def test_slice(self):
# with fasta
seq = SeqItem(name='s1', lines=['>s1\n', 'ACTGGTAC\n'])
seq = SeqWrapper(SEQITEM, seq, 'fasta')
expected_seq = SeqItem(name='s1', lines=['>s1\n', 'CTGG\n'])
expected_seq = SeqWrapper(SEQITEM, expected_seq, 'fasta')
assert slice_seq(seq, 1, 5) == expected_seq
# with fastq
seq = SeqItem(name='seq',
lines=['@seq\n', 'aata\n', '+\n', '!?!?\n'])
seq = SeqWrapper(SEQITEM, seq, 'fastq')
seq = slice_seq(seq, 1, 3)
assert list(get_int_qualities(seq)) == [30, 0]
assert get_str_seq(seq) == 'at'
assert seq.object.lines == ['@seq\n', 'at\n', '+\n', '?!\n']
# with multiline fastq
seq = SeqItem(name='seq', lines=['@seq\n', 'aaatcaaa\n', '+\n',
'@AAABBBB\n'])
seq = SeqWrapper(SEQITEM, seq, 'fastq-illumina')
seq_ = slice_seq(seq, 1, 5)
assert list(get_int_qualities(seq_)) == [1, 1, 1, 2]
assert get_str_seq(seq_) == get_str_seq(seq)[1: 5]
# It tests the stop is None
seq = SeqItem('seq', ['>seq\n', 'aCTG'])
seq = SeqWrapper(SEQITEM, seq, 'fasta')
assert get_str_seq(slice_seq(seq, 1, None)) == 'aCTG'[1:]
assert get_str_seq(slice_seq(seq, None, 1)) == 'aCTG'[:1]
def test_slice(self):
# with fasta
seq = SeqItem(name='s1', lines=['>s1\n', 'ACTGGTAC\n'])
seq = SeqWrapper(SEQITEM, seq, 'fasta')
expected_seq = SeqItem(name='s1', lines=['>s1\n', 'CTGG\n'])
expected_seq = SeqWrapper(SEQITEM, expected_seq, 'fasta')
assert slice_seq(seq, 1, 5) == expected_seq
# with fastq
seq = SeqItem(name='seq', lines=['@seq\n', 'aata\n', '+\n', '!?!?\n'])
seq = SeqWrapper(SEQITEM, seq, 'fastq')
seq = slice_seq(seq, 1, 3)
assert list(get_int_qualities(seq)) == [30, 0]
assert get_str_seq(seq) == 'at'
assert seq.object.lines == ['@seq\n', 'at\n', '+\n', '?!\n']
# with multiline fastq
seq = SeqItem(name='seq',
lines=['@seq\n', 'aaatcaaa\n', '+\n', '@AAABBBB\n'])
seq = SeqWrapper(SEQITEM, seq, 'fastq-illumina')
seq_ = slice_seq(seq, 1, 5)
assert list(get_int_qualities(seq_)) == [1, 1, 1, 2]
assert get_str_seq(seq_) == get_str_seq(seq)[1:5]
# It tests the stop is None
seq = SeqItem('seq', ['>seq\n', 'aCTG'])
seq = SeqWrapper(SEQITEM, seq, 'fasta')
assert get_str_seq(slice_seq(seq, 1, None)) == 'aCTG'[1:]
assert get_str_seq(slice_seq(seq, None, 1)) == 'aCTG'[:1]
def _do_trim(self, seq):
'It trims the edges of the given seqs.'
annots = get_annotations(seq)
if not TRIMMING_RECOMMENDATIONS in annots:
return seq
trim_rec = annots[TRIMMING_RECOMMENDATIONS]
# fixing the trimming recommendations
if TRIMMING_RECOMMENDATIONS in annots:
del annots[TRIMMING_RECOMMENDATIONS]
trim_segments = []
for trim_kind in TRIMMING_KINDS:
trim_segments.extend(trim_rec.get(trim_kind, []))
# masking
if self.mask:
seq = _mask_sequence(seq, trim_segments)
else:
# trimming
if trim_segments:
trim_limits = get_longest_complementary_segment(
trim_segments, get_length(seq))
if trim_limits is None:
# there's no sequence left
return None
else:
trim_limits = []
if trim_limits:
seq = slice_seq(seq, trim_limits[0], trim_limits[1] + 1)
return seq
def _do_trim(self, seq):
'It trims the edges of the given seqs.'
