPSIS Eqtl o output z zmapfile m mapfile t trait M Model a size S t
108. he sample could be a randomly generated one from Reross or a real one in the same format as the output of Rcross In addition the program requires the results of the stepwise linear regression analysis of SRmapqtl for composite interval mapping OPTIONS See OTLcart 1 for more information on the global options h for help A for automatic V for non Verbose W path for a working directory R file to specify a resource file e to specify the log file s to specify a seed for the random number generator and X stem to specify a filename stem The options below are specific to this program If you use this program without specifying any options then you will get into a menu that allows you to set them interactively o This requires a filename for output Zmapqtl will append the file if it exists and create a new file if it does not If not used then Zmapqtl will use gtlcart z i This requires an input filename This file must exist It should be in the same format as the output of Rcross The default file is gtlcart cro m Zmapqtl requires a genetic linkage map This option requires the name of a file con taining the map It should be in the same format that Rmap outputs The default file is gtlcart map t Use this to specify which trait Zmapqtl will analyze If this number is greater than the number of traits then all traits will be analyzed The default is to analyze trait 1 only 1 Allows the user to specify the name of the f
109. hey are doing You can also use finger to get info on a user on another machine Try finger basten statgen ncsu edu 5 4 4 Other commands e rlogin telnet and ftp allow you to initiate sessions on other machines You need to supply the IP address or nickname of the machine with these commands e exit closes a terminal window and clear clears it e history shows the last 40 commands issued They will be numbered and you can rerun them with an exclamation point and the number of the command e g 23 would run the command numbered 23 in the history list e lpr sends a file to the printer You can print up to 50 sheets from your account e alias allows you to assign Unix commands to more familar words For example alias dir ls would allow you to type dir to list the files in a directory alias with no arguments would list the current aliased commands 78 CHAPTER 5 TUTORIAL EXAMPLES QTL Cartographer 5 5 Simulating and Analyzing data Assuming that you have a qwork subdirectory folder create a new subdirectory folder within it You can call it anything you like but for the purposes of illustration it will be referred to as example1 Thus your working directory for this example will be qwork examplel if on a Macintosh or qwork example1 if on a PC If you are on a UNIX machine cd into the qwork example1 subdirectory and don t worry about set ting a working subdirectory Also note that if you
110. hreshold L lod DESCRIPTION Eqtl reformats the prodigous output of Zmapqtl The output file has a section that is suitable for input to Rcross There are other sections to the output that are more readable Eqtl can also detect whether a bootstrap permutation or jackknife analysis was performed and process the interim files produced by those analyses OPTIONS See QTLcart 1 for more information on the global options h for help A for automatic V for non Verbose W path for a working directory R file to specify a resource file e to specify the log file s to specify a seed for the random number generator and X stem to specify a filename stem The options below are specific to this program If you use this program without specifying any options then you will get into a menu that allows you to set them interactively o This requires a filename for output Eqtl will overwrite the file if it exists and create a new file if it does not If not used then Eqtl will use gtlcart eqt z This requires an input filename This file must exist It should be in the same format as the output of Zmapqtl The default file is gtlcart z m Eqtl requires a genetic linkage map This option requires the name of a file containing the map It should be in the same format that Rmap outputs The default file is gtlcart map H Allows the user to specify which hypothesis test results to process Use values 10 or 14 for data with two marker clas
111. if a permutation test was done and gtlcart z3a if a bootstrap was done Eqtl automatically detects these files and processes their results It will open a qtlcart z3e file and determine an experimentwise threshold based on the size specified with the a option If the gtlcart z3a file exists then Eqtl opens it and computes the means and standard devi ations at each test site of the likelihood ratio test statistic additive effect and dominance effect The results are printed to gtlcart z3b The jackknife procedure produces a gtlcart z3i that Eqtl opens computes the means and standard deviations at each test site of the likelihood ratio test statistic additive effect and dominance effect The results are printed to gtlcart z3 REFERENCES BUGS If the resource file indicates that there are more than one trait then Eqtl will try to estimate positions and additive effects for all the traits This will even if no analysis was done on the extra traits The output file will then have some null estimates When doing a jackknife with Zmapqtl the user should check that the file ending in the letter i is truely the last version of the interim jackknife file Zmapqtl switches between a file ending in i and another ending in j so check both and move the j file onto the i file if required If you set the significance threshold too high then Eqtl may find no QTL in the gtlcart z output If this is the case then Eqtl will crash 132 CHAPTER 8 UN
112. ife summary file Zijacks out Table 1 6 Miscellaneous Files and File types 29 June 22 2000 CHAPTER 1 INTRODUCTION 30 Chapter 2 Simulating Reformatting Data The first phase in using QTL Cartographer is to create some data You have two options for this You can either simulate a data set or collect one yourself The end result will be to have two files One will contain the information on a genetic linkage map marker order chromosome assignment and recombination fractions and the other a data set from a cross which contains the markers trait values and other explanatory variables QTL Cartographer cannot create a genetic linkage map from a data set You will have to use another program such as MAPMAKER EXP for that task Figures 2 1 2 2 present a schematic of the data simulation reformatting process There are four main programs involved in this phase Rmap Rqtl Rcross and Prune Rmap is a program designed to create random genetic linkage maps or reformat linkage maps that were prepared by MAPMAKER EXP Rqtl is a program that creates a genetic model for simulation One can specify the positions effects and the number of loci for each trait or have the program do it randomly Finally Rcross uses the genetic linkage map and the model to create a random data set by simulating a cross Reross can also reformat MAPMAKER QTL raw data files or specially formatted data files The fourth member of this group is Prune With Prune
113. ifies this number which must be in the range 0 0 100 0 If option 6 is selected Prune will eliminate individuals with this percentage of missing marker data Resampling data Options 7 8 and 9 allow the user to bootstrap permute or simulate missing data for the dataset If a bootstrap is chosen then a new dataset of the same size will be resampled with replacement from the original data A permutation simply permutes the trait values The simulation of missing data requires a percentage level to simulate That percentage of markers will then be set to unknown These options are examined in more detail in Section 2 4 2 Selecting 7 8 or 9 will do the requested action write the output and exit The other options require you to specify when to write and exit You also have the option of exiting without writing anything 2 4 2 Recreating Datasets Bootstrapping The b option with a value of 1 tells Prune to create a single bootstrapped data set This option should be used alone It will sample the data set with replacement creating a new data set of the same sample size and writing it to the file gtlcart crb Of course you can change the output file name by changing the output filename stem with the o option Using Prune one can perform a bootstrap experiment on the data set This is much easier to do on a UNIX workstation than a Macintosh or MS Windows machine because it can be automated in a batch file For example if one wanted to
114. ile containing results from LRmapqtl Zmapqtl reads those results and uses the information to choose cofactors for some of the anal ysis methods 121 June 22 2000 CHAPTER 8 UNIX MAN PAGES S Allows the user to specify the name of the file containing results from SRmapqtl Zmapgqtl reads the results and uses the information to choose cofactors for composite interval mapping model 6 M Zmapqtl assumes the specified model see below in the analysis Model 3 is default c The user can specify a specific chromosome for Zmapqtl to analyze If zero then all will be analyzed d Zmapqtl walks along the chromosome at a rate that can be specified with this option The default is to do an analysis every 2 centiMorgans along the chromosome n Use this to indicate how many background parameters Zmapqtl uses in composite interval mapping This is used only with model 6 and gives an upper bound If fewer than this number of markers are ranked in the SRmapgtl out file then less than the specified number of markers will be used w Zmapqtl blocks out a region of this many centiMorgans on either side of the markers flanking the test position when picking background markers It is 10 by default and is only used in models 5 and 6 We refer to it as the window size r Zmapqtl can do a permutation test to determine the threshold for rejecting the null hy pothesis of no QTL at a site By default this option sets the number of permutations equa
115. ile will be stem hp You could use the hpljii option for Terminal and then edit the stem plt file to change the type of printer to anything that GNUPLOT supports See the GNUPLOT manual Williams and Kelley 1993 for more details Extension Meaning s Significance Threshold lr Linear Regression results Zit Composite interval mapping results q Quantitative trait locus data from Rqtl Table 4 3 Filename extensions for Preplot output 43 GNUPLOT GNUPLOT is free plotting software available for UNIX Macintosh and Windows ma chines It is an interactive package The basic idea behind the program is to read in simple files of numbers and plot them The files of numbers contain two columns one for the abscissa and one for the ordinate Preplot takes care of reformatting the output of the analysis so that GNUPLOT can read the results and plot them We have placed copies of GNUPLOT for the three platforms on our ftp server 73 June 22 2000 CHAPTER 4 VISUALIZATION OF RESULTS 4 3 1 Basic GNUPLOT In many ways GNUPLOT is similar to MAPMAKER in that it is an interactive command driven program Once GNUPLOT has been started the user can type help to get in formation on how to use the program There are commands to change the terminal type load files and specify the output device Thus one can view or print the images created by GNUPLOT If you have run GNUPLOT you should have a plot control file
