Dosage to best guess, handler for ricopili script, to correct best guess to include low freq variants

bg_fix.pbs

#PBS -lnodes=1
#PBS -lwalltime=02:05:00

#!/bin/bash

#Make walltime a function of N?

while getopts i:f:o:p:d:n: option
do
  case "${option}"
    in
      i) indir=${OPTARG};;
      f) file_list=${OPTARG};;
      p) plinkloc=${OPTARG};;
      d) outdir=${OPTARG};;
      n) nodeuse=${OPTARG};;
    esac
done

 
 #Write the start and stop points of the file
 jstart=$((($PBS_ARRAYID-1)*$nodeuse +1))
 jstop=$(($PBS_ARRAYID*$nodeuse))

 for j in $(seq -w $jstart 1 $jstop)
 do
  file_use=$(awk -v lineno=$j '{if(NR==lineno) print}' $file_list)
  perl /home/maihofer/ricopili/rp_bin_manc/daner_bg3_am  $file_use --freq_th 0 --info_th 0.02  --ploc $plinkloc --indir $indir  --outdir $outdir &
 done
wait
 


  

dosage_to_bg.sh

#Specify where probability format genotypes are stored
probdir=/home/maihofer/yehuda/WRBY/qc/imputation/dasuqc1_pts_wrby_mix_am-qc.hg19.ch.fl/qc1

#Specify where output will be stored 
outdir=/home/maihofer/yehuda/WRBY/test

#Specify working directory (error logs will go here)
workingdir=/home/maihofer/yehuda/WRBY


ls $probdir | grep .gz$ > files_to_make_dose.txt

#Number of commands to run is a function of the number of files
ncommands=$(wc -l files_To_make_dose.txt | awk '{print $1}' )

#Make a job code, where 'nodesize' processes will run on each node simultaneously
 nodesize=16
 nodeuse=$(($nodesize - 1))

#Total number of jobs = Number of commands / number of commands used per job, rounded up 
totjobs=$(( ($ncommands + $nodeuse - 1 ) / $nodeuse ))


qsub bg_fix.pbs  -t 1-$totjobs -d $workingdir -F "-f files_To_make_dose.txt -d $outdir -i $probdir -n $nodeuse -p /home/maihofer/yehuda/plink"