annots = get_annotations(seq)
if not TRIMMING_RECOMMENDATIONS in annots:
return seq
trim_rec = annots[TRIMMING_RECOMMENDATIONS]
# fixing the trimming recommendations
if TRIMMING_RECOMMENDATIONS in annots:
del annots[TRIMMING_RECOMMENDATIONS]
trim_segments = []
for trim_kind in TRIMMING_KINDS:
trim_segments.extend(trim_rec.get(trim_kind, []))
# masking
if self.mask:
seq = _mask_sequence(seq, trim_segments)
else:
# trimming
if trim_segments:
trim_limits = get_longest_complementary_segment(
trim_segments, get_length(seq))
if trim_limits is None:
# there's no sequence left
return None
else:
trim_limits = []
if trim_limits:
seq = slice_seq(seq, trim_limits[0], trim_limits[1] + 1)
return seq
class MatePairSplitter(object):
'It splits the input sequences with the provided linkers.'
def __init__(self, linkers=None):
'The initiator'
if linkers is None:
linkers = get_setting('LINKERS')
linkers = [
SeqItem(str(i), '>%d\n%s\n' % (i, l))
for i, l in enumerate(linkers)
]
linkers = assing_kind_to_seqs(SEQITEM, linkers, 'fasta')
self.linkers = list(linkers)
def __call__(self, seqs):
'It splits a list of sequences with the provided linkers'
seq_fhand = write_seqs(seqs, file_format='fasta')
seq_fhand.flush()
min_identity = 87.0
min_len = 13
filters = [{
'kind': 'min_length',
'min_num_residues': min_len,
'length_in_query': False,
'filter_match_parts': True
}, {
'kind': 'score_threshold',
'score_key': 'identity',
'min_score': min_identity
}]
matcher = BlasterForFewSubjects(seq_fhand.name,
self.linkers,
program='blastn',
filters=filters,
params={'task': 'blastn-short'},
elongate_for_global=True,
seqs_type=NUCL)
new_seqs = []
for seq in seqs:
segments = matcher.get_matched_segments_for_read(get_name(seq))
if segments is not None:
split_seqs = self._split_by_mate_linker(seq, segments)
else:
split_seqs = [seq]
for seq in split_seqs:
new_seqs.append(seq)
return new_seqs
def _split_by_mate_linker(self, seq, (segments, is_partial)):
'It splits the seqs using segments'
if not segments:
return [copy_seq(seq)]
elongated_match = is_partial
if len(segments) == 1:
segment_start = segments[0][0]
segment_end = segments[0][1]
seq_end = get_length(seq) - 1
if segment_start == 0:
new_seq = slice_seq(seq, segment_end + 1, None)
return [new_seq]
elif segment_end == seq_end:
new_seq = slice_seq(seq, None, segment_start)
return [new_seq]
elif segment_end > seq_end:
msg = 'The segment ends after the sequence has ended'
raise RuntimeError(msg)
else:
new_seq1 = slice_seq(seq, None, segment_start)
new_seq2 = slice_seq(seq, segment_end + 1, None)
if elongated_match:
name = get_name(seq) + '_pl'
else:
name = get_name(seq)
new_seq1 = copy_seq(new_seq1, name=name + r'\1')
new_seq2 = copy_seq(new_seq2, name=name + r'\2')
return [new_seq1, new_seq2]
else:
seqs = []
counter = 1
seq_start = 0
for segment_start, segment_end in segments:
if segment_start == 0:
continue
new_seq = slice_seq(seq, seq_start, segment_start)
seq_name = get_name(seq) + '_mlc.part{0:d}'.format(counter)
new_seq = copy_seq(new_seq, name=seq_name)
seqs.append(new_seq)
counter += 1
seq_start = segment_end + 1
else:
if segment_end != get_length(seq) + 1:
new_seq = slice_seq(seq, segment_end + 1, None)
name = get_name(seq) + '_mlc.part{0:d}'.format(counter)
new_seq = copy_seq(new_seq, name=name)
seqs.append(new_seq)
return seqs