116. influenced by the Free Software Foundation in that we charge no fee for this program package We have attempted to integrate these programs with other free software most notably GNUPLOT and MAPMAKER 13 Copyright Information and Acknowledgments 1 3 1 QTL Cartographer Copyright Information Copyright C 1994 2000 C J Basten B S Weir and Z B Zeng Permission to use copy and distribute this software and its documentation for any pur pose with or without fee is hereby granted provided that the above copyright notice ap pear in all copies and that both that copyright notice and this permission notice appear in supporting documentation Permission to modify the software is granted but not the right to distribute the modified code Modifications are to be distributed as patches to released version This software is provided as is without express or implied warranty 16 CHAPTER 1 INTRODUCTION QTL Cartographer 1 3 2 Citing OTL Cartographer In publications you should cite our original short announcement Basten Weir and Zeng 1994 and this manual e C J Basten B S Weir and Z B Zeng 1994 Zmap a QTL cartographer IN Proceed ings of the 5th World Congress on Genetics Applied to Livestock Production Computing Strategies and Software edited by C Smith J S Gavora B Benkel J Chesnais W Fairfull J P Gibson B W Kennedy and E B Burnside Volume 22 pages 65 66 Published by the Organizing Committee 5t
117. into any programs The default behav ior of Preplot is what we term the automagic mode Preplot reads the Zmapqtl output file determines what analyses have been done and then reformats all of these analyses in a logical way There will be a separate graph for each trait and each chromosome Pre plot will attempt to put the results from different models in Zmapqtl and from LRmapqtl on the graphs along with any information from the Rqtl output file if it exists and a significance threshold which can be set in the interactive menu or on the command line Table 4 2 shows the command line options specific to Preplot In general it will not be necessary to change any options to Preplot Most of the proper values should have been set by other programs in the QTL Cartographer suite You might want to use the L command to tell Preplot to convert LR values into LOD scores In any case the output of Preplot is ready for import into GNUPLOT There will be a number of output files One is a plot control file that has commands that GNUPLOT understands The other files simply contain two columns of numbers for the x and y coordinates to plot The names of the files indicate what the numbers are for They all start with a lower case c which indicates chromosome Following the c is an integer indicating which chromosome then there is a t followed by an integer indicating the trait Then there is a period and a file extension that indicates the results c
118. its 2 Trait_1 Trait_2 named Ind_1 5 0 End 2 543 Ind_3 6 2 Ind_4 4 1 Indio 545 Ind_6 5 8 Ind_7 6 7 Ind_8 6 1 Ind_9 3 Ind_10 6 4 155 ES 16 24 253 29 16 26 33 16 0 NOFA OOF ND W 4 stop individuals traits start ind Ind_1 M HHHH a agoa A Ay AES S H Q H o H H Q Ke DONN E EH HI aa o o ST DD Q O 0 DOB WN H Q H H 5 z o mn G ividuals otraits 2 sex brood named 2 stop individuals otraits 92 Chapter 7 Benchmarks Tables 7 1 7 2 summarize the timings for Zmapqtl to do interval mapping Lander and Botstein 1989 and composite interval mapping Zeng 1993 Zeng 1994 on various com puting platforms under different operating systems All timings were done in the winter of 1999 The simulated data set has been used previously Zeng 1994 and consists of a genetic linkage map that has four chromosomes with 16 markers on each chromosome The markers are evenly spaced at 10 cM and the simulated data has one trait The entire genome was scanned at a walking speed of 2 cM The programs were run in automatic mode with no recourse to the interactive menus They indicate the amount of time to read in the data perform the analysis and write the output Table 7 1 summarizes the timings for interval mapping Machine Speed Mhz Time seconds Ratio to UltraSparc 60 PowerMac G3 266 7 1 7 PowerMac G3 400 5 S
119. kdir 24 CHAPTER 1 INTRODUCTION QTL Cartographer Option Default Explanation e qtlcart log Error and Log File S 795793333 Random Number Seed h off Show help and exit R qtlcart rc Resource File W none Working Directory A off Automatic mode X qtlcart Filename Stem V on Verbosity Table 1 4 Command Line Options for all programs would indicate go up one level from the binary subdirectory where you will find a workdir subdirectory In UNIX it might look like W workdir For the Macintosh you use extra colons If the binaries are in the bin ppc folder inside the qtlcart folder then W workdir would indicate that there is a folder called workdir in the qtlcart whereas W workdir would indicate that the workdir folder is inside the bin ppc folder Listing options Using the h option will print out a list of all command line options and their values The program will then exit without doing anything I find this most useful when I just want a reminder of what the programs expect This may not seem as useful now that there is an interactive menu to set options but if you only want to use the programs in batch mode it is a quick way to see what the values of all parameters are Random Number Seed Many of the simulation programs make use of a pseudo random number generator that requires a seed If none is provided the number of seconds since some date in the past is used The s op
120. ke io ON te a a Mk a ed ek en An 131 Bibliography 134 Index 136 June 22 2000 CONTENTS List of Figures 1 1 2 1 2 2 3 1 4 1 Basic Grosse in et yc hb oe ae eds Et a See nus Net 15 Reformattineg Data Lati eat adage E a AAA e 32 Simulating Dar A A a a ete ee ie hae Ne 32 Analysis SCHEINAUC i s sa e as Grek ee We bok Pee e De fit 50 Visualization Schematic 70 June 22 2000 LIST OF FIGURES 10 List of Tables 1 1 1 2 1 3 1 4 1 5 1 6 2 1 2 2 2 3 2 4 2 5 3 1 3 2 3 3 3 4 3 5 3 6 4 1 4 2 4 3 7 1 7 2 Summary of Experimental Design Codes 15 Subroutines from Numerical Recipes ME ue hs Se dia Rte 18 Contactior Help es SS o ae eae en tds QE 23 Command Line Options for all programs 25 Standard Filename Extensions and File types for Output Files 29 Miscellaneous Files and File types 29 Command Line Options for Rimap 2 4 6 roue Gok Rues aS Pees 35 Command Line Options for Rap 1 6 84 ee asada 36 Command Line Options for Ratl eee Desire Mee es ee Sed 37 Command Line Options for Rcross 38 Command Line Options for PURE s42 5 4 Moses sans eae 43 Command Line Options for Qstats 52 Command Line Options for LRmapqtl 54 Command Line Options for SRmapqtl 55 Command Line Options for Zmapqtl ici ie Ro
121. l to 0 which means no permutation test is run You can set it to a number lt 10000 to do the test See Churchill and Doerge 1994 for more details The results are in an interim file Use Eqtl to summarize them when enough repetitions have been done You need to run Zmapqtl without permutations or bootstraps at least once before you can do the permutation tests This option only allows for interval mapping Model 3 or composite interval mapping Model 6 b When used with argument 1 Zmapqtl will do a single bootstrap You need to run Prune to actually create the bootstrapped data set This option merely analyzes it and stores summary statistics in an interim file gtlcart z3b by default for model 3 You should also run Zmapqtl without bootstraps or permutation tests before doing a bootstrap analysis When used with an argument 2 Zmapqtl will do a jackknife analysis Again Zmapqtl should be run without this argument prior to doing a jackknife INPUT FORMAT The input format of the molecular map should be the same as that of the output format from the program Rmap The input format of the individual data should be the same as the output format of the program Rcross 122 CHAPTER 8 UNIX MAN PAGES QTL Cartographer EXAMPLES Zmapqtl Calculates the likelihood ratio test statistics of the dataset in gtlcart cro using the map in gtlcart map nice Zmapqtl A V i corn cro m corn map M 6 r 500 Calculates the likelihood r
122. llowing options can be used with any of the programs in the QTL Cartographer suite The current programs are Rmap Rqtl Rcross Qstats LRmapqtl SRmapqtl JZmapqtl Eqtl Prune and Preplot h Prints out the current values of all program options and information on what the pro gram does It then exits V Turns the verbosity mode off The programs in the suite print out messages while running This option turns off those messages This is useful for batch files A Skips the interactive screen for setting options All programs start up with a menu that allows setting of options This turns the menu off It is also very useful for batch files R The programs will read the default parameters from a file specified with this option If a file called gtlcart rc is in the current working directory it will be opened by default and all parameter values read If no such file exists then default parameter values will be assumed and the file will be created It is probably better to simply rename a resource file gtlcart rc than to use this option W This option allows one to set the work directory This directory must exist All the input files must be in this directory and the output files will be placed there s This requires a long integer to act as the random number seed By default it is the value returned by the ANSI C function time which is usually the number of seconds since some arbitrary past date often 1 January 1970 This numb
123. mute csh Usage Permute csh stem permutations email where stem is the filename stem permutations is the number of permutations and email is the user s email address Note This only works if you have set and used a filename stem if 1 h then echo Usage Permute csh stem model permutations email echo Where echo stem filename stem echo model Zmapqtl Model echo permutations number of permutations echo email user s email address echo echo Now exiting exit endif set tlog temp log usr bin rm f tlog echo Permutation test started gt tlog usr bin date gt gt Stlog echo Stem 1 gt gt tlog echo Model 2 gt gt tlog echo Reps 3 gt gt tlog echo Email 4 gt gt tlog set bindir usr local bin set i l mv 1 log 1 logsave while i lt 3 Sbindir Prune A V i l cro b 2 gt gt amp Stlog nice Sbindir Zmapqtl A V M 2 i l crb r 1 n 5 gt gt amp Stlog i end 46 CHAPTER 2 SIMULATING REFORMATTING DATA QTL Cartographer mv l logsave 1 log usr bin date gt gt Stlog usr ucb mail 4 lt tlog Upon completion you can run Eqtl to get the experimentwise significance thresholds You could also have SRmapqtl redo the stepwise regression in the above script so that the background markers in composite interval mapping reflect the permuted data set rather than the original There i
