def _getBestFrontBackRecord(self, domFN):
"""Parses DOM output from phmmer and fill in best_of_front, best_of_back
bestOf: sequence id ---> DOMRecord
"""
logging.info("Get the best front & back primer hits.")
# bestOf_ = {} # key: sid --> primer name --> DOMRecord
best_of_front = defaultdict(lambda: None)
best_of_back = defaultdict(lambda: None)
reader = DOMReader(domFN)
for r in reader:
# allow missing adapter
if r.sStart > 48 or r.pStart > 48:
continue
if r.sid.endswith('_front'): # _front
bestOf = best_of_front
r.sid = r.sid[:-6]
elif r.sid.endswith('_back'): # _back
bestOf = best_of_back
r.sid = r.sid[:-5]
else:
raise ClassifierException(
"Unable to parse a read {r} in phmmer dom file {f}.".
format(r=r.sid, f=domFN))
if r.sid not in bestOf:
bestOf[r.sid] = {}
if (r.pid in bestOf[r.sid] and
bestOf[r.sid][r.pid].score < r.score) or \
(r.pid not in bestOf[r.sid]):
bestOf[r.sid][r.pid] = r
return (best_of_front, best_of_back)
def test_Reader(self):
"""Test DOMReader."""
inDOMFN = op.join(self.testDir, "data/test_DOMReader.dom")
reader = DOMReader(inDOMFN)
res = [r for r in reader]
expected_0 = DOMRecord(
"F1",
"m131018_081703_42161_c100585152550000001823088404281404_s1_p0/45/ccs",
23.7, 0, 31, 31, 2170, 2201, 3931)
expected_1 = DOMRecord(
"R1",
"m131018_081703_42161_c100585152550000001823088404281404_s1_p0/45/ccs",
16.2, 0, 25, 25, 3906, 3931, 3931)
self.assertEqual(res[0], expected_0)
self.assertEqual(res[1], expected_1)
def _getChimeraRecord(self, domFN, opts):
"""Parses phmmer DOM output from trimmed reads for chimera
detection, return DOMRecord of suspicious chimeras, which
have primer hits in the MIDDLE of the sequence.
"""
logging.info("Identify chimera records from {f}.".format(f=domFN))
# sid --> list of DOMRecord with primer hits in the middle
# of sequence.
suspicous_hits = defaultdict(lambda: [])
reader = DOMReader(domFN)
for r in reader:
# A hit has to be in the middle of sequence, and with
# decent score.
if r.sStart > opts.min_dist_from_end and \
r.sEnd < r.sLen - opts.min_dist_from_end and \
r.score > opts.min_score:
suspicous_hits[r.sid].append(r)
return suspicous_hits
def _getBestFrontBackRecord(self, domFN, min_score):
"""Parses DOM output from phmmer and fill in best_of_front, best_of_back
bestOf: sequence id ---> DOMRecord
"""
logging.info("Get the best front & back primer hits.")
# bestOf_ = {} # key: sid --> (score, primer
best_of_front = defaultdict(lambda: None)
best_of_back = defaultdict(lambda: None)
reader = DOMReader(domFN)
for r in reader:
# allow missing adapter
if r.sStart > 48 or r.pStart > 48:
continue
if r.score < min_score:
continue
# ex: sid m160213_091647_42134_c100957952550000001823213806221633_s1_p0/54497/ccs_front
# ex: pid f_G11
if r.sid.endswith('_front'): # _front
bestOf = best_of_front
r.sid = r.sid[:-6]
elif r.sid.endswith('_back'): # _back
bestOf = best_of_back
r.sid = r.sid[:-5]
else:
raise ClassifierException(
"Unable to parse a read {r} in phmmer dom file {f}.".
format(r=r.sid, f=domFN))
if r.sid not in bestOf:
bestOf[r.sid] = {}
if (r.pid not in bestOf[r.sid]) or \
(bestOf[r.sid][r.pid].score < r.score):
bestOf[r.sid][r.pid] = r
return (best_of_front, best_of_back)