def test_read_evidence_variant_matching_gatk_mini_bundle_extract():
handle = Samfile(data_path("gatk_mini_bundle_extract.bam"))
loci = [
Locus.from_inclusive_coordinates("20", 10008951), # 0
Locus.from_inclusive_coordinates("20", 10009053), # 1
Locus.from_inclusive_coordinates("20", 10009053, 10009054), # 2
Locus.from_inclusive_coordinates("20", 10006822), # 3
Locus.from_inclusive_coordinates("20", 10006822, 10006823), # 4
]
evidence = PileupCollection.from_bam(handle, loci)
eq_(evidence.match_summary(Variant(loci[0], "A", "C")), [('A', 1),
('C', 4)])
eq_(
evidence.filter(drop_duplicates=True).match_summary(
Variant(loci[0], "A", "C")), [('A', 0), ('C', 3)])
eq_(evidence.match_summary(Variant(loci[1], "A", "C")), [('A', 3),
('C', 0)])
eq_(evidence.match_summary(Variant(loci[1], "A", "CC")), [('A', 3),
('CC', 0)])
eq_(evidence.match_summary(Variant(loci[1], "A", "")), [('A', 3), ('', 0)])
eq_(evidence.match_summary(Variant(loci[1], "A", "")), [('A', 3), ('', 0)])
eq_(evidence.match_summary(Variant(loci[2], "AT", "")), [('AT', 3),
('', 0)])
eq_(evidence.match_summary(Variant(loci[3], "A", "")), [('A', 2), ('', 6)])
eq_(evidence.match_summary(Variant(loci[4], "AC", "")), [('AC', 2),
('', 6)])
eq_(
evidence.match_summary(
Variant(loci[4], "AC", ""),
lambda e: e.read_attributes().mapping_quality.mean()),
[('AC', 60.0), ('', 65.0)])
def test_read_evidence_variant_matching_gatk_bundle_native_varcode_variant():
# Try native varcode Variant.
handle = Samfile(data_path("gatk_mini_bundle_extract.bam"))
locus = Locus.from_inclusive_coordinates("20", 10008951)
variant = VarcodeVariant(
locus.contig,
locus.position + 1, # inclusive not interbase
"A",
"C")
evidence = PileupCollection.from_bam(handle, [variant])
eq_(evidence.match_summary(variant), [('A', 1), ('C', 4)])
def from_bam(pysam_samfile, loci):
"""
Create a PileupCollection for a set of loci from a BAM file.
Parameters
----------
pysam_samfile : `pysam.csamfile.Samfile` instance, or filename string
to a BAM file. The BAM file must be indexed.
loci : list of Locus instances
Loci to collect pileups for.
Returns
----------
PileupCollection instance containing pileups for the specified loci.
All alignments in the BAM file are included (e.g. duplicate reads,
secondary alignments, etc.). See `PileupCollection.filter` if these
need to be removed.
"""
loci = [to_locus(obj) for obj in loci]
close_on_completion = False
if typechecks.is_string(pysam_samfile):
pysam_samfile = Samfile(pysam_samfile)
close_on_completion = True
try:
# Map from pyensembl normalized chromosome names used in Variant to
# the names used in the BAM file.
chromosome_name_map = {}
for name in pysam_samfile.references:
normalized = pyensembl.locus.normalize_chromosome(name)
chromosome_name_map[normalized] = name
result = PileupCollection({})
# Optimization: we sort variants so our BAM reads are localized.
locus_iterator = itertools.chain.from_iterable(
(Locus.from_interbase_coordinates(locus_interval.contig, pos) for pos in locus_interval.positions)
for locus_interval in sorted(loci)
)
for locus in locus_iterator:
result.pileups[locus] = Pileup(locus, [])
try:
chromosome = chromosome_name_map[locus.contig]
except KeyError:
logging.warn("No such contig in bam: %s" % locus.contig)
continue
columns = pysam_samfile.pileup(
chromosome,
locus.position,
locus.position + 1, # exclusive, 0-indexed
truncate=True,
stepper="nofilter",
)
try:
column = next(columns)
except StopIteration:
# No reads align to this locus.
continue
# Note that storing the pileups here is necessary, since the
# subsequent assertion will invalidate our column.
pileups = column.pileups
assert list(columns) == [] # column is invalid after this.
for pileup_read in pileups:
if not pileup_read.is_refskip:
element = PileupElement.from_pysam_alignment(locus, pileup_read)
result.pileups[locus].append(element)
return result
finally:
if close_on_completion:
pysam_samfile.close()
def from_bam(pysam_samfile, loci, normalized_contig_names=True):
'''
Create a PileupCollection for a set of loci from a BAM file.
Parameters
----------
pysam_samfile : `pysam.csamfile.Samfile` instance, or filename string
to a BAM file. The BAM file must be indexed.
loci : list of Locus instances
Loci to collect pileups for.
normalized_contig_names : whether the contig names have been normalized
(e.g. pyensembl removes the 'chr' prefix). Set to true to
de-normalize the names when querying the BAM file.
Returns
----------
PileupCollection instance containing pileups for the specified loci.
All alignments in the BAM file are included (e.g. duplicate reads,
secondary alignments, etc.). See `PileupCollection.filter` if these
need to be removed.