124. ncements of updates and bug fixes To subscribe send the following two line message to the server subscribe qtlcart end The second line in the message stops MajorDomo from interpreting your sig Note that the subject line of your mail message will be ignored If the subscription was successful you will receive a confirmation note saying as much You may also put an email address after the subscribe qtlcart on the same line to subscribe that address subscribe qtlcart basten statgen ncsu edu end A message like the above with unsubscribe rather than subscribe would unsubscribe the address The command help would cause the server to return a list of commands that can be sent to the MajorDomo server Remember that all commands should be di rected to MajorDomo statgen ncsu edu while messages for people on the list go to gtl cart statgen ncsu edu 1 5 2 Bug Reports Send any bug reports to gtlcart bug statgen ncsu edu There is certain information that will greatly aid in diagnosing the problem The QTL Cartographer distribution should come with a file called problems txt with the following questions in it 1 Computing platform a What machine are you using Is it a i UNIX based workstation ii PC running Windows iii PowerPC based Macintosh b What operating system is it running c What is the version of the Operating system d How much memory and free hard disk space do you have 2 Progr
125. nd line options specific to Zmapqtl One can select a trait to analyze a model for analysis and a walking speed along the genome that is the interval between successive analysis points The user can analyze just one chromosome or the entire genome Finally permutation tests or bootstraps can be performed by setting the number of permutations or bootstraps to a number greater than 0 Explanatory variables such as Sex or Line are automatically included in the analysis if their names are preceded by a plus sign in the data file This is similar to LRmapqtl except that interaction terms are not yet used Option Default Explanation i qtlcart cro Input File 0 qtlcart z Output File m qtlcartmap Genetic Linkage Map File l qtlcart lr LRmapqtl Results file S qtlcart sr SRmapqtl Results file Model 6 M 3 Model t 1 Trait to analyze c 0 Chromosome to analyze d 2 0 Walking speed in cM n 5 Number of Background Parameters Model 6 W 10 0 Window Size in cM Models 5 and 6 r 0 Number of Permutations b 0 Number of Bootstraps Table 3 4 Command Line Options for Zmapqtl Traits and Chromosomes The t option allows the user to specify which trait in a data set with multiple traits is to be analyzed For multiple trait analysis use JZmapqtl If you set the trait number to one 59 June 22 2000 CHAPTER3 ANALYSIS more than the total number of traits then all traits except for those whose names begin with a
126. nd line parameters that are valid for all the programs in QTL Cartographer Working directory A working subdirectory folder to hold all input and output files is a convenient way to organize your work We suggest using a different subdirectory folder for each data set In the UNIX world you can simply change into such a subdirectory and run the programs In the Macintosh and MS Windows environs you need to run the programs from where they reside and specify where the working directory is Use the W command line option to specify a working directory or set it in the interactive menu Be sure to follow the conventions of the particular operating system that you are working on For UNIX you might specify it as W home myaccount qtlcart workdir While for MS Windows it might look like W C qtlcart workdir And on a Macintosh assuming that your Hard drive is called MacintoshHD W MacintoshHD qtlcart workdir The programs will automatically add a file separator to the end of the path if you don t put it in Thus W MacintoshHD qtlcart workdir is equivalent to the first incarnation of the Macintosh work directory The Macintosh file separator is equivalent to the DOS and the UNIX 7 You may also use relative pathnames for the working subdirectory In the UNIX and Win dows environments a single period means from here and a pair of periods indicates one higher directory level Thus W wor
127. ndom variables sampled from the gamma distribution Zeng 1992 page 993 equation 12 and reprinted here 38 q8 le a8 f a 0 lt a lt 0 lt 8 lt 2 7 G 37 June 22 2000 CHAPTER 2 SIMULATING REFORMATTING DATA The shape parameter 5 allows a wide variety of different genetic models to be generated The additive effect of substituting an Q allele for an Q allele is a When multiple traits are simulated the number of QTLs per trait is simulated as a random variable with mean specified by the q option If an input file is specified then it is translated into a format readable by Reross and the options in Table 2 3 from Number of Traits and below are ignored The input file format qtls inp is defined in Section 6 2 1 This input file format will allow a wide variety of genetic models to be simulated 2 3 Rcross Rcross uses the information generated by Rmap and Rqtl and randomly simulates a data set Alternatively it can also reformat MAPMAKER raw data files and cross inp filetype formatted files Table 2 4 presents the options for Rcross The default values would create a simulated sample of 200 individuals backcrossed to P with a heritability of 0 5 for the quantitative trait Option Default Explanation i None Input File O qtlcart cro Output File m qtlcart map Genetic Linkage Map File q qtlcart qtl QTL Data File n 200 Sample Size C 1 Type of Cross H 0 5 Herita
128. ng The automatic mode should only be used by those familiar with the QTL Cartographer programs Resource File A resource file is an ASCII text file that keeps track of the parameters that the user specifies in using the programs The same file is read and updated by all the programs in the suite You can specify a resource file using the R option It is qtlcart rc by default and should be in the directory that you are currently working in for UNIX machines or where the binaries are for PCs and Macintoshes If you change any options either via the command line or the menus they will be saved to the file specified If you decide to use a file other than qtlcart rc as the resource file you will need to specify it for each program you run Initially the user may want to create a resource file with three lines in it The three lines will specify the working subdirectory online help file and a stem for filenames Here is an example of a resource file for the Macintosh version of the programs workdir test The working directory stem corn Stem for filenames helpfile qtlcart hlp The help file The working directory must be specified according to the rules of the operating system This was explained in using the W option in previous section In the above example a relative pathname was used The programs will assume that there is a directory folder called test in the directory folder one level up from the directory
129. ng data and calculate some basic statistics on your quan titative traits LRmapgqtl to do a simple linear regression of the data on the markers SRmapqftl to do a stepwise linear regression of the data on the markers to rank the markers This should be run with model 2 Zmapqtl to do interval or composite interval mapping This should be run twice once with model 3 and a second time with model 6 Preplot to reformat the output of the analysis for Gnuplot GNUPLOT to see the results graphically We recommend that the new user tries a simulation to gain an understanding of the pro grams REFERENCES 1 T Williams and C Kelley 1993 GNUPLOT An Interactive Plotting Program Ver sion 3 5 100 CHAPTER 8 UNIX MAN PAGES QTL Cartographer BUGS Many UNIX systems have been known to get upset when trying to run the QTL Car tographer programs from out of the front end It has something to do with the memory management Try running the individual programs one by one A good test is to simply run each program without changing any parameters SEE ALSO Rmap 1 Rqtl 1 Reross 1 Ostats 1 LRmapqtl 1 SRmapqtl 1 JZmapqtl 1 Eqtl 1 Prune 1 Preplot 1 AUTHORS In general it is best to contact us via email basten statgen ncsu edu Christopher J Basten B S Weir and Z B Zeng Department of Statistics North Carolina State University Raleigh NC 27695 8203 USA Phone 919 515 1934 101 Jun
130. nge 0 0 to 1 0 It is 0 5 by default E Allows the user to specify an environmental variance for the trait If used it requires a positive value and will disable the heritability This is ignored by default I is the flag to turn on interactive crosses By default it has a value of 0 To do interactive crosses use this option with the value 1 c Allows the user to specify the type of cross It requires a string such as B1 SF2 or RI1 See below for more on the values of the cross n This is the sample size of the offspring It is 200 by default and requires some integer value greater than 0 if used INPUT FORMAT The input format of the molecular map should be the same as that of the output format from the program Rmap The input form of the QTL data should be that of the output format from Rqtl If an input file for the data is used then it can have one of two formats The first is identical to the raw files required by MAPMAKER You must first use MAP MAKER to create a genetic map then run the map through Rmap to reformat it then use the map and the original raw file to reformat the data for subsequent use An alternative format is defined in a file cross inp that is included with the distribution The file can be annotated freely Look at the cross inp file and use it as a template for your data CROSSES A pair of inbred parental lines P1 and P2 that differ in the trait of interest and marker genotypes are crossed to produc
131. nning with the first program you run a resource file called qtlcart rc is created and updated for each subsequent program This file keeps track of all the parameters and file names that you use In addition a log file will record which specific parameters were used with which specific programs and when the programs were run Thus the qtlcart rc file keeps track of the current settings and the qtlcart log file records the history of parameter settings You can look at any of these files or any other files that QTL Cartographer creates by opening them in any text editor Macintoshes and PCs work a little differently than the examples below They maintain one copy of the qtlcart rc file in the subdirectory folder where the applications are lo cated You can specify a working subdirectory folder in any of the QTL Cartographer programs and this will be recorded in the qtlcart rc file The Introduction has more exten sive instructions on how to do this If you are on a Macintosh or a PC create a subdirectory folder called qwork in the subdirectory folder where the binaries are If you are on a UNIX machine create the qwork subdirectory in your root directory There is a web page for QTL Cartographer http statgen ncsu edu qtlcart cartographer html which is the good place to keep abreast of new information The readme file from the ftp server is linked to the web page The programs are also linked to the web page so you c
