User Manual for GIGI v1.03
Contents
1. 101111222 102120022 103111212 6 allele frequencies of the dense marker file line 7 Each row contains the allele frequencies of the dense markers The first column is the allele frequency of allele 1 the second column is the allele frequency of allele 2 etc Each row must sum to 1 The allele frequencies file is space delimited allele frequencies of marker 1 allele frequencies of marker 2 allele frequencies of marker 3 eg 0 4 0 6 0 2 0 3 0 5 0 8 0 2 Output files Three or Four output files are created in the directory where GIGI is run 1 impute geno The imputed genotype file has the same format as the input dense genotype file 2 impute prob The imputed genotype probabilities file displays the genotype probabilities for each marker For each bi allelic marker three columns are used to store the probability of genotype configuration 1 1 2 1 and 2 2 For each tri allelic marker the probabilities are displayed and space delimited with 1 1 2 1 2 2 3 1 3 2 3 3 For multi allelic markers in general the appropriate number of columns are used to store all the combinations Markers are separated by tabs e g ID genotype_11_markerl genotype_21_markerl genotype 27 markerl genotype_11_marker2 genotype 21 marker2 genotype_22_marker2 101 0 9 0 10 0 8 0 0 2 102 0 9 0 1 0 100 3 impute consistentIV The consistency file counts the number of IVs consistent with the observed genotypes of each dense marker This f2. potentially useful for genotype imputation on trios Since the current version of GIGI does not leverage information from linkage disequilibrium you may want to use existing linkage disequilibrium based software that supports imputation on trios e g BEAGLE Appendix A GIGI v1 00 used an old format for the parameter file It is not recommended for new users For old users if you want to continue to use this old format use the flag longParam when you run 11 GIGI This file format allows you to specify the exact filename and location of the output files e g GIGI example oldparamFormat txt longParam The old file format line 1 filename of the pedigree meiosis file line 2 number of pedigrees current max is 1 Just use the value 1 line 3 filename of the dense marker file line 4 IVs file line 5 number of saved realizations in the IVs file line 6 filename of the map positions of the framework markers file CM Haldane map distance line 7 filename of the map positions of the dense markers file cM Haldane map distance line 8 filename of the allele frequencies of dense markers file line 9 random seed number e g 1234 line 10 filename of the output 1 called genotypes line 11 Use the value 2 This is the calling method to use 1 most likely genotype calling 2 threshold based calling line 12 the first threshold of threshold based calling tl to call both alleles this number should be
3. Foundation either version 3 of the License or at your option any later version This program is distributed in the hope that it will be useful but WITHOUT ANY WARRANTY without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE See the GNU General Public License for more details There is NO WARRANTY for the program to the extent permitted by applicable law In no event unless required by applicable law will any GIGI copyright holder be liable to you for damages including any general special incidental or consequential damages arising out of the use or inability to use the program including but not limited to loss of data or data being rendered inaccurate or losses sustained by you or third parties or a failure of the program to operate with any other programs Acknowledgement Supported by funding from the National Institutes of Health grants R37GM046255 P01HL030086 and PS0AG05136 I would also like to thank Elizabeth Marchani and Mohamad Saad for providing feedback as one of the beta testers of GIGI 13 Future Direction 1 Enable GIGI to read marker genotype data in the long file format similar to the format used in BEAGLE e g each row contains a marker 2 Convert the code from C to C
4. User Manual for GIGI v1 03 Author Charles Y K Cheung cykc uw edu Department of Biostatistics University of Washington Last Modified on 7 14 2013 Contents TAT OCI CTO E 3 Citing GIG WE 3 SOWIE E EE 3 Files in GIGI software distribution eege 3 Inferring ENEE E 3 Installing GIG EG 5 Running GIG E 5 EH 5 Making The Parameter Fil issiisgchernenareseineio noreo ee aE REEE ENEE NE ia 6 File E 7 EE H Frequently Asked Questions FAQs srs iuccauss eege GE 10 Appendix As iar su saceanussn nen n E E E E A E RE E ERE EEEE 10 PDE BE ysx5scsacwssusstanwncduaetoussadnndeseiousesutassasdseltinuusbrniesasconudomabsunncduenviaaeutivestesuamiMastanncdsumbincebrsnsceswentes 11 EE 12 ER 12 Acknowledgement arseeni Ke n E E TET ERE ATELE E aE 12 F t re Direction ee Ee a a E es 13 Introduction GIGI is a C program to impute genotypes in pedigree data It imputes missing genotypes of dense markers conditional on genotypes available on a subset of relatives in the pedigree and framework genotypes used to infer Inheritance Vectors IVs Our pedigree based genotype imputation approach consists of two steps The first step is to infer IVs at the positions of framework markers using gl auto which is a MCMC based program that is part of the MORGAN package The second step is to impute dense genotypes by GIGI using the IVs output by gl auto and pedigree structure file from MORGAN GIGI is developed under the linux environment In this documentat
