bin/jbrowse_vcf_tools/generate_indiv_vcfs.sh

   1 #!/bin/bash
   2 #-------------------------------------------------------------------------------------------------------------------------------
   3 # NAME
   4 #
   5 # generate_indiv_vcfs.sh
   6 #
   7 # SYNOPSIS
   8 # Shell script for creating multiple versions of individual VCF files from a single multivcf and imputed dosage file.
   9 #
  10 # ./generate_indiv_vcfs.sh -v [multivcf file] -d [dosage file]
  11 #
  12 # To run, this script requires create_indiv.pl
  13 #                      finish_indiv.pl
  14 #-------------------------------------------------------------------------------------------------------------------------------
  15
  16
  17 #--------------------------------------------------------------------------------
  18 # 1 Parse command line arguments:
  19 #-------------------------------------------------------------------------------
  20
  21 while [[ $# > 1 ]]
  22 do
  23 key="$1"
  24
  25 case $key in
  26     -v|--multi.vcf)
  27     MULTI_VCF="$2"
  28     shift
  29     ;;
  30     -d|--dosage)
  31     DOSAGE="$2"
  32     ;;
  33 esac
  34 shift
  35 done
  36 echo MULTI_VCF  = "$MULTI_VCF"
  37 echo DOSAGE  = "$DOSAGE"
  38 if [ -z "$MULTI_VCF" ] || [ -z "$DOSAGE" ]
  39 then
  40     echo "Trouble reading command line arguments, make sure
  41     -v [multi vcf file] and
  42     -d [dosage file] are both specified";
  43     exit
  44 fi
  45
  46
  47 #----------------------------------------------------------------------------------
  48 # 2 create a nearly empty vcf file for each accession in the multi-vcf
  49 #----------------------------------------------------------------------------------
  50
  51 echo Creating starter vcf files...
  52
  53 mkdir output
  54
  55 ./create_indiv.pl -v $MULTI_VCF -o output
  56
  57 #--------------------------------------------------------------------------------
  58 # 3 add genotype data to complete indiv vcf files. then generate filt and imputed files too.  Requires long operations, so do it in parallel to speed it up
  59 #-------------------------------------------------------------------------------
  60
  61 ls output/* | parallel -j 30 --gnu --verbose "./finish_indiv.pl -v $MULTI_VCF -d $DOSAGE -f {}"
  62
  63