132. ns QTL Cartographer presently allows for eight map functions specified by an integer The numbers 1 2 or 3 correspond to the Haldane Kosambi and Morgan formerly Fixed mapping functions respectively The default is the Haldane mapping function If r corresponds to the recombination frequency between a pair of markers and dm is the distance between them in Morgans then the Haldane mapping function is defined by dm mG 2r 2 1 p zll exp 2dw 2 2 The Kosambi function is 1 exp 4dyz 21 exp 4dyy oo dm emf 2 4 35 June 22 2000 CHAPTER 2 SIMULATING REFORMATTING DATA and the Morgan function assumes dy r which is complete interference All eight map ping functions are discussed at length in Ben Lui s book Liu 1998 We direct the reader there for the details Table 2 2 lists the mapping functions and their integer codes for QTL Cartographer Some of these map functions require an extra parameter This parameter can be set in the Rmap menu See Section 10 3 1 of Liu 1998 for the details Code Reference Note 1 Haldane 1919 default 2 Kosambi 1944 3 Morgan 1994 Fixed 4 Carter and Falconer 1951 5 Rao et al 1979 0 lt p lt 1 6 Sturt 1976 L 7 Felsenstein 1979 00 lt K lt oo K 2 8 Karlin 1984 binomial N gt 0 Table 2 2 Command Line Options for Rmap Output Flags The output flag takes on values of 1 2 or 3 A 1 indicates
133. o you right away but will try to Name Dr Christopher J Basten Email basten statgen ncsu edu Phone 919 515 1934 Fax 919 515 7315 Address Program in Statistical Genetics Department of Statistics North Carolina State University Raleigh NC 27695 8203 USA MajorDomo MajorDomo statgen ncsu edu Bug Report qtlcart bug statgen ncsu edu Table 1 3 Contact for Help 1 6 General Usage of the Programs The programs in the QTL Cartographer suite all have the same look and feel and are heav ily influenced by UNIX programs They can be used as command line programs or in an 23 June 22 2000 CHAPTER 1 INTRODUCTION interactive mode where a menu of options is presented Some command line options that are common to all the programs are discussed in 1 6 1 The new user should become famil iar with these options In addition to the command line interface all the programs have an interactive menu for setting options The user need only start up any program in the suite and a list of options will appear Selecting the number of an option will allow the user to change the value of the option When all options are set to the user s satisfaction choos ing a zero 0 will cause the program to run Choosing the penultimate numbered option will allow you to exit the program without changing any files The last option saves any parameters you have set before exiting 1 6 1 Options for all programs Table 1 4 shows the comma
134. om Both statistics are calculated and presented in a table in the Qstats output 52 CHAPTER3 ANALYSIS QTL Cartographer 3 2 LRmapqtl LRmapqftl fits the data to a simple linear regression model For each marker in turn it fits a simple linear model to the trait data It is a quick way to get an idea of where the QTLs may reside 3 2 1 Simple Linear Regression For each marker in turn LRmapqtl fits the phenotypic data to the linear model Yi bo bizi e 3 2 where y is the phenotype of the ith individual and x is an indicator variable for the marker genotype Generally 2 if AA Ti 1 if Ao 0 if A242 but for B crosses fiat AA A 0 if AA If the marker is missing or dominant then an expected value for the marker is calculated from the flanking markers Fisch Ragot and Gay 1996 Jiang and Zeng 1997 The regres sion parameters b and b can be estimated and e is assumed to have a normal distribution LRmapqtl can also take into account categorical traits that is other variables such as sex or brood in its analysis If your data set contains such information then there should be a list of the names of these other variables near the beginning of the Rcross out formatted file These names might look as follows Names of the other traits 1 Sex 2 Line If you would like to include Sex and Sex by Marker interaction terms in your analysis then you need to indicate as much to LRmapqtl If
135. on a corn data set We might use corn as the filename stem Then Rmap would write its output to corn map and its error messages to corn log Rqtl would write its output to corn qtl etc Table 1 5 sum marizes the standard file name extensions in the QTL Cartographer system Beginning with version 1 12 the default behavior of QTL Cartographer is to use a filename stem If none is given then qtlcart will be the stem Unless specifically written in the qtlcart rc file the old default names of Rmap out Rqtl out etc will no longer be used These old default names will be used as filetype identifiers In the output files there will be a to ken filetype followed by a token from the fourth column of Tables 1 5 1 6 Note that Zmapqtl creates some interim files and that Preplot will create many other files in addi tion to the GNUPLOT control file See Section 4 2 for details The filetype specifier will greatly aid programs such as Rmap and Rcross in translating files As QTL Cartographer develops this feature will be used more extensively Once the stem is set in the menu it will be remembered as long as a resource file is present In the interactive menu if you pick an item to change say a filename you can wipe it out by inputting a solitary period This way if you had specified an input file in an earlier run you can delete it In addition to the files specified in the table we assume that files with extensions map
136. ontained in the file For the results of composite interval mapping the z filename extension will be followed by an integer from 1 to 7 indicating the model used for the analysis For example the file c2t3 z6 would have the results of composite interval mapping for trait 3 on chromosome 2 in it 4 2 1 Printing Results One option that is useful to change is the Terminal setting This will be set correctly if all you want to do is view the graphs on your screen with GNUPLOT If you want to 72 CHAPTER 4 VISUALIZATION OF RESULTS QTL Cartographer Option Default Explanation O qtlcart Gnuplot Control File Name m qtlcart map Genetic Linkage Map File q qtlcart qtl QTL or Estimated QTL file l qtlcart lr LRmapqtl Output File Z qtlcart z Zmapqtl Output File S 10 0 Significance Threshold T x11 Terminal L 0 Output LOD scores 0 no 1 yes i 1 Hypothesis for F2 design Table 4 2 Command Line Options for Preplot get a hardcopy printout you have two alternate choices for the Terminal option If you have a postscript printer then use postscript as the terminal Run Preplot and then run GNUPLOT as explained in Section 4 3 You will not see any output but a file qtlcart ps will be created or stem ps where stem is your filename stem This file can be sent to any postscript printer The other alternative is hpljii which does something similar for HP LaserJet Ils the output f
137. or the fit of the model to the data A third method FB is to start with forward stepwise regression but only keep adding markers while the p value of the partial F statistic of the marker to be added is below a defined threshold p F When a step is reached in which no more markers can be added all of the markers are retested to see if they are still significant Each marker in turn is deleted from the model a p value is calculated for the partial F statistic and if the p value is greater than a specified level p Fout it is deleted As with LRmapqtl any otraits that begin with a plus sign are also used in the regression model Unlike LRmapqtl no interaction terms are used The command line parameters for SRmapqtl are listed in Table 3 3 One added feature is that if you use the t option with an integer value one greater than the number of traits then all traits will be analyzed in turn 3 3 1 Output For the specified trait SRmapqtl will output a small table 55 June 22 2000 CHAPTER3 ANALYSIS Chromosome Marker Rank F Stat DOF start 1 1 2 13 38778 114 2 3 4 10 12742 110 3 5 3555928 108 3 2 3 11 15490 112 4 3 T 28 85236 116 end The first two columns indicate the chromosome and marker The third column gives the rank of that marker as determined by the stepwise regression mode of choice Then there will be an F statistic indicating the difference between having that variable in the model or not Finally
138. ory created for QTL Cartographer so let s assume that it is c qtlcart You may also want to download GNUPLOT for MS Windows In binary format get the self extracting archive gnuplot exe Put it in a subdirectory say C gnuplot and while in that subdirectory run gnuplot from the DOS command line The programs can be run by double clicking their icons in the Windows Explorer appli cation An alternate method is to open a Command Window and type in the program names You can view the output files in any text editor although you should be aware that some editors in MS Windows cannot load large files 14 2 UNIX Download the file QTLCart tar Z in binary form from statgen ncsu edu It is in the same directory that README file came from On your local machine create a subdirectory for the distribution then move the file QTLCart tar Z to it Uncompress and untar the file as follows AS uncompress OTLCart tar Z tar xf QTLCart tar AS Follow these steps to compile and install QTL Cartographer 1 Move into the src directory and copy the file LocalD h UNIX to LocalD h This file may be fine for your system It is annotated and you can follow the directions in the file if compilation doesn t work the first time 20 CHAPTER 1 INTRODUCTION QTL Cartographer 2 You will also need to edit the Makefile and choose a compiler The default is gcc which is the compiler used on our Sun workstations running Solaris If you don t hav
139. ose results and uses the information to choose cofactors for some of the anal ysis methods S Allows the user to specify the name of the file containing results from SRmapqtl JZmapqtl reads the results and uses the information to choose cofactors for com posite interval mapping model 6 M JZmapqtl assumes the specified model see below in the analysis Model 3 is default c The user can specify a specific chromosome for Zmapqtl to analyze If zero then all will be analyzed d Zmapqtl walks along the chromosome at this rate The default is to do an analysis every 2 centiMorgans along the chromosome n Use this to indicate how many background parameters JZmapqtl uses in composite interval mapping This is used only with model 6 and gives an upper bound If fewer than this number of markers are ranked in the SRmapgtl out file then less than the specified number of markers will be used w JZmapqtl blocks out a region of this many centiMorgans on either side of the markers flanking the test position when picking background markers It is 10 by default and is only used in models 5 and 6 We refer to it as the window size I JZmapqtl requires the user to specify which hypotheses to test For backcrosses there are two hypotheses numbered 1 and 0 Use 10 for backcrosses or a 14 to do GxE tests as well For crosses in which there are three genotypic classes there are hypotheses 0 1 2 and 3 Use 30 31 32 in that case or 34 to do GxE
