def test_from_vcf_to_records(self):
vs = VariantSet.from_vcf(pkg_file('genomvar.test',
'data/example1.vcf'),
parse_info=True,
parse_samples=True)
self.assertEqual(vs._samples, ['SAMP1'])
# Test nested dtype
recs = vs.to_records(nested=True)
self.assertEqual(list(recs.dtype.fields), [
'chrom', 'start', 'end', 'ref', 'alt', 'vartype', 'phase_group',
'info', 'SAMPLES'
])
self.assertEqual(
list(recs['info'].dtype.fields),
['NSV', 'AF', 'DP4', 'ECNT', 'pl', 'mt', 'RECN', 'STR'])
self.assertEqual(list(recs['SAMPLES'].dtype.fields), ['SAMP1'])
self.assertEqual(list(recs['SAMPLES']['SAMP1'].dtype.fields), ['GT'])
# Test not nested
recs = vs.to_records(nested=False)
self.assertEqual(list(recs.dtype.fields), [
'chrom', 'start', 'end', 'ref', 'alt', 'vartype', 'phase_group',
'info_NSV', 'info_AF', 'info_DP4', 'info_ECNT', 'info_pl',
'info_mt', 'info_RECN', 'info_STR', 'SAMPLES_SAMP1_GT'
])
def test_asterisk_variant(self):
vset = VariantSet.from_vcf(pkg_file(
'genomvar.test', 'data/example_with_asterisk.vcf.gz'),
parse_info=True)
vrt = list(vset.find_vrt('chr1', 995507, 995515))
self.assertEqual(len(vrt), 3)
def test_from_vcf_with_attr(self):
s = VariantSet.from_vcf(pkg_file('genomvar.test', 'data/example1.vcf'),
parse_info=True)
_vrt = list(s.find_vrt('chr24', 150, 160))
self.assertEqual(len(_vrt), 1)
vrt = _vrt[0]
self.assertEqual(vrt.attrib['info']['AF'], 1.0)
# Check multiallelic locus
_vrt = list(s.find_vrt('chr24', 20, 30))
self.assertEqual(len(_vrt), 2)
for vrt in _vrt:
if not vrt.is_variant_instance(variant.Null):
self.assertEqual(vrt.attrib['info']['AF'], 0.5)
# Check None/KeyError cases (".",field absent...)
_vrt = list(
filter(lambda o: not o.is_variant_instance(variant.Null),
s.find_vrt('chr24', 450, 460)))
self.assertEqual(len(_vrt), 1)
vrt = _vrt[0]
with self.assertRaises(ValueError):
vrt.attrib['info']['Randomfields']
_vrt = list(
filter(lambda o: not o.is_variant_instance(variant.Null),
s.find_vrt('chr24', 4750, 4760)))
self.assertEqual(len(_vrt), 1)
vrt = _vrt[0]
self.assertEqual(vrt.attrib['info']['STR'], True)
def test_empty_vcf(self):
buf = io.StringIO()
with open(pkg_file('genomvar.test', 'data/example1.vcf')) as fh:
for line in itertools.takewhile(lambda l: l.startswith('#'), fh):
buf.write(line)
buf.seek(0)
vs = VariantSet.from_vcf(buf)
self.assertEqual(vs.nof_unit_vrt(), 0)
def test_sv_types(self):
with warnings.catch_warnings(record=True) as wrn:
vs = VariantSet.from_vcf(
pkg_file('genomvar.test', 'data/example4.vcf.gz'))
warnings.simplefilter('always')
self.assertEqual(vs.nof_unit_vrt(), 100)
self.assertGreater(len(wrn), 1)
self.assertIn('Structural', str(wrn[-1].message))
def test_sort_chroms(self):
vs = VariantSet.from_vcf(
pkg_file('genomvar.test', 'data/example2.vcf.gz'))
vs.sort_chroms()
self.assertEqual(list(vs.get_chroms()), ['chr23', 'chr24'])
vs.sort_chroms(key=lambda c: 1 if c == 'chr24' else 2)
self.assertEqual(list(vs.get_chroms()), ['chr24', 'chr23'])
def test_from_string_buffer(self):
buf = io.StringIO()
with open(pkg_file('genomvar.test', 'data/example1.vcf'), 'rt') as fh:
for line in fh:
buf.write(line)
buf.seek(0)
vs = VariantSet.from_vcf(buf)
self.assertEqual(len(list(vs.find_vrt('chr24', 150, 160))), 1)
self.assertEqual(len(list(vs.find_vrt('chr24', 20, 30))), 2)
def test_from_vcf_with_samples(self):
vset = VariantSet.from_vcf(pkg_file('genomvar.test',
'data/example1.vcf'),
parse_samples=True)
