bcf_msmc_prep.sh

#!/bin/bash
#SBATCH -D /home/dmvelasc/Projects/Prunus/Analysis/VCF
#SBATCH -o /home/dmvelasc/Projects/Prunus/slurm-log/%j-stdout-vcf.txt
#SBATCH -e /home/dmvelasc/Projects/Prunus/slurm-log/%j-stderr-vcf.txt
#SBATCH -J vcf
#SBATCH -p bigmemm
#SBATCH -n 1
#SBATCH -c 1
#SBATCH -a 1-12
set -e
set -u


# To compare with MSMC and Josh's SAMtools results using reduced sampling of almond and peach genotypes

x=$SLURM_ARRAY_TASK_ID
i=$(( x-1 ))

# Declare array
declare -a sample=(PD03 PD04 PD05 PD06 PD07 PD09 PP02 PP03 PP04 PP05 PP13 PP15)

# Load zlib 1.2.8
module load zlib

# Declare directories
dir1="/home/dmvelasc/bin"					# software binary directory
dir2="/home/dmvelasc/Projects/Prunus/Analysis/VCF"		# VCF directory
dir3="/home/dmvelasc/Data/references/persica-SCF"		# FASTA reference directory
dir4="/group/jrigrp3/Velasco/Prunus/BAM"

##### FILTER SAMPLES, CALL A FEW STATISTICS #####
# pileup and filtering of genomic SNPs
# with -A removed (consider removing -v, which only calls variant sites)
#"$dir1"/samtools mpileup -BRug -t DP,SP -f "$dir3"/Prunus_persica_v1.0_scaffolds.fa -b all.prunusbam.txt | "$dir1"/bcftools call -O z -vm -f GQ,GP - > "$dir2"/prunus-A.raw.vcf.bzip
#"$dir1"/bcftools index "$dir2"/prunus-A.raw.vcf.bzip

for j in {1..8}; do
	"$dir1"/bcftools view -O z -V indels -o "$dir2"/"${sample["$i"]}"_"$j"_20_msmc.flt.vcf.bzip -s "${sample["$i"]}".bam -r scaffold_"$j" -i "%QUAL>=30 && %AVG(DP)>20 && %MIN(GQ)>=30" "$dir2"/prunus-A.raw.vcf.bzip
done