140. out the files than ls 77 June 22 2000 CHAPTER 5 TUTORIAL EXAMPLES e pwd tells you where you are This can be useful if you have created many subdirec tories e cd allows you to change the current working directory You can give it an absolute or a relative argument cd would move you to the next highest subdirectory cd ncsu pams046 bin would move you to the the ncsu pams046 bin subdirectory etc e mkdir allows you to create a subdirectory mkdir test would create the subdirectory test rmdir would remove it You can only remove empty subdirectories e rm allows you to remove a file It is aliased as rm i which means that it will ask if you really want to remove the file rm filename would remove the file filename e mv moves a file mv file orig file new would move the file file orig to file new You can think of it as renaming e cp copies one or more files cp file1 file2 copies the file filel to the file file2 cp file1 file2 direct would copy the files file1 file2 to the directory direct e chmod is a rather complex command to change the permissions on files You can write batch files and use chmod to allow execution of them e more will display the contents of a file Use it as more filename While in more typing a q will get you out 5 4 3 Curious There are a couple of commands to find out who is on your machine and what they are doing w who and finger tell you who is logged on to your machine and what t
141. output of Eqtl is identical to the output format of Rqtl This is convenient if the experimenter would like to do simulation studies with a set of estimated QTLs The output of Eqtl can be used as the input to Reross with the appropriate genetic linkage map and new data sets can be simulated to examine the power of the different methods to detect the QTLs Finally the output of Eqtl can be read by Zmapqtl and used to create virtual markers to be used as covariates in composite interval mapping see model seven of Section 3 4 2 The remaining output is more readable and is appropriate if the experimenter is not inter ested in doing further simulations The positions of the QTL are given in Morgans from the telomere rather than recombination frequencies from the flanking markers In addition to reformatting the output of Zmapqtl Eqtl will automatically detect whether a permutation test jackknife or bootstrap experiment had been done If such results ex ist Eqtl will open and summarize them For example if you do a permutation test with 69 June 22 2000 CHAPTER 4 VISUALIZATION OF RESULTS qtlcart qtl rem c t files 1 Eqtl 2 Preplot 3 GNUPLOT Figure 4 1 Visualization Schematic Zmapqtl using interval mapping an interim file qtlcart z3e is created and appended to for each permutation Eqtl will read this file and calculate experimentwise threshold values from it Standard significance thresholds will be written to the log file
142. p that could be a random one produced by Rmap or a real one in the same format as the output of Rmap The sample could be a randomly generated one from Reross or a real one in the same format as the output of Rcross OPTIONS See OTLcart 1 for more information on the global options h for help A for automatic V for non Verbose W path for a working directory R file to specify a resource file e to specify the log file s to specify a seed for the random number generator and X stem to specify a filename stem The options below are specific to this program If you use this program without specifying any options then you will get into a menu that allows you to set them interactively o This requires a filename for output Qstats will append the file if it exists and create a new file if it does not If not used then Ostats will use gtlcart gst i This requires an input filename This file must exist It should be in the same format as the output of Rcross The default file is gtlcart cro m Ostats requires a genetic linkage map This option require the name of a file containing the map It should be in the same format that Rmap outputs The default file is gtlcart map INPUT FORMAT The input format of the molecular map should be the same as that of the output format from the program Rmap The input format of the individual data should be the same as the output format of the program Rcross 115 June 22 2000 CHAPTER 8 UNI
143. parc 10 36 34 8 5 UltraSparc 2170 167 8 2 0 UltraSparc 60 300 4 1 0 Pentium NT 4 0 260 8 2 0 Pentium NT 40 450 4 1 0 Table 7 1 Timings for Interval Mapping Table 7 2 summarizes timings for composite interval mapping The model for analysis was Model 6 with a window size set to 10 0 cM and using up to 5 markers to control for the genetic background Some of the ratios for the same machine change from interval mapping to composite interval mapping Model 6 uses quite a lot more double precision arithmetic and this may account for the differences For N replications of a permutation test or bootstrap the computing time should be less than N times the values in Tables 7 1 7 2 The jackknife analysis should be around n times these values where n is the sample size 93 June 22 2000 CHAPTER 7 BENCHMARKS Machine Speed Mhz Time seconds Ratio to UltraSparc 60 PowerMac G3 266 7 1 2 Sparc 10 36 45 7 5 UltraSparc 2170 167 10 1 7 UltraSparc 60 300 6 1 0 Pentium NT 4 0 260 10 1 7 Table 7 2 Timings for Composite Interval Mapping 94 Chapter 8 UNIX Man Pages In the UNIX world a standard way of providing online documentation of programs is to write man pages These are ASCII text files with embedded troff commands UNIX versions of QTL Cartographer have man pages for all the programs in the suite On a UNIX system if the man pages are in the correct subdirectory in essence if the subdirectory that cont
144. r analysis 1 Use all the markers to control for the genetic background This is model 1 from Zeng 1994 2 Use all unlinked markers to control for the genetic background This is model 2 from Zeng 1994 3 Don t use any markers to control for the genetic background This is also known as interval mapping and is the same as Lander and Botstein s method Lander and Botstein 1989 4 This is an ad hoc model One marker from each chromosome except for the chro mosome on which we are testing is used to control for the genetic background The results of LRmapqtl are scanned and the marker that showed the highest test statis tic from each chromosome is used 5 This is another ad hoc model Two markers from each chromosome are used to con trol for the genetic background They are the top two markers as determined by LRmapqtl In addition all the other markers on the chromosome of the test position that are more than 10 cM away from the flanking markers are also thrown in It may be ad hoc but tends to work best at this time The value of 10 centimorgans can be changed with the w option 6 Model six will be explained in the next subsection 7 Model seven requires the results of a prior run of Zmapqtl and Eqtl Initially the user may want to run Zmapqtl with interval mapping summarize the positions and effects of that analysis using Eqtl and then use those estimates as the covariates in the regression model Virtual markers
145. rated in Figure 1 1 We can then look for correlations between the trait in question and marker genes that have been mapped previously We have also included options for more complex experimental designs including recom binant inbred lines general F lines produced by selfing or random crossing of F _ lines etc The programs in the QTL Cartographer system will need to know the type of exper imental design used to create the data This design is encoded by a string of characters If the letter stands for some integer then the possible crosses will be B SF RF RI T X X SF and T X X RF The B stands for a backcross and the integer attached to it will indicate the parental line to which the F line was crossed to either 1 or 2 If there was repeated backcrossing to one of the parental lines this can be indicated by attaching two integers to the B B indicates that there were j generations of backcrossing to parental line i B is equivalent to B SF stands for selfed intercross lines and the integer indicates the generation i 2 3 RF stands for randomly mated intercross lines RI means recombinant inbred lines and the integer can take on one of three values 0 1 and 2 A 1 indicates RI lines derived by selfing a 2 by sib mating and a 0 means doubled haploid lines The T indicates that the data are the result of a test cross For a test cross genotyping is done on an intercross SF or RF and phenotyping on a cross
146. rited at the quantitative trait loci For each such locus there will be an additive effect a that is defined in the file prepared by Rqtl Genotypes will then have the following values Genotype Q1Q1 Q1Q2 Q2Q2 Genotypic Value 2a a 0 An individual s genotypic value is the sum over all loci of these values This gives a vec tor of genotypic values one entry per individual in the simulated data set The genetic variance is the sample variance of this vector of genotypic values Call it 07 The environ mental variance 0 is defined by 2 1 AS 1 2 8 39 June 22 2000 CHAPTER 2 SIMULATING REFORMATTING DATA where h is the heritability of the trait The extra environmental effect is taken from a nor mal distribution with mean 0 and variance o If the environmental variance is specified the heritability is ignored and the environmental variance is used directly For each in dividual in the data set a random variable with mean zero and variance o is generated and added to the genotypic value This is the phenotypic value of that individual and is printed in the output file 2 3 2 Translating Data Similar to Rmap and Rqtl Rcross can translate files in a pair of special formats The first format is the input format for MAPMAKER QTL These would be the MAPMAKER QTL raw files Simply invoke Reross and specify that the input file is one of these files The parameters that are for simulations are then ignored REMEM
