def main():
args = get_args()
outfile = fastq.FasterFastqWriter(args.output)
read1 = fastq.FasterFastqReader(args.read1)
read2 = fastq.FasterFastqReader(args.read2)
rc = 0
sys.stdout.write("Interleaving reads (1 dot = 10,000 pairs): ")
for r1, r2 in izip(read1, read2):
assert r1[0].split(" ")[0] == r2[0].split(" ")[0], \
"Read FASTQ headers mismatch."
outfile.write(r1)
outfile.write(r2)
if rc != 0 and rc % 10000 == 0:
sys.stdout.write(".")
sys.stdout.flush()
rc += 1
outfile.close()
read1.close()
read2.close()
def main():
args = get_args()
r1 = fastq.FasterFastqWriter(args.read1)
r2 = fastq.FasterFastqWriter(args.read2)
# read all of our files into fastq iterators
rc = 0
reads = fastq.FasterFastqReader(args.input)
sys.stdout.write("Splitting reads (1 dot = 10,000 pairs): ")
for read in reads:
if read[0].split(' ')[1].split(':')[0] == '1':
r1.write(read)
first = read[0].split(' ')[0]
else:
assert first == read[0].split(' ')[0], "File does not appear interleaved."
r2.write(read)
if rc != 0 and rc % 10000 == 0:
sys.stdout.write('.')
sys.stdout.flush()
rc += 1
print ""
reads.close()
r1.close()
r2.close()
def main():
"""main loop"""
motd()
args = get_args()
# setup tags object
tags = Tags(args.tagmap, args.section, args.no_correct)
# create output files
tags.create_zip_files(args.output)
if not args.hamming:
# vectorize the levenshtein function so we only call once per read
distance = numpy.vectorize(levenshtein)
else:
# vectorize the hamming function so we only call once per read
distance = numpy.vectorize(hamming)
read = 0
for f in glob.glob(os.path.join(args.reads, '*_R1_*')):
print "Working on ", f
read1 = fastq.FasterFastqReader(f)
# get basename of R1 file
read1_basename = os.path.basename(f)
index = fastq.FasterFastqReader(
os.path.join(args.reads, read1_basename.replace('R1', 'R2')))
read2 = fastq.FasterFastqReader(
os.path.join(args.reads, read1_basename.replace('R1', 'R3')))
# read all of our files into fastq iterators
for r1, i, r2 in izip(read1, index, read2):
if read % 100000 == 0:
print "{:,}".format(read)
if not filtered(r1):
# see if we need to trim the index
idx, idx_qual = get_index(i, args.tag_length)
# get index sequence differences from tags
dist = distance(tags.seqs, idx)
# get quality values
good_quality = get_quality(idx_qual, args.min_qual,
args.min_mean_qual)
# find tags with 0 distance from other tags
positions = numpy.where(dist == 0)[0]
# if not a perfect match, check distance 1 matches
if positions.size == 0 and good_quality and idx not in tags.no_correct:
positions = numpy.where(dist == 1)[0]
if positions.size == 1 and good_quality:
# assert headers match
assert r1[0].split(' ')[0] == r2[0].split(
' ')[0], "Header mismatch"
# get tag for match
match = tags.seqs[positions[0]]
# change header to add tag
r1 = change_read_num(r1, 1, match)
r2 = change_read_num(r2, 2, match)
# write to output
tags.files[match][1].write(r1)
tags.files[match][2].write(r2)
# put low quality tags into their own file
elif (positions is not None) and (len(positions)
== 1) and not good_quality:
tags.files['lowqual'][1].write(r1)
tags.files['lowqual'][2].write(i)
tags.files['lowqual'][3].write(r2)
# put everything else into unknown
else:
tags.files['unknown'][1].write(r1)
tags.files['unknown'][2].write(i)
tags.files['unknown'][3].write(r2)
# if for some reason there are reads not passing filter,
# put those into unknown, too.
else:
tags.files['unknown'][1].write(r1)
tags.files['unknown'][2].write(i)
tags.files['unknown'][3].write(r2)
read += 1
read1.close()
index.close()
read2.close()
tags.close_zip_files()