def _subcontigs_from_strand_bias(self, bam, ctg_name):
ctg_length = len(self.contigs[ctg_name])
fwd_cov = mapping.get_bam_region_coverage(bam, ctg_name, ctg_length)
rev_cov = mapping.get_bam_region_coverage(bam,
ctg_name,
ctg_length,
rev=True)
good_intervals = self._good_intervals_from_strand_coverage(
fwd_cov, rev_cov)
new_contigs = []
if len(good_intervals) == 1:
self.contigs[ctg_name].fa.seq = self.contigs[ctg_name].fa.seq[
good_intervals[0][0]:good_intervals[0][1] + 1]
elif len(good_intervals) > 1:
for i in range(len(good_intervals)):
start = good_intervals[i][0]
end = good_intervals[i][1]
if end - start + 1 >= 100:
new_contigs.append(
pyfastaq.sequences.Fasta(
ctg_name + '.' + str(i + 1),
self.contigs[ctg_name].fa[start:end + 1]))
return new_contigs
def _trim_contig_for_strand_bias(self, bam, ctg_name):
assert os.path.exists(bam)
if ctg_name in self.contigs_trimmed_for_strand_bias:
return
ctg_length = len(self.contigs[ctg_name])
fwd_cov = mapping.get_bam_region_coverage(bam, ctg_name, ctg_length)
rev_cov = mapping.get_bam_region_coverage(bam,
ctg_name,
ctg_length,
rev=True)
first_good_base = 0
while first_good_base < ctg_length:
total_cov = fwd_cov[first_good_base] + rev_cov[first_good_base]
if total_cov >= self.ext_min_cov and min(
fwd_cov[first_good_base],
rev_cov[first_good_base]) / total_cov >= self.strand_bias:
break
first_good_base += 1
last_good_base = ctg_length - 1
while last_good_base > first_good_base:
total_cov = fwd_cov[last_good_base] + rev_cov[last_good_base]
if total_cov >= self.ext_min_cov and min(
fwd_cov[last_good_base],
rev_cov[last_good_base]) / total_cov >= self.strand_bias:
break
last_good_base -= 1
if self.verbose >= 2:
print('Trimming strand biased ends of contig', ctg_name,
'- good base range is', first_good_base + 1, 'to',
last_good_base + 1, 'from', ctg_length, 'bases')
self.contigs[ctg_name].fa.seq = self.contigs[ctg_name].fa.seq[
first_good_base:last_good_base + 1]
def test_get_bam_region_coverage_fwd_And_rev(self):
'''Test get_bam_region_coverage both strands'''
bam = os.path.join(data_dir, 'mapping_test.smalt.out.sorted.bam')
cov = mapping.get_bam_region_coverage(bam, 'ref', 190, verbose=3, both_strands=True)
f = open(os.path.join(data_dir, 'mapping_test.smalt.out.sorted.bam.fwd_and_rev.cov'), 'rb')
expected = pickle.load(f)
f.close()
self.assertListEqual(cov, expected)
def _subcontigs_from_strand_bias(self, bam, ctg_name):
ctg_length = len(self.contigs[ctg_name])
fwd_cov = mapping.get_bam_region_coverage(bam, ctg_name, ctg_length)
rev_cov = mapping.get_bam_region_coverage(bam, ctg_name, ctg_length, rev=True)
good_intervals = self._good_intervals_from_strand_coverage(fwd_cov, rev_cov)
new_contigs = []
if len(good_intervals) == 1:
self.contigs[ctg_name].fa.seq = self.contigs[ctg_name].fa.seq[good_intervals[0][0]:good_intervals[0][1]+1]
elif len(good_intervals) > 1:
for i in range(len(good_intervals)):
start = good_intervals[i][0]
end = good_intervals[i][1]
if end - start + 1 >= 100:
new_contigs.append(pyfastaq.sequences.Fasta(ctg_name + '.' + str(i+1), self.contigs[ctg_name].fa[start:end+1]))
return new_contigs