147. s and raw are MAPMAKER genetic linkage map and raw data files respectively These and other files recognized by QTL Cartographer are listed in Table 1 6 28 CHAPTER 1 INTRODUCTION QTL Cartographer Program Extension Contents filetype Rmap map genetic linkage map Rmap out Rqtl qtl QTL model Rqtl out Reross Cro data file markers traits Reross out Ostats qst Qstats Analysis Qstats out LRmapqtl lr Single Marker Analysis LRmapqatl out SRmapqtl sr Stepwise Regression Analysis SRmapqftl out Zmapqtl z IM CIM Results Zmapqtl out JZmapqtl z Multitrait Results JZmapqtl out Prune mpb Pruned genetic linkage map Rmap out Prune crb Pruned data file Rcross out Preplot pit Gnuplot Control file Preplot plt Eqtl eqt Summary of Zmapqtl Results Eqtl out Table 1 5 Standard Filename Extensions and File types for Output Files Program Example Contents filetype Rmap qtlcartm inp genetic linkage map map inp Rmap qtlcartmaps MAPMAKER EXP output mapmaker maps Rqtl qtlcartq inp genetic model file qtls inp Rcross qtlcartc inp data file markers traits cross inp Rcross qtlcart raw MAPMAKER EXP input mapmakerraw Zmapqtl qtlcart z3c Perm test interim file ZipermC out Zmapqtl qtlcart z3e Perm test interim file ZipermE out Zmapqtl qtlcart z3a Bootstrap interim file Ziboot out Eqtl qtlcart z3b Bootstrap summary file Ziboots out Zmapqtl qtlcart z3i Jackknife interim file Zijack out Eqtl qtlcart z3j Jackkn
148. s Remember Rmap will overwrite output files If you specify an output file that already exists Rmap will destroy it when creating a new file For this reason we recommend that all work is done in a working subdirectory on copies of the original input files 2 1 1 Simulating a Map As an exercise in learning to use the programs you can simulate a genetic linkage map The main parameters that you will need to specify are the haploid number of chromo somes average number of markers per chromosome and average intermarker distance between consecutive markers You can also simulate linkage maps in which the telomeres don t have marker information To see how Rmap simulates a genetic linkage map denote the number of chromosomes by c the average number of markers per chromosome by m and the average intermarker distance by d in centimorgans Furthermore the average amount of tail DNA DNA outside the most telomeric markers will be specified by t again in centimorgans The standard deviations of m and d will by om and oq respectively All of these variables can be specified by command line options the resource file or by the interactive menu The standard deviation of t will be o Loa For each chromosome Rmap decides how many markers are on that chromosome by picking a random number from a normal distribution with mean m and standard deviation om Once this is done the amount of DNA between consecutive markers is simulated as a normal
149. s 81 5 8 1 Using MAPMAKER EXP 22 ace ot Sa EA A ee aves 81 5 8 2 Using the MAPMAKER files er AAA 82 6 Input File Formats 85 6 1 Genetic Linkage Maps 000 0 a eee Boe So mes dv ea 85 6 1 1 MAPMAKER output files 220 224 444 DS des 85 6 1 2 Rmap input files Fie so a RES 85 61 3 Bina piOUl purines as ane eaa ee amp WAS a i OE ee eps BE Ne 87 6 2 OTL information aa we eke ae add oe Go a A tae ee 87 621 Rgtlinp ttiles Beanie es E RN 87 6 2 2 Rgtkoutputfil s serna ta nae po A we eS ool each a 88 6 37 Data files eneh it PN ee Le wh ee 89 6 3 1 MAPMAKER raw files 89 6 62 Reross input Mes ota mises SEAS misent ii 89 7 Benchmarks 93 8 UNIX Man Pages 95 Bl OTE CARY e al O aus He edd Bone ene Me Ab we ae ate ee Sed BA ge uns 96 8 22 ARMAR A din gh ihe oe STE aies tt ad rae En mL 102 SANO Ed Bet Ob ee tee icra AS Re RAA Sa ah eA es Sew 106 SV REROS SS M od Ah can sition Be SORT ay asado den lisa SN ENT 109 30 PRUNE sie Si Gece tte E e see eens Re Re ae 112 8 6 OSTATSE ied odd aha con ee ee Rss a ee Boal are te See Bee a a Ls 115 8 7 ERMA POSTE 52 4 mine LU nn ee ees a wd a de Ga Bl ate ae de 117 3 87 SRMAPOT 2d nd lan du WS pe aie oe LK ee a Bd epee 119 8 9 ZMARP OT ind ia ce Ee an ses pct BB ee ae den ei bok 8 Da iaa 121 8 10 J JZMIAPR ODT ia cite tue Sen ee bo eB eee GR tales ates Meet ta ida RS 125 8 11 PREP LOW ta ay ete Cad her wens me ah dut ne I an A oo 129 8 125 BOTE te Bem ete
150. s a molecular map that was used in the analysis of the data with LRmapqtl and Zmapqtl OPTIONS See OTLcart 1 for more information on the global options h for help A for automatic V for non Verbose W path for a working directory R file to specify a resource file e to specify the log file s to specify a seed for the random number generator and X stem to specify a filename stem The options below are specific to this program If you use this program without specifying any options then you will get into a menu that allows you to set them interactively o This requires a filename stem for output Preplot will overwrite the file if it exists and create a new file if it does not If not used then Preplot will use gtlcart The GNUPLOT file will be gtlcart plt in that case m Zmapqtl requires a genetic linkage map This option requires the name of a file con taining the map It should be in the same format that Rmap outputs The default file is gtlcart map 1 This requires an input filename This file must exist It should be in the same format as the output of LRmapqtl The default file is gtlcart lr q This requires an input filename This file may or may not exist It should be in the same format as the output of Rqtl The default file is gtlcart qtl z This requires an input filename This file must exist It should be in the same format as the output of Zmapqtl The default file is gtlcart z T Allows the user to set t
151. s a small quirk in this type of simulation if you are us ing SRmapqtl with stepwise forward backward regression and Zmapqtl with model 6 Sometimes a permuted data set will result in no markers being sufficiently correlated with the trait of interest to be added in the forward phase of the stepwise regression Thus Zmapatl will think there are no markers to be used as covariates and default to interval mapping Thus you may not get the exact number of permutations specified to the above script Simulating Missing Data You can also use Prune to simulate missing data You set the amount of missing marker data you would like to simulate with the M option This will be a percent and should be specified before you invoke the bootstrap option which actually does the simulation Use a value of 3 to tell Prune to randomly set some of the markers to missing Over the entire data set approximately the percentage of markers that had been set with the M option will be set to 10 The results will be in a file with the filename extension crb Similar to simulating missing data some of the markers can be made dominant by using a value of 4 with the bootstrap option The percentage of markers transformed is set with the M The direction of dominance is random Half of those changed will convert the P allele to dominant while the other half will convert the P allele 47 June 22 2000 CHAPTER 2 SIMULATING REFORMATTING DATA 48 Chapter 3
152. s front end for QTL Cartog rapher 19 June 22 2000 CHAPTER 1 INTRODUCTION You can usually download a file by using the get command with a filename On Macin toshes using the server mode may require you to use the put command as you are putting the files onto your local machine rather than getting them from the remote server It is best to do the transfer in an empty subdirectory so that you don t inadvertently delete some important files You will also want to download the README file if you don t already have a copy of it The README file in the pub qtlcart subdirectory will often be more recent than the one in the archive The manual pdf and manpages pdf files are Adobe Portable Document Files of the man ual and the UNIX manpages The manual is the present document and the manpages are meant to be appended to this document You can view or print these files with Adobe Ac robat Reader which is freely available from the Adobe website http www adobe com The following sections indicate how to install the programs onto various computing plat forms 1 4 1 MS Windows Download the file QTLCartWin zip in binary format to your computer s hard drive Move the program to a directory where you want QTL Cartographer to reside Use the program unzip utility to unpack the MS Windows distribution this can be done with a double click The programs will be unpacked in the directory you choose You will want to do this ina direct
153. ses and 30 31 32 34 for those with three marker classes S Tells Eqtl the significance threshold It assumes that the test statistic is significant if greater than this value It is 3 84 by default a Eqtl uses the specified size alpha to determine the significance threshold from the experiment wise permutation results If used the S option is ignored and the sig nificance threshold is set and saved from the experiment wise permutation test re sults The size is 0 05 by default 131 June 22 2000 CHAPTER 8 UNIX MAN PAGES L If used with argument 1 it causes LOD scores to be output rather than the LR statistics It is O by default INPUT FORMAT The input format of the molecular map should be the same as that of the output format from the program Rmap The input format of the individual data should be the same as the output format of the program Rcross The other files should have been created by Zmapqtl Take care that Zmapqtl completed its analysis An incomplete gtlcart z file can cause Eqtl to crash EXAMPLES o Eqtl m example map z example z S 13 2 reprocesses the results of example z based on the map in example map using a significance threshold of 13 2 BOOTSTRAPS JACKKNIVES AND PERMUTATIONS If Zmapqtl was used to do a bootstrap experiment or a permutation test then there will be interim results files With the default filename stem and model 3 there will be files gtlcart z3c and gtlcart z3e
154. sts in Zmapqtl It is more efficient to use Prune and a batch file to do the same job This paradigm will allow users to do permutation tests with any of the programs SEE ALSO Rmap 1 Rqtl 1 Reross 1 Ostats 1 LRmapqtl 1 SRmapqtl 1 JZmapqtl 1 Eqtl 1 Prune 1 Preplot 1 OTLcart 1 AUTHORS In general it is best to contact us via email basten statgen ncsu edu Christopher J Basten B S Weir and Z B Zeng Department of Statistics North Carolina State University Raleigh NC 27695 8203 USA Phone 919 515 1934 124 CHAPTER 8 UNIX MAN PAGES QTL Cartographer 8 10 JZMAPOTL NAME JZmapqtl Multitrait mapping module SYNOPSIS JZmapqtl o output iinput m mapfile E egtfile S srfile trait M Model c chrom d walk n nbp w window I hypo DESCRIPTION JZmapqtl uses composite interval mapping to map quantitative trait loci to a map of molecular markers and can analyze multiple traits simultaneously It requires a molecular map that could be a random one produced by Rmap or a real one in the same format as the output of Rmap The sample could be a randomly generated one from Rcross or a real one in the same format as the output of Rcross In addition the program requires the results of the stepwise linear regression analysis of SRmapqtl for composite interval mapping OPTIONS See QTLcart 1 for more information on the global options h for help A for
