Python script for running multiple bowtie2 jobs with a variety of input fastq and read groups.

INSTRUCTIONS.md

1. Modify the sample_info.txt appropriately. I'd recommend doing this in excel and exporting the sheet as a CSV file. The first lane is a header and cannot be changed.
2. Run the bowtie2 commands with python run_bowtie2.py

run_bowtie2.py

import shlex
from subprocess import Popen, PIPE


def run_bowtie2(args_dict):
    print 'running {0[sample_id]}\n'.format(args_dict)

    cli1 = "\
    bowtie2 \
    -p {0[processors]} \
    --very-sensitive-local \
    --rg ID:{0[plate]}.{0[lane]} \
    --rg SM:{0[sample_id]} \
    --rg LB:{0[library_id]} \
    --rg PL:ILLUMINA \
    --rg-id {0[plate]}.{0[lane]} \
    -x {0[bwt2_idx]} \
    -1 {0[fq1]} \
    -2 {0[fq2]}".format(args_dict)

    cli2 = "samtools view -Sb -"

    cli1 = shlex.split(cli1)
    c1_out = Popen(cli1, stdout=PIPE, stderr=PIPE)
    cli2 = shlex.split(cli2)
    c2_out = Popen(cli2,
                   stdin=c1_out.stdout,
                   stdout=open('{0[sample_id]}.bam'.format(args_dict), 'wb')).communicate()
    c1_out.stdout.close()              # Allow p1 to receive a SIGPIPE if p2 exits.

if __name__ == '__main__':

    IDs = None
    with open('sample_info.txt','rU') as fin:
        for count, line in enumerate(fin):
            line_parts = line.strip().split(",")
            if count is 0:
                IDs = line_parts
                continue
            if len(line_parts) != len(IDs):
                break
            else:
                b = dict(zip(IDs, line_parts))
                run_bowtie2(b)

sample_info.txt

processors,plate,lane,bwt2_idx,sample_id,library_id,fq1,fq2
4,P1,4,test_run.RADnome,sample1,sample1,fq1.10k.fq,fq2.10k.fq
4,P1,4,test_run.RADnome,sample1,sample1,fq1.10k.fq,fq2.10k.fq