Core_variants.sh

#!/bin/bash
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=1
#SBATCH --cpus-per-task=1
#SBATCH --mem=20gb
#SBATCH -t 02:00:00
#SBATCH --mail-type=AL
#SBATCH --mail-user=pmorrell@umn.edu
#SBATCH -p ram1t
#SBATCH -o %j.out
#SBATCH -e %j.err

set -e
set -o pipefail

module load bcftools


#    This script is intended to partition a VCF file into "core" and "noncore" variants
#    We will use a GFF file to find positions of core or noncore genes
#    We can

GFF3=
VCF=
OUT_DIR=
CORE=
NONCORE=

#    Cut the GFF3 file down to just gene positions to reduce our search space
GENES=$(grep 'gene' ${GFF3})

#    Create a sorted bed file with only CORE or only NONCORE gene positions
CORE_POS=$(grep -E -f ${CORE} <(echo ${GENES}) | cut -f 1,4,5 | sort -k1,1 -k2,2n -k3,3n)
NONCORE_POS=$(grep -E -f ${NONCORE} <(echo ${GENES}) | cut -f 1,4,5| sort -k1,1 -k2,2n -k3,3n)

#    Use bcftools to cut the VCF down to individual SNPs within genes from each list
bcftools --regions-file <(CORE_POS) ${VCF} --output-type z --output ${OUT_DIR}/CORE_vcf.gz
bcftools --regions-file <(NONCORE_POS) ${VCF} --output-type z --output ${OUT_DIR}/NONCORE_vcf.gz