def generate_wiggle(reads, ofps, num_threads=1, contig=None ):
all_args = []
for chrm_length, chrm in sorted(izip(reads.lengths, reads.references)):
strands = ['+', '-'] if len(ofps) == 2 else [None,]
# skip regions not in the specified contig, if requested
if contig is not None and clean_chr_name(chrm) != clean_chr_name(contig):
continue
for strand in strands:
ofp = ofps[strand]
assert (ofp, reads, chrm, chrm_length, strand ) not in all_args
all_args.append((ofp, reads, chrm, chrm_length, strand ))
if num_threads == 1:
for args in reversed(all_args):
populate_cvg_array_for_contig( *args )
else:
ps = [None]*num_threads
while len( all_args ) > 0:
for i in xrange(num_threads):
if ps[i] is None or not ps[i].is_alive():
ps[i] = multiprocessing.Process(
target=populate_cvg_array_for_contig,
args=all_args.pop() )
ps[i].start()
break
time.sleep( 0.1 )
for p in ps:
if p is not None: p.join()
for fp in ofps.values(): fp.close()
return
def write_array_to_opstream(ofp, buffer, buff_start, chrm, chrm_length,
strand):
"""write buffer to disk, buff_start determines the start of buffer in
genomic coordinates.
"""
chrm = fix_chrm_name(clean_chr_name(chrm))
prev_pos = 0
prev_val = buffer[0]
for pos, val in enumerate(buffer[1:]):
# make sure this doesn't extend past the end of the chromosome
# bedGraphs are 0-based, so use chrm_length-1
if buff_start + pos + 1 >= chrm_length:
pos = chrm_length - buff_start - 1
break
if val != prev_val:
if prev_val > 1e-12:
write_val = -prev_val if strand == '-' else prev_val
line = "%s\t%i\t%i\t%.2f" % (chrm, buff_start + prev_pos,
buff_start + pos + 1, write_val)
ofp.write(line + "\n")
prev_pos, prev_val = pos + 1, val
if prev_val > 1e-12:
write_val = -prev_val if strand == '-' else prev_val
line = "%s\t%i\t%i\t%.2f" % (chrm, buff_start + prev_pos,
buff_start + pos + 1, write_val)
ofp.write(line + "\n")
return
def parse_arguments():
allowed_assays = ['cage', 'rampage', 'rnaseq', 'polya']
import argparse
parser = argparse.ArgumentParser(
description='Extract elements from a bed or gtf file.')
parser.add_argument( 'region',
help='A region string of the form contig:strand:start-stop')
parser.add_argument( 'input_file', type=file,
help='An bed file to extract elements from')
args = parser.parse_args()
chrm, strand, poss = args.region.strip().split(":")
start, stop = [int(x) for x in poss.replace(",", "").split('-')]
return ( GenomicInterval(clean_chr_name(chrm), strand, start, stop),
args.input_file )
def parse_arguments():
allowed_assays = ['cage', 'rampage', 'rnaseq', 'polya']
import argparse
parser = argparse.ArgumentParser(
description='Extract elements from a bed or gtf file.')
parser.add_argument(
'region', help='A region string of the form contig:strand:start-stop')
parser.add_argument('input_file',
type=file,
help='An bed file to extract elements from')
args = parser.parse_args()
chrm, strand, poss = args.region.strip().split(":")
start, stop = [int(x) for x in poss.replace(",", "").split('-')]
return (GenomicInterval(clean_chr_name(chrm), strand, start,
stop), args.input_file)
def fix_chr_name(x):
return "chr" + clean_chr_name(x)
def fix_chr_name(x):
return "chr" + clean_chr_name(x)