sra_downloader.py

import subprocess
import shutil
import argparse
import pathlib

parser = argparse.ArgumentParser(description='Download all SRAs from a GSM.')
parser.add_argument('GSMs', metavar='GSMs', type=str, nargs='+',
                   help='GSMs to download, can be more than one')
parser.add_argument('-g', '--guess-title',
                   help='If specified, guess the title of each GSMs and use it to store SRRs',
                   action='store_true')


args = parser.parse_args()

for GSM in args.GSMs:
    print(f'downloading {GSM} from NCBI GEO...')

    output_folder=f'./{GSM}'
    if args.guess_title: 
        raw_title=subprocess.check_output(
            f'esearch -db sra -query {GSM} '
            '| efetch -format docsum '
            '| xtract -pattern DocumentSummary -element Title',
            shell=True
        ).decode()
        title = raw_title.split(GSM+':')[1].split(';')[0].strip()
        print(f'Extract title "{title}" from a full title "{raw_title}"')
        output_folder=f'./{title}'
    pathlib.Path(output_folder).mkdir(parents=True, exist_ok=True)

    SRRs=subprocess.check_output(
        f'esearch -db sra -query {GSM}'
        '|efetch -format docsum '
        '|xtract -pattern DocumentSummary -element Run@acc',
        shell=True)

    SRRs = SRRs.decode().strip().split('\t')
    print(f'found SRRs: {SRRs}')
    for SRR in SRRs:
        print(f'downloading {SRR}')

        SRRs=subprocess.check_output(
            f'fastq-dump -F {SRR} --split-files --gzip -O {output_folder}',
            shell=True)