vrt = list(vset.find_vrt('chr24', 1200, 1210))
self.assertEqual(len(vrt), 2)
v1, v2 = vrt
self.assertEqual(v1.attrib['samples']['SAMP1']['GT'], (1, 0))
def test_from_vcf_problematic(self):
vset = VariantSet.from_vcf(pkg_file('genomvar.test',
'data/example5.vcf.gz'),
parse_info=True)
v = list(vset.find_vrt('1', 160109406, 160109409))
self.assertEqual(len(v), 1)
v = v[0]
self.assertEqual(v.attrib['info']['PH'], ('.', ))
def test_VariantSet_cmp2(self):
vcf1 = pkg_file('genomvar.test', 'data/example3.vcf')
s1 = VariantSet.from_vcf(vcf1)
s2 = VariantSet.from_variants(reversed(list(s1.iter_vrt())))
diff = s1.diff(s2)
self.assertEqual(diff.nof_unit_vrt(), 0)
comm = s1.comm(s2)
self.assertEqual(comm.nof_unit_vrt(), s1.nof_unit_vrt())
self.assertEqual(s2.comm(s1).nof_unit_vrt(), comm.nof_unit_vrt())
def test_from_vcf_with_info(self):
vset = VariantSet.from_vcf(pkg_file('genomvar.test',
'data/example1.vcf'),
parse_info=True)
vrt = list(vset.find_vrt('chr24', 1200, 1210))
self.assertEqual(len(vrt), 2)
v1, v2 = vrt
self.assertEqual(v1.alt, 'C')
self.assertEqual(v1.attrib['info']['NSV'], 1)
def test_VariantSet_cmp(self):
vcf1 = pkg_file('genomvar.test', 'data/example1.vcf')
vcf2 = pkg_file('genomvar.test', 'data/example2.vcf.gz')
s1 = VariantSet.from_vcf(vcf1, parse_info=True, parse_samples=True)
s2 = VariantSet.from_vcf(vcf2)
diff = s1.diff(s2)
self.assertEqual(diff.nof_unit_vrt(), 14)
# Now same diff but without loading in memory
N = 0
for vrt in s1.diff_vrt(s2).iter_vrt():
N += vrt.nof_unit_vrt()
self.assertEqual(N, 14)
comm = s1.comm(s2)
self.assertEqual(len(list(comm.iter_vrt())), 4)
v1, v2 = sorted(comm.iter_vrt(), key=lambda v: v.key)[:2]
self.assertEqual(v1.attrib['info']['NSV'], 1)
self.assertEqual(v1.attrib['samples']['SAMP1']['GT'], (0, 1))
self.assertEqual(s2.comm(s1).nof_unit_vrt(), comm.nof_unit_vrt())
def test_to_vcf_no_ref(self):
vs1 = VariantSet.from_variants(
[variant.Del("chr24", 23, 24),
variant.SNP("chr24", 1206, "C")])
buf = io.StringIO()
vs1.to_vcf(buf, reference=self.chr24)
buf.seek(0)
vs2 = VariantSet.from_vcf(buf)
self.assertEqual(vs1.comm(vs2).nof_unit_vrt(), 2)
def test_many_chroms_shuffled(self):
vs1 = VariantSet.from_vcf(
pkg_file('genomvar.test', 'data/example3.vcf'))
vrt = list(vs1.iter_vrt())
vrt2 = [vrt[i] for i in (0, 7, 1, 6, 2, 5, 3, 4)]
vs2 = VariantSet.from_variants(vrt2)
self.assertEqual(vs1.comm(vs2).nof_unit_vrt(), vs1.nof_unit_vrt())
self.assertEqual(vs1.diff(vs2).nof_unit_vrt(), 0)
self.assertEqual(vs2.comm(vs1).nof_unit_vrt(), vs2.nof_unit_vrt())
self.assertEqual(vs2.diff(vs1).nof_unit_vrt(), 0)
def test_find_vrt(self):
ivfs = VariantSetFromFile(pkg_file('genomvar.test',
'data/example2.vcf.gz'),
index=True)
vs = VariantSet.from_vcf(
pkg_file('genomvar.test', 'data/example2.vcf.gz'))
self.assertEqual(
sum([v.nof_unit_vrt() for v in ivfs.find_vrt('chr24')]),
sum([v.nof_unit_vrt() for v in vs.find_vrt('chr24')]))
self.assertEqual(sum([v.nof_unit_vrt() for v in ivfs.iter_vrt()]),
sum([v.nof_unit_vrt() for v in vs.iter_vrt()]))
def test_match(self):
# REF TGG TT
# 2093 2099
# vs1 CCC GG
# vs2 CG
# r1 r2,r3
vs1 = VariantSet.from_vcf(
pkg_file('genomvar.test', 'data/example1.vcf'))
vrt_CG = factory.from_edit('chr24', 2098, 'TT', 'CG')
vrt_CC = factory.from_edit('chr24', 2098, 'TT', 'CC')
self.assertEqual(len(vs1.match(vrt_CG)), 1)
self.assertEqual(len(vs1.match(vrt_CG, match_partial=False)), 0)
self.assertEqual(len(vs1.match(vrt_CC)), 0)
