QMSim's User Guide - University of Guelph
Contents
1. Fimpute User s Guide Version 2 2 Semex Alliance Ontario and Centre for Genetic Improvement of Livestock University of Guelph Ontario Mehdi Sargolzaei Jacques Chesnais and Flavio Schenkel Jan 2014 Disclaimer The FlImpute software is distributed AS IS solely for non commercial use The authors and their organizations will not be liable for any general special incidental or consequential damages arising from using FImpute By the use of this software the user agrees to bear all risk resulting from using the software Citing FImpute Sargolzaei M J P Chesnais and F S Schenkel 2014 A new approach for efficient genotype imputation using information from relatives BMC Genomics 15 478 DOI 10 1186 1471 2164 15 478 Contact Mehdi Sargolzaei msargol uoguelph ca IMPORTANT If you have a problem with a specific imputation run please include report txt and control files with your message Overview FImpute ef impute was mainly developed for large scale genotype imputation in livestock where hundreds of thousands of individuals are genotypes with different panels Flmpute uses an overlapping sliding window approach to efficiently exploit relationships or haplotype similarities between target and reference individuals The process starts with long windows to capture haplotype similarity between close relatives After each chromosome sweep the window size is shrunk by a constant factor allowing2. and v3 are the maximum numbers of SNP in the same order of the chips Type Optional Default Automate Note If set to zero for a specified chip the program uses default value trim_segm_pop Description Trim head and tail of segment in population imputation Usage trim_segm_pop v v is the portion of segment to be trimmed Type Optional turnoff _fam Description This command turns off family imputation Usage turnoff fam Type Optional turnoff pop Description This command turns off population imputation Usage turnoff_pop Type Optional save_partial Description Save partial calls 6 7 8 and 9 See hap_lib_file command for partial codes Usage save_partial Type Optional Note In output statistics partial calls are treated as missing 10 save_genotype Description Saves genotypes instead of haplotypes heterozygous loci are saved as code 1 Usage save_genotype File format ID genotype codes Genotype codes 0 A1A1 1 A1A2 or A2A1 2 A2A2 5 missing Type Optional save_hap_lib Description Save haplotype library built from reference individuals Usage save_hap_lib option Option diplotype This options force the program to combine two haplotypes together to save memory File format SNP IDs are listed in the first line Haplotypes start from the second line with no space between haplotype codes Haplotype codes 1 Al 2 A2 5 missing When diplotype optio
3. ID length is 30 characters Multiple pedigree files can be read in as ped_file filename1l filename2 Pedigree files with overlap will be combined to create one pedigree with unique IDs If pedigree file is not defined family imputation is automatically turned off If sex chromosome is to be analyzed pedigree file should always be defined In this case if pedigree is not known set parents to missing but provide correct sex hap_lib_file Description Usage Type Input Format Haplotype library file hap_lib_file filename option filename is input haplotype library file name option diplotype Compressed format Two haplotypes are combined in one line mr value Missing rate threshold Haplotypes with larger missing rate will be discarded Default is 0 2 Optional First line should contain SNP IDs Haplotypes start from the second line There should be no space between haplotype codes Haplotype codes 1 Al 2 A2 5 missing When diplotype option is specified the codes are 0 A1A1 1 treated as missing A2A2 A1A2 A2A1 missing A1 second haplotype is missing A2 second haplotype is missing A1 first haplotype is missing A2 first haplotype is missing Note Multiple haplotype library files can be inputed as Hap_lib_file filenamel filename2 output_folder Description Output folder Usage output_folder foldername folde
4. all frequencies missing rate and minor allele frequency Missing calls are ignored for statistics on MAF and calls 0 1 and 2 stat_snp_imp txt Reports statistics on SNPs after imputation stat_anim txt Reports statistics on individuals genotypes ID chip number call frequencies homozygosity and missing rate Missing calls are ignored for statistics on homozygosity and calls 0 1 and 2 stat_anim_imp txt Reports statistics on individuals genotypes after imputation 134 org_vs_imp txt Reports the difference between original genotypes and imputed genotypes Large changes in the original genotypes may indicate progeny parent conflict Animals are sorted by change ref_pop txt Contains list of reference individuals used for population phasing and imputation report txt Detailed report on the steps carried out by the software 14 Running the application FImpute control filename o If control file name is not specified program will prompt the user to enter it Option o forces the program to overwrite output folder if it already exists 15 Example 1 title Familyt tpopulation imputation genotype file example data genotypes ld txt snp info file example data snp info txt ped file example data ped txt output _folder outputl parentage test ert _mm 0 02 remove conflict add_ungen min fsize 4 save_sep save hap lib diplotype njob 5 Note ped file and add_ungen command
