05_map_parallel.sh

#!/bin/bash

#############
## written by RMPeery 2025
## Doesn't always work well - if fail use alt_map_loop.sh
## USE: ./map_parallel.sh
#############

set -eou pipefail

# Define dir
# change to match dir structure
DIR="/home/user/path" 
RDS="$DIR/reads/Pcin_trimRds"
REF="$DIR/ref/Phyci1_bwaRef"
OUT="$DIR/bam"

mkdir -p "$OUT"

# function to map reads started by ChatGPT4 from loop version (alt_map_loop.sh)
# modified by RMPeery to use paths and streamline
mapping() {
    FWDRDS="$1"
    # grab reverse reads; check file extensions
    REVRDS="${FWDRDS/_1P.fastq.gz/_2P.fastq.gz}"
    # grab sample ID
    ID=$(basename "$FWDRDS" _1P.fastq.gz)
    # check in on processing
    echo "Processing sample: $ID"
    # run mapping and sorting
    bwa-mem2 mem -M -t 24 "$REF" "$FWDRDS" "$REVRDS" | \
        samtools sort -@ 16 -o "$OUT/${ID}.mapSort.bam"
    # check in on processing
    echo "Mapping for $ID completed."
}

export -f mapping

# run in parallel (j = no. concurrent jobs)
find "$RDS" -name "*_1P.fastq.gz" | parallel -j 3 mapping {}

echo "Mapping complete for all files."