def get_fancy_results_dict(self,
max_per_query=10,
defline_white_space_mask=None):
b6 = b6lib.B6Source(self.output)
input_fasta = u.SequenceSource(self.input)
target_db = u.SequenceSource(self.target)
query_counts = {}
fancy_results_dict = {}
while b6.next():
if not query_counts.has_key(b6.entry.query_id):
query_counts[b6.entry.query_id] = 1
if query_counts[b6.entry.query_id] - 1 == max_per_query:
continue
else:
query_counts[b6.entry.query_id] += 1
if not fancy_results_dict.has_key(b6.entry.query_id):
fancy_results_dict[b6.entry.query_id] = []
query_seq = input_fasta.get_seq_by_read_id(
b6.entry.query_id).replace('-', '')
target_seq = target_db.get_seq_by_read_id(b6.entry.subject_id)
if defline_white_space_mask:
b6.entry = remove_white_space_mask_from_B6_entry(
b6.entry, defline_white_space_mask)
# parts that were aligned during the search are being aligned to each other to generate
# hsp_match data to include into results
query_aligned, target_aligned = nw_align(query_seq[int(b6.entry.q_start) - 1:int(b6.entry.q_end)],\
target_seq[int(b6.entry.s_start) - 1:int(b6.entry.s_end)])
query_aligned, target_aligned = query_aligned.upper(
), target_aligned.upper()
coverage = (b6.entry.q_end -
(b6.entry.q_start - 1)) * 100.0 / b6.entry.q_len
hsp_match = ''.join([
'|' if query_aligned[i] == target_aligned[i] else ' '
for i in range(0, len(query_aligned))
])
entry = copy.deepcopy(b6.entry)
entry.coverage = coverage
entry.hsp_query = query_aligned
entry.hsp_subject = target_aligned
entry.hsp_match = hsp_match
entry = remove_white_space_mask_from_B6_entry(entry)
fancy_results_dict[entry.query_id].append(entry)
return fancy_results_dict
def homopolymer_indel_exists(seq1, seq2):
seq1, seq2 = trim_uninformative_gaps_from_sequences(seq1, seq2)
# sometimes alignments look like this:
#
# CCCGAAAAAA--TAT
# CCCGAAA---AATAT
#
# where the correct alignment should look like this:
#
# CCCGAAAAAATAT
# CCCGAAAAA-TAT
#
# causes this function to return false. in order to fix that problem
# we perform needleman-wunch alignment here:
if sum([seq1.count('-'), seq2.count('-')]) > 1:
seq1, seq2 = nw_align(seq1.replace('-', ''), seq2.replace('-', ''))
gap_index = seq1.find('-')
if gap_index == -1:
gap_index = seq2.find('-')
# so the gap is in seq2. replace seq1 and 2 so it would be certain
# that the sequence with gap is seq1:
seq1, seq2 = seq2, seq1
if gap_index == -1:
return False
isHP = lambda x: len(set(x)) == 1
isHPindel = lambda (s, e): seq1[s:e] == seq2[s:e] and isHP(seq1[
s:e]) == 1 and seq2[gap_index] == seq2[s]
def DownStream(sequence):
i = 3
while isHP(sequence[gap_index - i - 1:gap_index]):
i += 1
return (gap_index - i, gap_index)
def UpStream(sequence):
i = 4
while isHP(sequence[gap_index + 1:gap_index + i + 1]):
i += 1
return (gap_index + 1, gap_index + i)
# check downstream of the gap
if gap_index >= 3:
if isHPindel(DownStream(seq1)):
return True
# check upstream of the gap
if len(seq1) - gap_index > 3:
if isHPindel(UpStream(seq1)):
return True
return None
def homopolymer_indel_exists(seq1, seq2):
seq1, seq2 = trim_uninformative_gaps_from_sequences(seq1, seq2)
# sometimes alignments look like this:
#
# CCCGAAAAAA--TAT
# CCCGAAA---AATAT
#
# where the correct alignment should look like this:
#
# CCCGAAAAAATAT
# CCCGAAAAA-TAT
#
# causes this function to return false. in order to fix that problem
# we perform needleman-wunch alignment here:
if sum([seq1.count('-'), seq2.count('-')]) > 1:
seq1, seq2 = nw_align(seq1.replace('-', ''), seq2.replace('-', ''))
gap_index = seq1.find('-')
if gap_index == -1:
gap_index = seq2.find('-')
# so the gap is in seq2. replace seq1 and 2 so it would be certain
# that the sequence with gap is seq1:
seq1, seq2 = seq2, seq1
if gap_index == -1:
return False
isHP = lambda x: len(set(x)) == 1
isHPindel = lambda (s, e): seq1[s:e] == seq2[s:e] and isHP(seq1[s:e]) == 1 and seq2[gap_index] == seq2[s]
def DownStream(sequence):
i = 3
while isHP(sequence[gap_index - i - 1:gap_index]):
i += 1
return (gap_index - i, gap_index)
def UpStream(sequence):
i = 4
while isHP(sequence[gap_index + 1:gap_index + i + 1]):
i += 1
return (gap_index + 1, gap_index + i)
# check downstream of the gap
if gap_index >= 3:
if isHPindel(DownStream(seq1)):
return True
