def testSort(self):
hits = HHpredHitList([self.h2, self.h1])
self.assertEqual(hits[0], self.h2)
self.assertEqual(hits[1], self.h1)
hits.sort()
self.assertEqual(hits[0], self.h1)
self.assertEqual(hits[1], self.h2)
def testSort(self):
hits = HHpredHitList([self.h2, self.h1])
self.assertEqual(hits[0], self.h2)
self.assertEqual(hits[1], self.h1)
hits.sort()
self.assertEqual(hits[0], self.h1)
self.assertEqual(hits[1], self.h2)
def setUp(self):
super(TestHitList, self).setUp()
self.h1 = HHpredHit(1, 'hit1', 2, 5, 3, 6, 0.5, 10)
self.h2 = HHpredHit(2, 'hit2', 3, 5, 4, 6, 0.2, 10)
self.h1.add_alignment('A-CD', 'A-CD')
self.h2.add_alignment('BCD', 'BCD')
self.hits = HHpredHitList([self.h1, self.h2])
def _parse(self, stream, header_only):
qlen = None
in_hits = False
in_alis = False
has_alis = False
c_rank = 0
header = {}
hits = {}
alis = {}
for line in stream:
if not in_hits and not in_alis:
if line.replace(' ', '').startswith('NoHitProbE-value'):
in_hits = True
continue
elif line.strip() == '':
continue
else: # parse header data (stuff above the hits table)
columns = line.strip().split(None, 1)
if len(columns) == 2:
identifier, data = columns
if identifier in ('Query', 'Command'):
data = data.strip()
elif identifier == 'Neff':
data = float(data)
elif identifier in ('Searched_HMMs', 'Match_columns'):
data = int(data)
header[identifier] = data
if identifier == 'Match_columns':
qlen = data
if in_hits and not header_only:
if not line.strip(): # suboptimal way to handle block switch
in_hits = False
in_alis = True
if self.alignments:
continue
else:
break
elif line.strip() == 'Done':
in_hits = False
in_alis = False
break
description = line[:34].split()
rank = int(description[0])
id = description[1]
pos = line[85:94].strip()
start, end = map(int, pos.split('-'))
qpos = line[75:84].strip()
qstart, qend = map(int, qpos.split('-'))
probability = float(line[35:40]) / 100.0
hit = HHpredHit(rank, id, start, end, qstart, qend,
probability, qlen)
hit.evalue = float(line[41:48])
hit.pvalue = float(line[49:56])
hit.score = float(line[57:63])
hit.ss_score = float(line[64:69])
hit.slength = int(line[94:].replace('(', '').replace(')', ''))
hits[hit.rank] = hit
alis[hit.rank] = {'q': [], 's': []}
elif in_alis and not header_only:
if line.startswith('Done'):
in_alis = False
break
elif line.startswith('No '):
c_rank = int(line[3:])
if c_rank not in hits:
raise HHOutputFormatError(
'Alignment {0}. refers to a non-existing hit'.
format(c_rank))
elif line.startswith('>'):
hits[c_rank].name = line[1:].strip()
elif line.startswith('Probab='):
for pair in line.split():
key, value = pair.split('=')
if key == 'Identities':
hits[c_rank].identity = float(
value.replace('%', ''))
elif key == 'Similarity':
hits[c_rank].similarity = float(value)
elif key == 'Sum_probs':
hits[c_rank].prob_sum = float(value)
elif line.startswith('Q ') and not line[:11].rstrip() in (
'Q Consensus', 'Q ss_pred', 'Q ss_conf', 'Q ss_dssp'):
for residue in line[22:]:
if residue.isspace() or residue.isdigit():
break
else:
alis[c_rank]['q'].append(residue)
has_alis = True
elif line.startswith('T ') and not line[:11].rstrip() in (
'T Consensus', 'T ss_pred', 'T ss_conf', 'T ss_dssp'):
for residue in line[22:]:
if residue.isspace() or residue.isdigit():
break
else:
alis[c_rank]['s'].append(residue)
if self.alignments and has_alis:
for rank in alis:
try:
hits[rank].add_alignment(alis[rank]['q'], alis[rank]['s'])
except (KeyError, ValueError) as er:
raise HHOutputFormatError(
'Corrupt alignment at hit No {0}.\n {1}'.format(
rank, er))
del alis
hits = HHpredHitList(hits.values())
hits.sort()
## add data obtained from the header to the HHpredHitList
for identifier, data in header.items():
if identifier == 'Query':