155. te the data files into the QTL Cartographer format and then analyze the data 1 Start up Rmap Change the working subdirectory and then the filename stem You can use realdat for the stem Now select item 1 from the menu and enter re aldatm inp Now run the program Rmap should read in the prepared genetic linkage map file and reformat it properly 80 CHAPTER 5 TUTORIAL EXAMPLES QTL Cartographer 2 Start up Rcross Select item 1 from the menu and enter realdatc inp Now run the program Reross should read in the prepared data file match marker names from this data file to those in the map file and reformat the data properly Look at the output 3 Proceed with the analysis programs as in the previous examples Run Qstats LRmapqtl SRmapqtl and Zmapqtl Look at the output after each run 4 Start up Preplot Don t change any parameters Go ahead with the program 5 Start up GNUPLOT From the GNUPLOT command line type in load realdat plt This should display graphical results 6 Start up Eqtl Go ahead with the analysis Look at the output realdat eqt 5 8 Analyzing a MAPMAKER data set 5 8 1 Using MAPMAKER EXP You will need MAPMA KER EXP for this part If you don t want to use MAPMAKER EXP then you can use the already prepared files that come with the distribution Otherwise ftp to genome wi mit edu and cd to distribution mapmaker to get the programs A file sam ple raw com
156. tests are done for interval mapping within Zmapqtl and interim results are stored in the files qtlcart z3c and qtlcart z3e There are two distinct ways to perform the permutation test in QTL Cartographer The first is simply to have Zmapqtl do the permuting and analysis You would then use r with the number of permutations to perform If you choose to do the permutation test entirely within Zmapqtl you must set the number permutations to a value larger than number of permutations already com pleted In this way if you started a permutation test and your machine crashed before the test was complete you can restart Zmapqtl and finish it from where it left off An alternative way to do the permutation test is in a batch file For composite interval mapping one might want to reselect the background markers with SRmapqtl in each per mutation To this end one would need to permute the traits reselect the background markers and then run the composite interval mapping The shell script example in Sec tion 2 4 2 shows how to do this Since Prune has already permuted the traits we want Zmapatl to read in the data do the analysis without permuting the traits and write the interim results to the appropriate files Setting the number of permutations equal to one is a special indicator to Zmapqtl to do just that In the bootstrap new datasets are created from the original by sampling with replacement New datasets are the same size as the original The st
157. that Rmap should output a file in the Rmap out format A 2 indicates that a set of files that can be plotted in GNUPLOT should be created while a 3 indicates that both should be done The option to display the map in GNUPLOT allows a general overview of the spacing of markers If you choose to create the GNUPLOT files then Rmap will write one file per chromosome summarizing the linkage information Each file will have two columns The first indicating the position of the marker from the telomere and the second for the chromosome number The file for chromosome 1 will be Chrom 1 and other files are named accordingly Finally a con trol file Chrom plt will have the plotting commands understood by GNUPLOT This file should be loaded by GNUPLOT to view the linkage map Marker names are not written on the map Input Files Again note that if an input file is specified all options from Chromosomes down in Table 2 1 will be ignored and Rmap will attempt to translate the input file Remember that Rmap overwrites any files with the same name as its output file so avoid giving your input and output files the same name 36 CHAPTER 2 SIMULATING REFORMATTING DATA QTL Cartographer 2 2 Rqtl Given a genetic linkage map Rqtl can place a random set of quantitative trait loci on the map The program simulates the positions and additive and dominance effects It can also reformat a given set of QTLs defined in an input file of filetype
158. that the parameters are set correctly you can selct 0 to run the program If you want to quit simply select 10 Selecting 11 will update the resource file with any parameter changes you have made RESOURCE FILE The resource file keeps track of the most current parameter values used in the programs Each time the user runs a program the program accepts new values for parameters and writes them to the resource file This is unlike the log file which keeps track of the param eters used at the time of running each program The resource file that is generated by the programs in the suite is self documenting Look in the gtlcart rc file HELP FILE Online help requires that OTLcart and all the other programs in the QTL Cartographer suite know where the helpfile is If it is in the current working directory there will be no problem If not then the user should specify the location of the help file in the resource file The line helpfile Path Filename will allow the programs to find the helpfile This line would look different under Windows Macintosh and Unix systems For Unix a help file called gtlcart hlp in the usr local lib subdirectory would be specified by 98 CHAPTER 8 UNIX MAN PAGES QTL Cartographer helpfile usr local lib qtlcart hlp In Windows such a helpfile in c gtlcart would be specified by helpfile c qtlcart qtlcart hlp In Macintosh a help file on hard drive HardDrive in the folder
159. the additive effect of a QTL It is 0 5 by default See Zeng 1992 equation 12 and accompanying text for a discussion of this parameter Itis not the allelic effect of a QTL allele rather it is the shape parameter in the beta distribution 1 2 Allows you to specify the two parameters used to determine the dominance effect of a QTL The effect is simulated from a beta distribution See the manual for more details INPUT FORMAT The input format of the molecular map should be the same as that of the output format from the program Rmap If a file is specified with the i option then that file will be read for the positions and effects of the QTLs The format of this file should be identical to that of the output of Rqtl or of a special format defined in the file qtls inp included with the distribution EXAMPLES Rqtl HZ Places 9 QTLs on the map in Rmap out There is complete dominance of A over a o 5 Rqtl i gtls inp o test qtl Reads the file gtls inp and translates it into the output format of Rqtl The output is written to the file test qtl which is overwritten if it exists REFERENCES 1 Zeng Zhao Bang 1992 Correcting the bias of Wright s estimates of the number of genes affecting a quantitative trait A further improved method Genetics 132 823 839 BUGS The t option for the number of traits is rather primitive at this time The number of QTLs and their effects are randomly determined with means giv
160. the trait identical to the one from Qstats and the re sults of simple linear regression The results are displayed in a table with seven columns The first column indicates the chromosome while the second gives the number of the marker on the chromosome The name of the marker can be found in the genetic linkage map file The next two columns correspond to the parameters in the linear model Equa tion 3 2 Column three is the intercept and column four the slope of the least squares regression line fit to the data Column five is a likelihood ratio test statistic for the model and column six is the F statistic Column seven is the tail probability of the F statistic as suming one and n 1 degrees of freedom in the numerator and denominator respectively Asterisks attached to these probabilities indicate significance of the F statistics Significance at the 5 1 0 1 and 0 01 levels are indicated by one two three and four asterisks respectively The results of running LRmapqtl are used in Zmapqtl for analysis models four and five see Section 3 4 2 3 2 3 Permutation Tests The r option tells LRmapqtl to perform a permutation test Churchill and Doerge 1994 The argument to r indicates how many permutations should be performed In each per mutation the phenotypes are shuffled relative to the genotypes over individuals and the analysis is redone The results are summarized at the end of the LRmapqtl output file 54 CHAPTER3
161. timates for the parameters for that trait These files will end in z where is a number indicating the trait There will be one other file ending in z0 that contains the results of the joint liklihood ratio The joint results file ending in z0 will have four columns corresponding to the chromo some marker markername and test position Then there will be column giving the joint liklihoods for the test position for all possible hypothesis tests see next section The single trait files ending in z will have the results for the numbered trait In addi tion to the chromosome marker markername and test position the likelihood ratio and parameter estimates will be given All columns are labelled and the parameters are the same as explained in the Zmapqtl section 3 5 3 Usage Hints Trait Selection You can select traits to include in the analysis in three ways Suppose that you have t traits in your data file 1 Set the trait to analyze at 0 so that no traits except those beginning with a plus sign are analyzed You would need to edit the cro file first to prepend a to all traits you want in the analysis 2 Set the trait to a value in the range 1 t inclusive You will then get single trait results for the selected trait 3 Set the trait to a value greater than t All traits will be put in the analysis unless they begin with a minus sign Hypothesis tests You need to set the hypothesis test for SF and
162. tio of columns 5 and 6 At the end of the Ostats output file there will be a summary of missing data for each individual in the data set Ostats will indicate the number of marker systems quantitative traits and categorical traits It will then have a table with seven columns Column 1 is for the individual Column 2 indicates the number of markers for which the individual is typed and Column 3 indicates a percent Columns 4 and 5 do the same for traits while columns 6 and 7 summarize the data for categorical traits Something to keep in mind is that some of the analyses require large sample sizes For example if the sample sizes are too small the ECM algorthm may fail in Zmapqtl When difficulties in analysis are encountered check the missing data summaries in the Ostats output Such problems often correspond to areas with a lot of missing data 3 1 2 Segregation Ostats also tests for adherence to Mendelian segregation at all marker loci For a given locus suppose there are r genotypic classes Let p be the expected frequency and n the observed count for the ith class For a sample of size n the expected counts will be np and the observed frequencies will be n n We can construct a test statistics based on a contigency table r r 2 ni mpi ni n pi Ti rn 2 np 2 Pi i 1 or a comparison of likelihoods To 2 5 ni ln n ln np i 1 Both T and T should have a x distribution with one degree of freed