# check upstream of the gap
if len(seq1) - gap_index > 3:
if isHPindel(UpStream(seq1)):
return True
return None
def get_fancy_results_dict(self, max_per_query = 10, defline_white_space_mask = None):
b6 = b6lib.B6Source(self.output)
input_fasta = u.SequenceSource(self.input)
target_db = u.SequenceSource(self.target)
query_counts = {}
fancy_results_dict = {}
while b6.next():
if not query_counts.has_key(b6.entry.query_id):
query_counts[b6.entry.query_id] = 1
if query_counts[b6.entry.query_id] - 1 == max_per_query:
continue
else:
query_counts[b6.entry.query_id] += 1
if not fancy_results_dict.has_key(b6.entry.query_id):
fancy_results_dict[b6.entry.query_id] = []
query_seq = input_fasta.get_seq_by_read_id(b6.entry.query_id).replace('-', '')
target_seq = target_db.get_seq_by_read_id(b6.entry.subject_id)
if defline_white_space_mask:
b6.entry = remove_white_space_mask_from_B6_entry(b6.entry, defline_white_space_mask)
# parts that were aligned during the search are being aligned to each other to generate
# hsp_match data to include into results
query_aligned, target_aligned = nw_align(query_seq[int(b6.entry.q_start) - 1:int(b6.entry.q_end)],\
target_seq[int(b6.entry.s_start) - 1:int(b6.entry.s_end)])
query_aligned, target_aligned = query_aligned.upper(), target_aligned.upper()
coverage = (b6.entry.q_end - (b6.entry.q_start - 1)) * 100.0 / b6.entry.q_len
hsp_match = ''.join(['|' if query_aligned[i] == target_aligned[i] else ' ' for i in range(0, len(query_aligned))])
entry = copy.deepcopy(b6.entry)
entry.coverage = coverage
entry.hsp_query = query_aligned
entry.hsp_subject = target_aligned
entry.hsp_match = hsp_match
entry = remove_white_space_mask_from_B6_entry(entry)
fancy_results_dict[entry.query_id].append(entry)
return fancy_results_dict
def compare_seqs(knownseq, testseq, header):
#keep if exact match of substrings
if knownseq in testseq or testseq in knownseq:
return ">%s\n%s" % (header, testseq)
#START: keep if one difference between each sequence
alnknown, alnseq = nw_align(knownseq, testseq)
#find where to start comparison by trimming gaps at ends
knownstart, knownend = find_ends(alnknown)
seqstart, seqend = find_ends(alnseq)
start = min(seqstart, knownstart)
end = max(seqend, knownend)
if start < 0 or end < 0:
raise ValueError("Start and end must be greater than 0!")
#if only one difference, print out sequence
if distance(knownseq[start:end], testseq[start:end]) <= 1:
return ">%s\n%s" % (header, testseq)
return ""
def compare_seqs(knownseq, testseq, header):
#keep if exact match of substrings
if knownseq in testseq or testseq in knownseq:
return ">%s\n%s" % (header, testseq)
#START: keep if one difference between each sequence
alnknown, alnseq = nw_align(knownseq, testseq)
#find where to start comparison by trimming gaps at ends
knownstart, knownend = find_ends(alnknown)
seqstart, seqend = find_ends(alnseq)
start = min(seqstart, knownstart)
end = max(seqend, knownend)
if start < 0 or end < 0:
raise ValueError("Start and end must be greater than 0!")
#if only one difference, print out sequence
if distance(knownseq[start:end], testseq[start:end]) <= 1:
return ">%s\n%s" % (header, testseq)
return ""
def test_nw_align(self):
"""Tests for nw_align function."""
(first, second), score = nw_align('ACGU', 'CAGU', return_score=True)
self.assertEqual(first, 'AC-GU')
self.assertEqual(second, '-CAGU')
self.assertEqual(score, 1)
def test_nw_align_empty(self):
"""Tests for nw_align function."""
(first, second), score = nw_align('', '', return_score=True)
self.assertEqual(first, '')
self.assertEqual(second, '')
self.assertEqual(score, 0)
def test_nw_align(self):
"""Tests for nw_align function."""
(first, second), score = nw_align("ACGU", "CAGU", return_score=True)
self.assertEqual(first, "AC-GU")
self.assertEqual(second, "-CAGU")
self.assertEqual(score, 1)
def test_nw_align_empty(self):
"""Tests for nw_align function."""
(first, second), score = nw_align("", "", return_score=True)
self.assertEqual(first, "")
self.assertEqual(second, "")
self.assertEqual(score, 0)
def test_nw_align(self):
"""Tests for nw_align function."""
(first,second),score = nw_align('ACGU','CAGU',return_score=True)
self.assertEqual(first,'AC-GU')
self.assertEqual(second,'-CAGU')
self.assertEqual(score,1)