gatk4_pon_make_input.sh

#! /bin/bash

#########################################################
#
# Platform: NCI Gadi HPC
# Usage: sh gatk4_hc_make_input.sh /path/to/cohort.config
# Version: 1.0
#
# For more details see: https://github.com/Sydney-Informatics-Hub/Somatic-ShortV
#
# If you use this script towards a publication, support us by citing:
#
# Suggest citation:
# Sydney Informatics Hub, Core Research Facilities, University of Sydney,
# 2021, The Sydney Informatics Hub Bioinformatics Repository, <date accessed>,
# https://github.com/Sydney-Informatics-Hub/Germline-ShortV
#
# Please acknowledge the Sydney Informatics Hub and the facilities:
#
# Suggested acknowledgement:
# The authors acknowledge the technical assistance provided by the Sydney
# Informatics Hub, a Core Research Facility of the University of Sydney
# and the Australian BioCommons which is enabled by NCRIS via Bioplatforms
# Australia. The authors acknowledge the use of the National Computational
# Infrastructure (NCI) supported by the Australian Government.
#
#########################################################

if [ -z "$1" ]
then
	echo "Please provide the path to your cohort.config file, e.g. sh gatk4_pon_make_input.sh ../cohort.config"
	exit
fi

config=$1
cohort=$(basename "$config" | cut -d. -f 1) 
ref=../Reference/hs38DH.fasta
scatterdir=../Reference/ShortV_intervals
scatterlist=$(ls $scatterdir/*.list)
if [[ ${#scatterlist[@]} > 1 ]]; then
	echo "$(date): ERROR - more than one scatter list file found: ${scatterlist[@]}"
	exit
fi
bamdir=../Final_bams
outdir=../$cohort\_PoN
logs=./Logs/gatk4_pon
INPUTS=./Inputs
inputfile=${INPUTS}/gatk4_pon.inputs

mkdir -p ${INPUTS}
mkdir -p ${logs}

# Collect sample IDs from samples.config
# Only collect IDs from normal samples (labids ending in -B)
while read -r sampleid labid seq_center library; do
	if [[ ! ${sampleid} =~ ^#.*$ && ${labid} =~ -B.?$ || ${labid} =~ -N.?$ ]]; then
		samples+=("${labid}")
	fi
done < "${config}"

echo "$(date): Writing inputs for gatk4_pon_run_parallel.pbs for ${#samples[@]} samples and 3200 tasks per sample to ${inputfile}"
# Resource requirements:
echo "$(date): Normal samples found include ${samples[@]}"

rm -rf ${inputfile}

# Write gatk4_pon_mutect2.inputs file, using nt=1
while IFS= read -r intfile; do
	for sample in "${samples[@]}"; do
		out=${outdir}/${sample}
		bam=${bamdir}/${sample}.final.bam
		logdir=${logs}/${sample}
		interval="${scatterdir}/${intfile}"
		echo "${ref},${sample},${bam},${interval},${out},${logdir}" >> ${inputfile}
	done
done < "${scatterlist}"

num_inputs=`wc -l ${inputfile}`
ncpus=$(( ${#samples[@]}*48*4 ))
mem=$(( ${#samples[@]}*192*4 ))
num_tasks=`wc -l $inputfile | cut -d' ' -f 1`
echo "$(date): Number of tasks in $inputfile: $num_tasks"
echo "$(date): Suggested compute to request in gatk4_pon_run_parallel.pbs: walltime=02:00:00,ncpus=${ncpus},mem=${mem}GB,wd"