hits.query_name = data
elif identifier == 'Match_columns':
hits.match_columns = data
elif identifier == 'No_of_seqs':
hits.no_of_seqs = data
elif identifier == 'Neff':
hits.neff = data
elif identifier == 'Searched_HMMs':
hits.searched_hmms = data
elif identifier == 'Date':
hits.date = data
elif identifier == 'Command':
hits.command = data
return hits
def _parse(self, stream, header_only):
qlen = None
in_hits = False
in_alis = False
has_alis = False
c_rank = 0
header = {}
hits = {}
alis = {}
for line in stream:
if not in_hits and not in_alis:
if line.replace(' ', '').startswith('NoHitProbE-value'):
in_hits = True
continue
elif line.strip() == '':
continue
else: # parse header data (stuff above the hits table)
columns = line.strip().split(None, 1)
if len(columns) == 2:
identifier, data = columns
if identifier in ('Query', 'Command'):
data = data.strip()
elif identifier == 'Neff':
data = float(data)
elif identifier in ('Searched_HMMs', 'Match_columns'):
data = int(data)
header[identifier] = data
if identifier == 'Match_columns':
qlen = data
if in_hits and not header_only:
if not line.strip(): # suboptimal way to handle block switch
in_hits = False
in_alis = True
if self.alignments:
continue
else:
break
elif line.strip() == 'Done':
in_hits = False
in_alis = False
break
description = line[:34].split()
rank = int(description[0])
id = description[1]
pos = line[85:94].strip()
start, end = map(int, pos.split('-'))
qpos = line[75:84].strip()
qstart, qend = map(int, qpos.split('-'))
probability = float(line[35:40]) / 100.0
hit = HHpredHit(rank, id, start, end, qstart, qend, probability, qlen)
hit.evalue = float(line[41:48])
hit.pvalue = float(line[49:56])
hit.score = float(line[57:63])
hit.ss_score = float(line[64:69])
hit.slength = int(line[94:].replace('(', '').replace(')', ''))
hits[hit.rank] = hit
alis[hit.rank] = {'q': [], 's': []}
elif in_alis and not header_only:
if line.startswith('Done'):
in_alis = False
break
elif line.startswith('No '):
c_rank = int(line[3:])
if c_rank not in hits:
raise HHOutputFormatError('Alignment {0}. refers to a non-existing hit'.format(c_rank))
elif line.startswith('>'):
hits[c_rank].name = line[1:].strip()
elif line.startswith('Probab='):
for pair in line.split():
key, value = pair.split('=')
if key == 'Identities':
hits[c_rank].identity = float(
value.replace('%', ''))
elif key == 'Similarity':
hits[c_rank].similarity = float(value)
elif key == 'Sum_probs':
hits[c_rank].prob_sum = float(value)
elif line.startswith('Q ') and not line[:11].rstrip() in ('Q Consensus', 'Q ss_pred','Q ss_conf', 'Q ss_dssp'):
for residue in line[22:]:
if residue.isspace() or residue.isdigit():
break
else:
alis[c_rank]['q'].append(residue)
has_alis = True
elif line.startswith('T ') and not line[:11].rstrip() in ('T Consensus', 'T ss_pred','T ss_conf', 'T ss_dssp'):
for residue in line[22:]:
if residue.isspace() or residue.isdigit():
break
else:
alis[c_rank]['s'].append(residue)
if self.alignments and has_alis:
for rank in alis:
try:
hits[rank].add_alignment(alis[rank]['q'], alis[rank]['s'])
except (KeyError, ValueError) as er:
raise HHOutputFormatError('Corrupt alignment at hit No {0}.\n {1}'.format(rank, er))
del alis
hits = HHpredHitList(hits.values())
hits.sort()
## add data obtained from the header to the HHpredHitList
for identifier, data in header.items():
if identifier == 'Query':
hits.query_name = data
elif identifier == 'Match_columns':
hits.match_columns = data
elif identifier == 'No_of_seqs':
hits.no_of_seqs = data
elif identifier == 'Neff':
hits.neff = data
elif identifier == 'Searched_HMMs':
hits.searched_hmms = data
elif identifier == 'Date':
hits.date = data
elif identifier == 'Command':
hits.command = data
return hits