163. tion allows you to specify a seed for the random number generator You can use this to repeat simulations to see if the same answers are obtained If you don t use this option the random number seed is set to the number of seconds since some arbitrary past date for example 1 January 1970 for Sun Workstations The random number seed is printed to the output files of the programs on the first line This means that if you don t specify a random number seed each file should have a unique identifier associated with it This identifier will also be written to the log file 25 June 22 2000 CHAPTER 1 INTRODUCTION Verbosity For debugging purposes and simply to inform the user about what is happening many diagnostic messages will be printed out as the programs run The user can turn these diagnostic messages off When the messages are displayed we refer to this as the verbosity mode The verbosity mode can be turned off by using the the V option This means that the time and summary of options will not be printed on the standard output at runtime This is a useful flag for batch files Most of the messages printed to the screen are also printed to the log file Automatic Mode By default when the user starts up a program an interactive menu for setting program options is displayed The opposite of this is the Automatic mode The A flag turns off the interactive setting of program options This is another flag useful for batch programmi
164. to have intermarker distances vary at random t You can simulate maps where there are no markers on the telomeres with this option Give this option a value of tails and Rmap puts an average of tails Morgans of genetic material on the ends of the chromosomes By default it is 0 0 If the standard devi ation for intermarker distance is greater than 0 0 then then the amount of flanking DNA will have a normal distribution with mean given here and standard deviation proportional to that of the standard deviation of intermarker distances M Allows you to specify an alternate simulation mode If the M option is used with a value of 1 then the intermarker distance will be used as the chromosome length so you should make it longer and the markers are placed on the chromosomes following the uniform distribution 103 June 22 2000 CHAPTER 8 UNIX MAN PAGES INPUT FORMAT Rmap recognizes three types of files The first is the Rmap out format that Rmap itself creates The second is a special format defined in the example file map inp included in the distribution The third format is the output of MAPMAKER If the input file is a MAPMAKER output file Rmap translates this file into its own format If the input file is already in the correct format Rmap will output it dependant upon the flag given to the g option The units of intermarker distances will be in centiMorgans in the output EXAMPLES o Rmap Map out c 23 vm 3 vd 1 t
165. under H 62 CHAPTER 3 ANALYSIS QTL Cartographer The last 13 columns are not shown because they are only valid for F design experiments They would all be zeros if shown The output for an F design or any design in which dominance effects can be estimated is similar but has more information For an F gt you can estimate additive a and dominance d parameters at each position Thus there are four hypotheses e Ay a 0 d 0 eH a 0 d 0 e Hy a 0 d40 e H3 a40 d40 and twelve full columns of output corresponding to all possible hypothesis tests and pa rameter estimates The 21 columns correspond to Chromosome of test position Left flanking marker of test position Absolute position from left telomere in Morgans H3 Ho H3 H H3 H Estimate of a the additive effect under H4 Estimate of a the additive effect under H3 Estimate of d the dominance effect under Ha Likelihood ratio test statistic for Likelihood ratio test statistic for Likelihood ratio test statistic for AND 0 FP WN Ha Estimate of d the dominance effect under H3 Ha Ha Likelihood ratio test statistic for a N Likelihood ratio test statistic for pe H Ho H Ho k H3 Ho H Ho A H Ho 7 H3 Ho A S for Ay S for Ao S for H3 r for Ha eo r for cs r for mi Hi oO O1 r for Hua I r for pa 00 r for
166. ut SRmapqtl will append the file if it exists and create a new file if it does not If not used then SRmapqtl will use gtlcart sr i This requires an input filename This file must exist It should be in the same format as the output of Rcross The default file is gtlcart cro m SRmapqtl requires a genetic linkage map This option requires the name of a file containing the map It should be in the same format that Rmap outputs The default file is gtlcart map t Use this to specify which trait SRmapqtl will analyze If this number is greater than the number of traits then all traits will be analyzed The default is to analyze trait 1 only M This tells SRmapqtl what type of analysis to perform Use a 0 for forward stepwise FS regression a 1 for backward elimination BE and a 2 for forward regression with a backward elimination step at the end FB It is probably best to use Model 2 here 119 June 22 2000 CHAPTER 8 UNIX MAN PAGES F Requires a real number in the range 0 0 to 1 0 This is a threshold p value for adding markers in model 2 during the forward stepwise regression step The default is 0 05 B Requires a real number in the range 0 0 to 1 0 This is a threshold p value for deleting markers in model 2 during the backward elimination step It should probably be the same as the previous option The default is 0 05 INPUT FORMAT The input format of the molecular map should be the same as that of the output format
167. with a plt extension Suppose that this file was stem plt You can start up GNUPLOT and issue the command gnuplot gt load stem plt to see the plot specified by stem plt See the GNUPLOT manual for more information on this program Williams and Kelley 1993 Of special interest may be the different types of printers supported by GNUPLOT If you choose postscript as your terminal type in Preplot then you will find a pair of lines on the stem plt file that look like this set term postscript set output stem ps You can change the token postscript token in that file to any printer that GNUPLOT supports and sent the stem ps file to that printer 74 Chapter 5 Tutorial Examples 5 1 General tactics and notes Below we outline some general exercises using QTL Cartographer These exercises were used in a class Statistics 5910 and in the Summer Institute in Statistical Genetics at North Carolina State University These computer exercises were done in the Statistics Instruc tional Computing Laboratory SICL which is equipped with Sun workstations running Solaris 2 5 but the exercises can be done on any platform that QTL Cartographer runs on As a general rule we suggest creating a separate subdirectory folder for each data set Copy the original input files into that subdirectory This will help to organize your work In addition since you will be working with copies your original files will be safe Begi
168. wn 1 in the output Data by Markers and Traits One way to organize the data is by markers For each marker you give the genotypes of the individuals The order of the individuals has to be the same for each marker Below is an example After the start markers the program expects a repeating sequence of marker name then n marker genotypes where n is the sample size The marker names should match those in the map inp file start markers Markerl_1 22222 1 Marker1 2 2 2 2 1 1 90 CHAPTER 6 INPUT FILE FORMATS QTL Cartographer Marker1_3 2 0 2 2 Marker1_4 T2222 Marker1_5 2222 2 Marker2_1 2112 2 Marker2_2 2 2 2 Marker2_3 2 2 11112222 Marker2_4 2111111122 stop markers The traits are encoded in the same fashion After the start traits tokens the program expects a repeating sequence of trait name and then n values for the sample The order of the individuals has to be the same as in the markers In the following example a period indicates missing trait data missingtrait start traits Trait_1 000 548 021 llo DD 00 Gak TE 6 4 Trait_2 P5015 3 Loe 2 241 25 5 2528 16a 267 1 33 2 L624 stop traits indicates the end of the trait data Other traits otraits will be stored as character strings These will be things such as sex brood eye color etc Each token should be less than
169. y the log file s to specify a seed for the random number generator and X stem to specify a filename stem The options below are specific to this program If you use this program without specifying any options then you will get into a menu that allows you to set them interactively o This requires a filename stem for output Prune will overwrite the file ending in crb if it exists and create a new file if it does not If not used then Prune will use gtlcart crb If the map is recreated then a new map file will be written to qtlcart mpb by default or a file ending in mpb with the specified stem i This requires an input filename This file must exist It should be in the same format as the output of Rcross The default file is gtlcart cro m Prune requires a genetic linkage map This option requires the name of a file contain ing the map It should be in the same format that Rmap outputs The default file is gtlcart map I Sets the interactive level A zero means that Prune will do what it needs to without asking the default for bootstraps permutations or missing data simulations A one means that the user will be put into a repeating loop to manipulate the data set It has a value 1 by default but using the b option disables it 112 CHAPTER 8 UNIX MAN PAGES QTL Cartographer M This sets a level for the elimination of individuals with this much missing marker data or for the simulation of missing or dominant markers when
170. you prefix the name of one of these variables with a plus sign then it will be incorporated into the linear model Names of the other traits 1 Sex 2 Line In LRmapqtl this would consider both Sex and Sex by Marker interaction terms In Zmapqtl and SRmapqtl the Sex by Marker term wouldn t be incorporated but the Sex factor would All other variables that have no sign at the beginning of their names will be ignored in the analysis For the above example a pair of models will be considered y bo bizi boSex b3Sex x x e 3 3 y bo t beSex e 3 4 53 June 22 2000 CHAPTER 3 ANALYSIS Option Default Explanation i qtlcart cro Data Input File 0 qtlcart Ir Output File m qtlcartmap Genetic Linkage Map File r 0 Number of permutations t 1 Trait to analyze Table 3 2 Command Line Options for LRmapqtl The output will give probabilities that the marker is significant Table 3 2 shows the command line options specific to LRmapqtl As with Qstats there are few parameters to change The t option allows you to specify a trait to analyze It is trait 1 by default If you only have one trait you can ignore this option If your data set has more than one trait you can analyze a specific trait by using t with an integer from 1 to the number of traits If you want LRmapqtl to analyze all traits use a value greater than the number of traits 3 2 2 Output LRmapqtl prints out a histogram of

User Manual for QTL Cartographer

Contents

Download Pdf Manuals

Related Search

Related Contents