self.assertEqual(len(vs1.match(vrt_CC, match_partial=False)), 0)
def test_from_vcf(self):
vset = VariantSet.from_vcf(pkg_file('genomvar.test',
'data/example1.vcf'),
reference=self.chr24,
normindel=True)
vrt = list(vset.find_vrt('chr24', 1200, 1210))
self.assertEqual(len(vrt), 2)
# Test presence of null operation
vrt = list(vset.find_vrt('chr24', 20, 25))
self.assertEqual(len(vrt), 2)
for _v in vrt:
if not _v.is_variant_instance(variant.Null):
self.assertEqual(_v.attrib['id'], '1')
def test_drop_duplicates(self):
buf = io.StringIO()
with open(pkg_file('genomvar.test', 'data/example1.vcf')) as fh:
for line in fh:
if line.startswith('#'):
buf.write(line)
else:
for i in range(2):
buf.write(line)
buf.seek(0)
vs = VariantSet.from_vcf(buf, parse_info=True, parse_samples=True)
vs2, dropped = vs.drop_duplicates(return_dropped=True)
self.assertEqual(vs.nof_unit_vrt() / 2, vs2.nof_unit_vrt())
self.assertEqual(vs2.nof_unit_vrt(),
sum([v.nof_unit_vrt() for v in dropped]))
def test_from_to_vcf(self):
fl = pkg_file('genomvar.test', 'data/example1.vcf')
variants1 = sorted(VCFReader(fl).iter_vrt(parse_info=True),
key=lambda v: v.key)
vs = VariantSet.from_vcf(fl, parse_info=True)
tf = tempfile.NamedTemporaryFile(suffix='.vcf')
with open(tf.name, 'wt') as fh:
vs.to_vcf(fh)
with open(tf.name, 'rt') as fh:
fh.seek(0)
self.assertIn(
'##INFO=<ID=DP4,Number=4,Type=Integer,Description="Test for multinumber field">',
fh.read().splitlines())
variants2 = sorted(VCFReader(tf.name).iter_vrt(parse_info=True),
key=lambda v: v.key)
self.assertEqual(len(variants1), len(variants2))
cnt = 0
for v1, v2 in zip(variants1, variants2):
self.assertTrue(v1.edit_equal(v2))
self.assertEqual(v1.attrib['info']['NSV'],
v2.attrib['info']['NSV'])
def test_from_vcf(self):
vset = VariantSet.from_vcf(
pkg_file('genomvar.test', 'data/example1.vcf'))
self.assertEqual(vset.nof_unit_vrt(), 18)
self.assertEqual(list(vset.chroms), ['chr24'])
vrt = list(vset.find_vrt('chr24', 1200, 1210))
self.assertEqual(len(vrt), 2)
v1, v2 = vrt
self.assertTrue(v1.is_variant_instance(variant.SNP))
self.assertEqual(v1.attrib['vcf_notation']['ref'], 'G')
self.assertEqual(v1.attrib['vcf_notation']['start'], 1206)
self.assertEqual(v1.attrib['vcf_notation']['row'], 4)
self.assertEqual(v1.attrib['id'], '5')
self.assertEqual(v1.attrib['filter'], 'PASS')
self.assertEqual(v1.attrib['qual'], 100)
vrt = list(vset.find_vrt('chr24', 154, 156))
self.assertEqual(len(vrt), 1)
v1 = vrt[0]
self.assertTrue(v1.is_variant_instance(variant.Ins))
vrt = list(vset.find_vrt('chr24', 20, 25))
self.assertEqual(len(vrt), 2)
self.assertEqual(len(list(vset.iter_vrt())), 16)
def test_find_vrt_chrom_only(self):
s1 = VariantSet.from_vcf(pkg_file('genomvar.test',
'data/example2.vcf.gz'),
parse_info=True,
parse_samples=True)
self.assertEqual(len(list(s1.find_vrt('chr24'))), 4)
def test_VariantSet_no_common(self):
vcf1 = pkg_file('genomvar.test', 'data/example1.vcf')
vcf2 = pkg_file('genomvar.test', 'data/example_gnomad_1.vcf.gz')
s1 = VariantSet.from_vcf(vcf1)
s2 = VariantSet.from_vcf(vcf2)
self.assertEqual(s1.comm(s2).nof_unit_vrt(), 0)
def test_from_bcf(self):
vset = VariantSet.from_vcf(
pkg_file('genomvar.test', 'data/example3.bcf'))
self.assertEqual(len(list(vset.iter_vrt())), 8)
def test_random_sample(self):
vs = VariantSet.from_vcf(pkg_file('genomvar.test',
'data/example1.vcf'))
sample = vs.sample(5)
self.assertEqual(len(sample), 5)
