def mutate_genome(infile, mutations):
parser = khmer.ReadParser(infile)
for record in parser:
sequence = record.sequence
if record.name in mutations:
mutlist = sorted(mutations[record.name],
key=lambda m: m.pos,
reverse=True)
sequence = mutate_sequence(sequence, mutlist)
yield khmer.Read(name=record.name, sequence=sequence)
def test_BrokenPairedReader_lowercase_khmer_Read():
# use khmer.Read objects which should automatically have a `cleaned_seq`
# attribute
stream = [khmer.Read(name='seq1/1', sequence='acgtn'),
khmer.Read(name='seq1/2', sequence='AcGtN'),
khmer.Read(name='seq1/2', sequence='aCgTn')]
results = []
for num, is_pair, read1, read2 in broken_paired_reader(stream):
results.append((read1, read2))
a, b = results[0]
assert a.sequence == 'acgtn'
assert a.cleaned_seq == 'ACGTA'
assert b.sequence == 'AcGtN'
assert b.cleaned_seq == 'ACGTA'
c, d = results[1]
assert c.sequence == 'aCgTn'
assert c.cleaned_seq == 'ACGTA'
assert d is None
def localize(contigstream, refrfile, ksize=31, delta=25):
"""
Wrap the `kevlar localize` task as a generator.
Input is an iterable containing contigs (assembled by `kevlar assemble`)
stored as khmer or screed sequence records, the filename of the reference
genome sequence, and the desired k-size.
"""
seedmatches = KmerMatchSet(ksize)
for seqid, pos in get_exact_matches(contigstream, refrfile, ksize):
seedmatches.add(seqid, pos)
if len(seedmatches) == 0:
raise KevlarNoReferenceMatchesError()
refrstream = kevlar.open(refrfile, 'r')
for subseqid, subseq in extract_regions(refrstream,
seedmatches,
delta=delta):
yield khmer.Read(name=subseqid, sequence=subseq)
def test_clean_input_reads():
# all Read attributes are read only
stream = [khmer.Read(name='seq1/1', sequence='ACGT')]
with pytest.raises(AttributeError):
next(clean_input_reads(stream))