5. for shorter haplotype similarity arising from more distant relatives to be taken into account Because closer relatives usually share longer haplotypes while more distant relatives share shorter haplotypes the algorithm simply assumes that all individuals are related to each other at different degrees Note that if pedigree information is provided FImpute makes use of this information for more accurate imputation Pedigree information becomes more important as the low density panel becomes sparser High input genotype quality is the key for accurate imputation The current version of FImpute can handle SNP markers only Input control file The program requires a control file in which various parameters for imputation should be specified The input parameter file must be in ASCII format C like comments can be used to add descriptive comments anywhere in the parameter file All commands end with a semicolon title Description Set an arbitrary title Usage title string string indicates an arbitrary title Type Optional Default None genotype_file Description Usage Type Input Format Note Input genotype file genotype_file filename option filename is input genotype file name option phased Indicates that input genotypes are already phased Mandatory ID chip number genotype calls First line is header line Chip number starts from 1 and should be the order of chip in SNP info file There
6. is no space between genotypes and genotypes should be coded as O and 2 for homozygotes 1 for heterozygote and 5 for missing genotypes The number of genotypes for each animal must be exactly the same as the number of SNP on the chip for which the animal was genotyped with Genotype calls 0 A1A1 1 A1A2 or A2A1 2 A2A2 5 missing Maximum ID length is 30 characters Multiple genotype files can be read in as genotype_file filenamel filename2 snp_info_file Description Usage Type Input Format Note This file contains SNP map information snp_info_file filename option filename is input SNP map file name option chrx v specifies chromosome X Note that v should not contains pseudo autosomal regions of X Mandatory SNP ID chromosome number base pair position order of SNP for each chip First line is header line Maximum SNP ID length is 50 characters Maximum number of chips is 10 Positions of SNP on each chromosome should be defined as accurate as possible since FImpute uses base pair position to model recombination 1 000 000 base pairs is considered as 1 cM ped_file Description Usage Type Input Format Note Pedigree file ped_file filename filename is input pedigree file name optional ID sire ID dam ID sex First line is header line IDs can be alphanumeric and do not need to be sorted sex should be coded as M and F Maximum
7. n is specified the codes are 0 ALAIL 2 A2A2 3 ALA2 4 A2A1 5 missing 6 Al second haplotype is missing 7 A2 second haplotype is missing 8 Al first haplotype is missing 9 A2 first haplotype is missing O Type ptional 11 random_fill Description Random filling imputation based on allele frequency This command is useful to access minimum accuracy by random sampling of alleles based on their frequency Usage random fill Type Optional system Description Run a system command after FImpute finishes all processes Usage system command Command is a system command Type Optional 12 Output files genotypes_imp txt Contains ID chip number haplotypes Haplotype codes 0 A1A1 1 Unphased heterozygous 2 A2A2 3 A1A2 4 A2A1 5 missing 6 Al 7 A2 8 Al 9 A2 First allele is paternal and the second is maternal If save_genotype is specified in control file program outputs only genotype codes i e 3 and 4 are converted to 1 and 6 7 8 and 9 are set to 5 genotypes_imp_chip0 txt Contains ID chip number 0 imputed genotypes for ungenotyped individuals This file is created if command add_ungen with option save_sep is specified snp_info txt Contains SNP ID chromosome number position excluded_snp_list txt Contains list of excluded SNPs stat_snp txt Reports statistics on SNPs SNP ID chromosome number positions c
8. or the owner Because the output files are not executable the execute permission is not allowed If execute permission is specified the program automatically ignore it However the execute permission is always set for the output folder ped_depth Description Set maximum generations to be traced for family imputation Usage ped_depth value value is the number of generations Type Optional Default 10 Note If set to zero only parents are used In this case the accuracy is higher but the missing rate is also higher min_nprg_imp Description Set minimum number of progeny required for imputation from progeny Usage min_nprg_imp value value is the number of progeny Type Optional Default 4 min_nsib_imp Description Set minimum number of sibs required for sib imputation Usage min_nsib_imp value value is the number of sib Type Optional Default 4 min_segm_len_fam Description Set minimum segment length for family imputation Usage min_segm_len_fam L1 L2 L3 L1 L2andL3 are segment lengths in the same order of the chips Type Optional trim_segm_fam Description Trim head and tail of segment in family imputation Usage trim_segm_fam v v is the portion of segment to be trimmed Type Optional Default 0 05 ref Description Set parameters for population imputation Usagel ref n options n is the number of reference individuals option parent Consider only individuals wi