def _trim_contig_for_strand_bias(self, bam, ctg_name):
assert os.path.exists(bam)
if ctg_name in self.contigs_trimmed_for_strand_bias:
return
ctg_length = len(self.contigs[ctg_name])
fwd_cov = mapping.get_bam_region_coverage(bam, ctg_name, ctg_length)
rev_cov = mapping.get_bam_region_coverage(bam, ctg_name, ctg_length, rev=True)
first_good_base = 0
while first_good_base < ctg_length:
total_cov = fwd_cov[first_good_base] + rev_cov[first_good_base]
if total_cov >= self.ext_min_cov and min(fwd_cov[first_good_base], rev_cov[first_good_base]) / total_cov >= self.strand_bias:
break
first_good_base += 1
last_good_base = ctg_length - 1
while last_good_base > first_good_base:
total_cov = fwd_cov[last_good_base] + rev_cov[last_good_base]
if total_cov >= self.ext_min_cov and min(fwd_cov[last_good_base], rev_cov[last_good_base]) / total_cov >= self.strand_bias:
break
last_good_base -= 1
if self.verbose >= 2:
print('Trimming strand biased ends of contig', ctg_name, '- good base range is', first_good_base + 1, 'to', last_good_base + 1, 'from', ctg_length, 'bases')
self.contigs[ctg_name].fa.seq = self.contigs[ctg_name].fa.seq[first_good_base:last_good_base+1]
def _trim_ends(fasta_in,
fasta_out,
to_trim,
min_length=100,
min_dist_to_end=25,
window_length=10,
min_pc=90):
'''Trim sequences off contig ends.'''
tmpdir = tempfile.mkdtemp(prefix='tmp.adapter_trim.', dir=os.getcwd())
tmp_prefix = os.path.join(tmpdir, 'out')
sorted_bam = tmp_prefix + '.bam'
mapping.map_reads(to_trim,
None,
fasta_in,
tmp_prefix,
index_k=9,
index_s=1,
threads=1,
minid=0.75,
sort=True,
extra_smalt_map_ops='-d -1 -m 10')
f_out = pyfastaq.utils.open_file_write(fasta_out)
seq_reader = pyfastaq.sequences.file_reader(fasta_in)
for seq in seq_reader:
coverage = mapping.get_bam_region_coverage(sorted_bam,
seq.id,
len(seq),
both_strands=True)
good_coords = _coverage_to_trimmed_coords(
coverage,
min_dist_to_end=min_dist_to_end,
window_length=window_length,
min_pc=min_pc)
if good_coords is None:
continue
seq.seq = seq.seq[good_coords[0]:good_coords[1] + 1]
if len(seq) >= min_length:
print(seq, file=f_out)
pyfastaq.utils.close(f_out)
shutil.rmtree(tmpdir)
def _trim_ends(fasta_in, fasta_out, to_trim, min_length=100, min_dist_to_end=25, window_length=10, min_pc=90):
'''Trim sequences off contig ends.'''
tmpdir = tempfile.mkdtemp(prefix='tmp.adapter_trim.', dir=os.getcwd())
tmp_prefix = os.path.join(tmpdir, 'out')
sorted_bam = tmp_prefix + '.bam'
mapping.map_reads(to_trim, None, fasta_in, tmp_prefix, index_k=9, index_s=1, threads=1, minid=0.75, sort=True, extra_smalt_map_ops='-d -1 -m 10')
f_out = pyfastaq.utils.open_file_write(fasta_out)
seq_reader = pyfastaq.sequences.file_reader(fasta_in)
for seq in seq_reader:
coverage = mapping.get_bam_region_coverage(sorted_bam, seq.id, len(seq), both_strands=True)
good_coords = _coverage_to_trimmed_coords(coverage, min_dist_to_end=min_dist_to_end, window_length=window_length, min_pc=min_pc)
if good_coords is None:
continue
seq.seq = seq.seq[good_coords[0]:good_coords[1]+1]
if len(seq) >= min_length:
print(seq, file=f_out)
pyfastaq.utils.close(f_out)
shutil.rmtree(tmpdir)