'''
loci = [to_locus(obj) for obj in loci]
close_on_completion = False
if typechecks.is_string(pysam_samfile):
pysam_samfile = Samfile(pysam_samfile)
close_on_completion = True
try:
# Map from pyensembl normalized chromosome names used in Variant to
# the names used in the BAM file.
if normalized_contig_names:
chromosome_name_map = {}
for name in pysam_samfile.references:
normalized = pyensembl.locus.normalize_chromosome(name)
chromosome_name_map[normalized] = name
chromosome_name_map[name] = name
else:
chromosome_name_map = None
result = PileupCollection({})
# Optimization: we sort variants so our BAM reads are localized.
locus_iterator = itertools.chain.from_iterable(
(Locus.from_interbase_coordinates(locus_interval.contig, pos)
for pos in locus_interval.positions)
for locus_interval in sorted(loci))
for locus in locus_iterator:
result.pileups[locus] = Pileup(locus, [])
if normalized_contig_names:
try:
chromosome = chromosome_name_map[locus.contig]
except KeyError:
logging.warn("No such contig in bam: %s" %
locus.contig)
continue
else:
chromosome = locus.contig
columns = pysam_samfile.pileup(
chromosome,
locus.position,
locus.position + 1, # exclusive, 0-indexed
truncate=True,
stepper="nofilter")
try:
column = next(columns)
except StopIteration:
# No reads align to this locus.
continue
# Note that storing the pileups here is necessary, since the
# subsequent assertion will invalidate our column.
pileups = column.pileups
assert list(columns) == [] # column is invalid after this.
for pileup_read in pileups:
if not pileup_read.is_refskip:
element = PileupElement.from_pysam_alignment(
locus, pileup_read)
result.pileups[locus].append(element)
return result
finally:
if close_on_completion:
pysam_samfile.close()
def test_read_evidence_gatk_mini_bundle_extract():
loci = [
Locus.from_inclusive_coordinates("20", 9999996, 9999996), # 0
Locus.from_inclusive_coordinates("20", 10260442), # 1
Locus.from_inclusive_coordinates("20", 10006823), # 2
Locus.from_inclusive_coordinates("20", 10006819, 10006823), # 3
Locus.from_inclusive_coordinates("20", 10006819, 10006825), # 4
Locus.from_inclusive_coordinates("20", 10006822, 10006827), # 5
Locus.from_inclusive_coordinates("20", 10007175), # 6
Locus.from_inclusive_coordinates("20", 10007174, 10007176), # 7
Locus.from_inclusive_coordinates("20", 1, 3), # 8
Locus.from_inclusive_coordinates("20", 10008796), # 9
Locus.from_inclusive_coordinates("20", 10008921), # 10
]
handle = Samfile(data_path("gatk_mini_bundle_extract.bam"))
evidence = PileupCollection.from_bam(handle, loci)
eq_(evidence.allele_summary(loci[0]), [("ACT", 9)])
eq_(
evidence.filter(drop_duplicates=True).allele_summary(loci[0]),
[("ACT", 8)])
eq_(evidence.allele_summary(loci[1]), [("T", 7)])
eq_(evidence.filter().allele_summary(loci[2]), [("", 6), ("C", 2)])
eq_(
evidence.filter(drop_duplicates=True,
min_base_quality=50).allele_summary(loci[2]), [])
eq_(
evidence.filter(drop_duplicates=True).allele_summary(loci[2]),
[("", 5), ("C", 1)])
eq_(
evidence.filter(drop_duplicates=True,
min_mapping_quality=60).allele_summary(loci[2]),
[("", 5), ("C", 1)])
eq_(
evidence.filter(drop_duplicates=True,
min_mapping_quality=61).allele_summary(loci[2]),
[("", 2)])
eq_(
evidence.filter(drop_duplicates=True,
min_mapping_quality=61).allele_summary(loci[3]),
[("A", 2)])
eq_(
evidence.filter(drop_duplicates=True,
min_mapping_quality=61).allele_summary(loci[4]),
[("AAA", 2)])
eq_(
evidence.filter(drop_duplicates=True,
min_mapping_quality=61).allele_summary(loci[5]),
[("AAAC", 2)])
eq_(evidence.filter().allele_summary(loci[6]), [("T", 5), ("C", 3)])
eq_(
evidence.filter(min_base_quality=30).allele_summary(loci[6]),
[("T", 4), ("C", 3)])
eq_(evidence.filter().allele_summary(loci[7]), [("CTT", 5), ("CCT", 3)])
eq_(
evidence.filter(min_base_quality=30).allele_summary(loci[7]),
[("CTT", 3), ("CCT", 2)])
eq_(
evidence.filter(min_base_quality=32).allele_summary(loci[2]),
[("", 6), ("C", 1)])
eq_(filtered_read_names(evidence.at(loci[2]).filter(min_base_quality=32)),
{'20GAVAAXX100126:4:3:18352:43857'})
eq_(evidence.allele_summary(loci[8]), [])
eq_(evidence.filter(drop_duplicates=True).allele_summary(loci[8]), [])
assert_raises(KeyError, evidence.allele_summary,
Locus.from_inclusive_coordinates("20", 10009174, 10009176))
eq_(
filtered_read_names(
evidence.at(loci[9]).filter(drop_improper_mate_pairs=True)),
{'20FUKAAXX100202:8:68:1530:49310'})
eq_(len(evidence.at(loci[8]).read_attribute('mapping_quality')), 0)
eq_(list(evidence.at(loci[9]).read_attribute('mapping_quality')),
list(evidence.at(loci[9]).read_attributes().mapping_quality))
eq_(
evidence.filter(drop_duplicates=True).allele_summary(loci[10]),
[('C', 2), ('CA', 1), ('CAA', 1)])
eq_(
evidence.filter(drop_duplicates=True).allele_summary(
Locus.from_interbase_coordinates(loci[10].contig, loci[10].start,
loci[10].start)), [('', 2),
('A', 1),
('AA', 1)])