salmon_flength_dist

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
This script creates Salmon flenDist.txt files from fld.gz files,
which is parseable by MultiQC.
"""

import struct
import gzip
import argparse
import os


parser = argparse.ArgumentParser()
parser.add_argument('dataset', help='dataset to create flenDist.txt Salmon files for')
args = parser.parse_args()


def get_fld(path):
    '''
    Returns a tuple containing the fragment length distribution.
    Number of bins (integers): 1001 (from 0 to 1000).
    '''
    with gzip.open(path) as fld_file:
        fld = struct.unpack('i' * 1001, fld_file.read())
        return fld


def Main():
    dat = args.dataset
    subfolders = [f.path for f in os.scandir(str(dat) +
                                             '/expression/salmon/') if f.is_dir()]
    for sub in subfolders:
        if os.path.exists(sub + '/aux_info/fld.gz'):
            print('Writing flenDist.txt to {}...\n'.format(sub))
            fld_tuple = get_fld(sub + '/aux_info/fld.gz')
            with open(sub + '/libParams/flenDist.txt', 'w') as out_file:
                for val in fld_tuple:
                    out_file.write(str(val) + " ")
        else:
            print('{}/aux_info/fld.gz does not exists\n'.format(sub))


if __name__ == '__main__':
    Main()