def subtract_mouse_reads(
summary_file, in_fastq, out_fastq, sub_fastq, num_mismatches):
# Accept this as a mouse read if it contains less than or equal to
# num_mismatches mismatches from the mouse genome.
from genomicode import filelib
from genomicode import genomelib
# List the reads that look like mouse.
mouse_reads = {}
for d in filelib.read_row(summary_file, header=1):
if not d.NM: # ignore missing alignments
continue
if int(d.NM) <= num_mismatches:
mouse_reads[d.query_name] = 1
outhandle = open(out_fastq, 'w')
subhandle = open(sub_fastq, 'w')
for x in genomelib.read_fastq(in_fastq):
title, sequence, quality = x
x = title
if x.startswith("@"):
x = x[1:]
x = x.split()[0] # BAM file only contains the first part.
if x in mouse_reads:
genomelib.write_fastq(title, sequence, quality, subhandle)
else:
genomelib.write_fastq(title, sequence, quality, outhandle)
def copy_fastq_file(in_filename, out_filename, num_samples):
from genomicode import genomelib
outhandle = open(out_filename, 'w')
for i, x in enumerate(genomelib.read_fastq(in_filename)):
if i >= num_samples:
break
genomelib.write_fastq(*x, handle=outhandle)
def copy_fastq(in_filename, out_filename, MAX_READS=None):
from genomicode import genomelib
in_iter = genomelib.read_fastq(in_filename)
out_handle = open(out_filename, 'w')
i = 0
while MAX_READS is None or i < MAX_READS:
i += 1
x = in_iter.next()
if not x: # no more reads
assert i # make sure at least 1 read
break
genomelib.write_fastq(*x, **{"handle" : out_handle})