5. between 0 5 and lt 1 e g 0 9 line 13 the second threshold of threshold based calling t2 to call 1 of 2 alleles if GIGI can t call both alleles under the threshold specified in line 12 a number that is between the first threshold lt 1 e g 0 9 line 14 filename of output 2 estimated genotype probabilities line 15 filename of output 3 number of IVs consistent with each observed dense markers Appendix B Creating the Pedigree Meiosis file manually from the console output of gl auto Using a text editor we open the console_output txt and fetch the line that begins with name name pa name ma Compnt pat meio mat meio We copy this line and table below We paste this table to another file and save it The file includes the header line and looks like this name name pa name ma Compnt pat meio mat meio 2100 6 0 D Tt 0 2100 21 0 0 1 0 0 2100 25 0 0 1 0 0 2100 29 0 0 1 0 0 2100 31 0 0 1 0 0 2100 39 0 0 1 0 0 12 2100_907 0 0 Me MO 8 2100 908 2100901 2100902 1 2 1 2100_909 0 0 L 0 0 2100 910 2100 901 2100 902 1 4 3 2100 911 2100 901 2100 902 1 2100 915 2100 907 2100 908 1 until the end of table oo O NaN ChangeLog The ChangeLog describes the recent changes to the program It is located in the root directory of the uncompressed file License GIGI is free software you can redistribute it and or modify it under the terms of the GNU Lesser General Public License as published by the Free Software
6. e order used in gl auto Each line contains the position of a marker position of Marker1 position of Marker2 position of Marker3 position of MarkerN eg 1 0 2 0 3 0 4 0 e map positions of dense markers line 5 Similar to the marker map for framework panel of markers the marker map of dense markers is a text file which contains the map distance in centi Morgans cM based on the Haldane map function Markers must be ordered in ascending order Each line contains the position of a marker eg 0 5 0 7 0 9 1 1 LAS b dense marker genotype file line 6 The dense markers are the markers that we want to impute on some individuals in the pedigree The dense marker file contains the genotypes of observed individuals a This file is space delimited b The markers should be sorted by ascending map positions c Alleles are labeled numerically starting from 1 2 in ascending order We use 0 to indicate a missing allele Jf the original genotypes are in alpha numeric users first need to convert the genotypes to an indexed numerical format GIGI supports imputation of multi allelic markers Each row of the file contains the genotypes for an individual This file does not include the header line The first column specifies the ID of an individual The ID can be an alphanumeric string ID allelel_ marker allele2_ marker allelel_marker2 allele2_ marker2 allelel1_marker3 allele2_ marker3 An example of this file looks like
7. found in the example directory under example param txt In GIGI_v1 01 the parameter file is organized as follows line 1 filename of the pedigree meiosis file line 2 framework IVs file line 3 number of sampled realizations in the IV file line 4 filename of the map positions of the framework markers file CM based on Haldane map function line 5 filename of the map positions of the dense markers file CM based on Haldane map function line 6 filename of the dense marker genotype file line 7 filename of the allele frequencies of the dense markers file line 8 Call_method tt 3 numeric values separated by space eg 2 0 6 0 8 On line 8 we specified up to 3 values 1 Call method use the value 1 to call the most likely genotype and 2 to use threshold based calling 2 and 3 If threshold based calling is used then the threshold of t 0 8 and t2 0 9 are set as defaults This pair of thresholds can be changed by the user by specifying the 2 and 3 numerical values The thresholds should be between 0 5 and lt 1 If only t is specified tz will be set to t 1 t 2 The format of the parameter file used here is different than that of an older format used in GIGI v1 00 See Appendix A for the older format and how we can continue to use the older format of the parameter file not recommended Notes I suggest using the absolute paths of the filenames instead of relative paths A relative path is relative to the direct