9. progeny parent mismatches default is 0 01 Error rate threshold to find progeny parent matches default is 0 005 Error rate threshold to find individuals with identical genotypes default is 0 001 Error rate threshold to find sex conflict for males only default is 0 05 When a progeny parent conflict is detected set the conflicting parents to missing When pedigree information is not available or pedigree is not complete the program as default creates a pseudo pedigree which is only used in population imputation part This command skips search for pseudo pedigree Skip parentage test exclude_snp Description Usage Type Exclude user defined SNP exclude_snp filername filename is the file name that contains SNP list to be excluded no header line Optional exclude_chr Description Usage Type Exclude SNP that are located on specified chromosomes exclude _chr cl c2 c3 el 2 3 jf are chromosome numbers Optional exclude_chip Description Exclude the specified chip s Usage exclude chip cl c2 c3 cl c2 c3 are chip numbers Type Optional njob Description Number of jobs to be run in parallel Usage njob n Type Optional Default 1 chmod Description Set desired permission on output folder and files Usage chmod value value is a 3 digit number similar to that of Unix s chmod Type Optional Note Always set read and write permissions f
10. rname is output folder name Type Mandatory add_ungen Description Add ungenotyped individuals in imputation process and try to impute genotypes for these individuals Usage add_ungen option option min_fsize c Add ungenotyped individuals with minimum family size of c Default is 4 output_min_fsize d Save imputed genotypes for ungenotyped individuals with minimum family size of d Default is 4 output_min_call_rate e Save imputed genotypes for ungenotyped individuals with minimum call rate e Default is 0 9 Type Optional Note Adding ungenotyped individuals improves the overall imputation accuracy but imputation might not be highly successful for ungenotyped individuals with small family size parentage_test Description Check for parentage errors Usage parentage_test option option chip v find_match_cnflt find_match_mp find_match_ugp find_identical ert_mm vl ert_m v2 ert_i v3 ert_s v4 remove_conflict pseudo_ped_off off Type Optional Default Parentage test is on Chip to be used for parentage test v can be the chip number or can be file name pointing to pre defined SNP list Find match for individuals having conflict with their parent Find match for individuals with missing parent might be time consuming Find match for individuals with ungenotyped parent might be time consuming Find animal pairs with identical genotypes Error rate threshold to find
11. s can be removed when imputing from 50k to higher density Example 2 title Population imputation genotype file example data genotypes ld txt snp info file example data snp info txt output _folder output2 njob 5 Example 3 title Random fill in based on allele frequency genotype file example data genotypes ld txt snp info file example data snp info txt output _folder output3 random fill njob 5 Example 4 title Imputation using already built haplotype library genotype file example data genotypes ld txt snp info file example data snp _info txt ped file example data ped txt hap lib file outputl hap library txt diplotype ref 0 output _folder output4 njob 5 16
12. th progeny option male Consider only male individuals option female Consider only female individuals Usage ref filename filename contains user defined list of reference individuals multiple files can be selected Files should be separated by space Type Optional Default ref 20000 target Description Specify list of individuals to be imputed using population information Usagel target filename filename is user defined list of target individuals multiple files can be selected Files should be separated by space Usage target c1 c2 c3 cl c2 c3 are chip numbers Note This command is ignored for family imputation i e all individuals are considered for family imputation sw_shrink_factor Description Shrink factor 0 02 0 5 for sliding windows Usage sw_shrink factor vl v2 v3 vl v2 and v3 are shrink factors in the same order of the chips Type Optional Default 0 08 sw_overlap Description Set amount of overlap 0 01 0 95 for sliding windows Usage sw_overlap v1 v2 v3 vl v2 and v3 are overlap values in the same order of the chips Type Optional Default 0 75 Sw_min_size Description Set minimum sliding window size Usage sw_min_ size v1 v2 v3 vl v2 and v3 are the numbers of overlap SNP in the same order of the chips Type Optional Default 4 Sw_max_size Description Set maximum sliding window size Usage sw_max_size vl v2 v3 vl v2
Download Pdf Manuals
Related Search