Shuffle PE fastq files with preserved pairing

shuf_pe.sh

#!/bin/sh

# Usage: ./shuf_pe.sh <reads1> <reads2>
# The output files 1_shuffled.fastq and 1_shuffled.fastq
# contain shuffled reads with preserved pairing.

# For shuffling PE reads:
# https://gist.github.com/iam28th/49a245427ea2b8ed5f1f9889c13468bf

# remove files with output filenames if exist
# because awk will append to them
rm -f 1_shuffled.fastq
rm -f 2_shuffled.fastq

# interleave lines from forward and reversed files
paste -d '\n' $1 $2 | \
    awk '{

    # read 8 lines
    lines[1] = $0;
    for (i = 2; i <= 8; ++i)
        getline lines[i];

    # and print them tab-separated on a single line
    for (i = 1; i <= 8; ++i)
        printf("%s%s", lines[i], i == 8 ? "'"\n"'" : "'"\t"'")
    }' | \

    # shuffle
    shuf | \

    # replace all tabs back to newlines
    tr '\t' '\n' | \

    # print odd and even lines to different files
    awk 'NR % 2 == 1 { print >> "1_shuffled.fastq" }
         NR % 2 == 0 { print >> "2_shuffled.fastq" }'