8. ge Equilibrium If the framework markers are multiallelic this spacing should be greater e g not denser than 1 marker per 2 cM for good MCMC mixing 2 What do I have to be careful about the genetic map that I am using The map positions are in map distances based on the Haldane map function instead of Kosambi map function or sequence positions If this map is based on the Kosambi map function the user will need to convert the map positions to Haldane map function using an appropriate conversion method Also since recombination fractions are relative to each other we strongly encourage the user to generate both the framework map positions and the dense marker positions at the same time 3 Can GIGI process multiple pedigrees No In version 1 01 GIGI can only process 1 pedigree at a time For this reason gl auto should also be run individually for each pedigree 4 Can I use GIGI to impute genotypes on unrelated individuals No 5 What should I know when I use GIGI to impute genotypes on trios Since there is no way to infer recombination on trios using unphased genotypes gl auto produces an empty inheritance vectors file From the pedigree meiosis file GIGI recognizes that this inheritance vectors file is empty but will proceed to impute genotypes For trios imputing genotypes using GIGI is equivalent to imputing genotypes conditional on the pedigree structure and minor allele frequencies Information about linkage disequilibrium is
9. ile is used for a sanity check A number that is low compared to the number of IVs used may indicate incompatibility between observed genotypes and the sampled IVs 4 impute dosage Ifall dense markers to be imputed are di allelic the dosage file will also be created The dosage summarizes the expected percent of the allele 1 in each genotype dosage of a genotype 1 P genotype is 1 1 0 5 P genotype is 1 2 Markers are separated by space The file is formatted as ID dosage marker 1 dosage marker 2 dosage marker 3 dosage marker 4 101 0 5 1 0 875 0 102 1 0 5 0 5 0 Note allele 1 is not necessarily the rare allele It is the user s responsibility to define what allele 1 is 10 Frequently Asked Questions FAQs 1 How do I choose framework markers Because the framework markers are assumed to be sparse we want to choose framework markers that are informative about which chromosomes are being transmitted at the framework loci First framework markers typed on a large number of subjects tend to be most informative Second framework markers that are multi allelic tend to be more informative than di allelic markers if they are available Third if framework markers are SNPs markers with high minor allele frequencies in the sample tend to be more informative Fourth framework markers should be moderately but not too dense eg not denser than 1 marker per 0 3 cM because of concern about MCMC mixing and violation of the assumption of Linka
10. ion we use the following terminology Framework markers are a relatively sparse set of markers that are used to infer IVs on a chromosome of interest Dense markers are markers with missing genotypes on some subjects that we want to impute For example these dense markers may be genotypes obtained from sequence data or from a dense SNP panel and may be typed on fewer and even different subjects in the pedigree See the publication describing GIGI below for more information Citing GIGI Cheung CYK Thompson E A Wijsman E M GIGI An approach to effective imputation of dense genotypes on large pedigrees American Journal of Human Genetics 92 4 504 516 Software URL http faculty washington edu wijsman software shtml Files in GIGI software distribution GIGI software code and its dependency files the Mersenne random number generator e example folder Inferring IVs using gl auto The first step to impute genotypes is to use framework markers to infer IVs For this purpose we use gl auto a program in the MORGAN package that is freely available at http www stat washington edu thompson Genepi MORGAN Morgan shtml To infer IVs in gl auto we need to supply the required files in MORGAN format 1 Pedigree file 2 Marker file this is a composite file that contains the map positions of framework markers in centiMorgans assuming the Haldane map function allele frequencies of framework markers and genotype data of framework marke
11. ory containing the executable program The parameter file in the example folder is created using a relative path line 1 user creates this file from the gl auto s console output In version 1 01 GIGI can only process 1 pedigree at a time line 2 when you run gl auto you should instruct gl auto to print Meiosis Indicators instead of Founder Genome Labels line 3 this corresponds to the number of samples that the user actually prints to the meisois indicator file File Formats Examples of these files are provided in the example directory refer to the param txt for the filename of these files a pedigree meiosis file line 1 The creation of this file was discussed in the Inferring IVs using gl auto section The pedigree meiosis specifies the pedigree structure and the index of meioses that GIGI needs to use to read in the corresponding Inheritance Vectors GIGI imputes genotypes on subjects specified in this pedigree The content of this file is generated by gl auto b framework IVs file line 2 The Inheritance Vectors file describes the descent pattern of chromosomes at the positions of the framework markers It is the output file that gl_ auto generates c map positions of framework markers line 4 The marker map positions of the framework markers file is a text file which contains the map distance in centi Morgans cM based on the Haldane map function Markers must be ordered in ascending order and consistent with th
