o-get-sample-info-from-fasta

#!/usr/bin/env python
# -*- coding: utf-8 -*-

# Copyright (C) 2010 - 2012, A. Murat Eren
#
# This program is free software; you can redistribute it and/or modify it under
# the terms of the GNU General Public License as published by the Free
# Software Foundation; either version 2 of the License, or (at your option)
# any later version.
#
# Please read the COPYING file.

import sys
import operator

import Oligotyping.lib.fastalib as u
from Oligotyping.utils.utils import pretty_print as pp


fasta = u.SequenceSource(sys.argv[1])

samples = {}
while next(fasta):
    if fasta.pos % 1000 == 0:
        sys.stderr.write('\rreads processed so far: %d' % (fasta.pos))
        sys.stderr.flush()
    sample_name = '_'.join(fasta.id.split('_')[:-1])

    if sample_name in samples:
        samples[sample_name] += 1
    else:
        samples[sample_name] = 1

sys.stderr.write('\rSamples and read counts found in the FASTA file:\n')
for sample, read_count in sorted(iter(samples.items()), key=operator.itemgetter(1), reverse = True):
    print('%-30s %s' % (sample, pp(read_count))) 

print()
print()
print('Total number of samples: ', pp(len(samples)))
print('Total number of reads: ', pp(fasta.pos))
print()
fasta.close()