12. rs 3 Parameter file used to run gl auto Refer to the documentation of MORGAN for guidance on setting up these files and on running gl auto Example files used to infer IVs using gl auto are included under the example gl auto_example directory Overview From running gl auto we need to obtain 2 files a Framework IVs file GIGI uses this file which contains IVs at framework positions b Pedigree meiosis file this file contains the information about the structure of the pedigree on which we want to im c pute genotypes This file is different from the pedigree file used in gl auto In addition to the pedigree structure this file also contains information that GIGI needs to determine how the Inheritance Vectors Meiosis indicators are organized i e the i line of the meisosis indicator belongs to which subject in the pedigree and whether this meiosis indicator is this person s maternal or paternal chromosome We need to create this file from the console output of gl auto Obtaining the files a Framework IVs file In gl auto s parameter file we must instruct gl auto to display the output as Meiosis Indicators instead of Founder Genome Labels Please make sure we use the option output meiosis indicators instead of output founder genome labels in the control parameter file of el auto b Pedigree Meiosis file We need to create the pedigree meiosis file b from the console output of gl auto It is very ea
13. sy to make this file When we run gl auto the program prints a huge amount of output to the console This console output actually contains the content of the pedigree meiosis file that we need to extract 1 In order to extract this content we first need to direct the console output to a file by using the gt directive so we can subsequently extract the content from this file ie gl_auto gl auto parameter file gt glauto_console_output txt 2 Then we extract the pedigree meiosis content from the console output to a new file To simplify the creation of this file use the Perl script extractPedMeiosis pl e Usage perl extractPedMeiosis pl glauto_console_output txt FILENAME PED MEIO o We need to have Perl installed in linux o assuming glauto console output txt is in the same directory as extractPedMeiosis pl e Alternatively this file can also be easily created by the user See Appendix B for the creation of this file manually and also for an example of how this file looks like Installing GIGI Simply unzip the files navigate to the code directory and type make If make does not work go to the GIGI cpp s directory and install the program by g GIGI cpp o GIGI GIGI was compiled successfully under Ubuntu 10 04 Mandriva Linux release 2009 0 and RedHat release 5 8 with g version 4 1 2 4 3 2 and 4 5 2 Running GIGI GIGI accepts a parameter file To run GIGI type GIGI lt parameter file gt lt options g
14. t To run the example file go to the main GIGI program directory and type GIGI example param txt Options The flags are case sensitive Flag Purpose verbose to run GIGI with extended output to the console seed NUM _ to change the default seed NUM should be a number between 1 and 999999 prog NUM to print the progress of imputation at every NUM markers longParam to run GIGI using the old GIGI v1 00 s format of the parameter file General users should ignore outD to specify the absolute directory path of where the output files will be created If this flag is missing the output files will be saved to the user s current directory e g home charles GIGI_ output no character at the end of the directory name An example of running GIGI with additional options GIGI example param txt verbose outD home charles GIGL_ output Making the Parameter File The parameter file tells GIGI where to look for the required files and where to save the output files GIGI needs the pedigree meiosis file that user prepares from the output of gl auto dense marker file IVs file from gl auto map of the sparse marker file map of the dense marker file and allele frequencies of the dense marker file It produces the called genotype file genotype probability file and another file that prints the number of IVs that are consistent with the observed dense genotypes for sanity check An example of the parameter file is
Download Pdf Manuals
Related Search