def test_minimal_VCF_definition_io(self):
buf = io.StringIO()
with open(pkg_file('genomvar.test', 'data/example1.vcf'), 'rt') as fh:
for line in fh:
if line.startswith('##fileformat') \
or line.startswith('#CHROM') \
or not line.startswith('#'):
buf.write(line)
buf.seek(0)
reader = VCFReader(buf)
outbuf = io.StringIO()
writer = VCFWriter(format_spec=[RESERVED_FORMAT.GT],
samples=reader.samples)
variants1 = []
for vrt in reader.iter_vrt(parse_samples=True):
self.assertTrue(
isinstance(vrt.attrib['samples']['SAMP1']['GT'], str))
if vrt.attrib['samples']['SAMP1'].get('GT') == '0/1':
vrt.attrib['samples']['SAMP1']['GT'] = (0, 1)
else:
vrt.attrib['samples']['SAMP1']['GT'] = None
outbuf.write(str(writer.get_row(vrt)))
variants1.append(vrt)
variants1.sort(key=lambda v: v.start)
outbuf.seek(0)
variants2 = list(VCFReader(outbuf).iter_vrt())
variants2.sort(key=lambda v: v.start)
for v1, v2 in zip(variants1, variants2):
self.assertTrue(v1.edit_equal(v2))
def test_from_variants_to_vcf_with_sampdata(self):
file = pkg_file('genomvar.test', 'data/example3.vcf')
variants1 = sorted(VCFReader(file).iter_vrt(parse_samples=True),
key=lambda v: v.key)
vs = VariantSet.from_variants(variants1)
tf = tempfile.NamedTemporaryFile(suffix='.vcf')
with open(tf.name, 'wt') as fh:
vs.to_vcf(
fh,
format_spec=[RESERVED_FORMAT.GT, ('AD', 'R', 'Integer', '')],
samples=['SAMP1'])
with open(tf.name, 'rt') as fh:
fh.seek(0)
self.assertIn(
'##FORMAT=<ID=AD,Number=R,Type=Integer,'\
+'Description="">',
fh.read().splitlines())
variants2 = sorted(VCFReader(tf.name).iter_vrt(parse_samples=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['samples']['SAMP1']['AD'],
v2.attrib['samples']['SAMP1']['AD'])
def test_iter_vrt_gzipped(self):
reader = VCFReader(pkg_file('genomvar.test', 'data/example2.vcf.gz'),
index=True)
self.assertEqual(list(reader.chroms), ['chr23', 'chr24'])
for vrt in reader.iter_vrt():
self.assertIn(vrt.chrom, ['chr24', 'chr23'])
self.assertEqual(len(list(reader.find_vrt(chrom='chr24'))), 4)
self.assertEqual(len(list(reader.find_vrt('chr23', 7464, 7465))), 3)
def test_example3(self):
reader = VCFReader(pkg_file('genomvar.test', 'data/example3.vcf'))
self.assertEqual(list(reader.get_chroms(allow_no_index=True)),
['chr1', 'chr2', 'chr10'])
vrt = list(reader.iter_vrt(parse_info=True, parse_samples=True))
self.assertGreater(len(vrt), 0)
v = vrt[3]
self.assertEqual(v.attrib['id'], None)
def test_sv_types(self):
reader = VCFReader(pkg_file('genomvar.test', 'data/example4.vcf.gz'))
with warnings.catch_warnings(record=True) as wrn:
# warnings.simplefilter(append=True)
for cnt, vrt in enumerate(reader.iter_vrt()):
pass
self.assertEqual(cnt, 99)
self.assertGreater(len(wrn), 1)
self.assertIn('Structural', str(wrn[-1].message))
def test_check_getting_vrt_is_sorted(self):
reader = VCFReader(pkg_file('genomvar.test',
'data/example_gnomad_2.vcf.gz'),
index=True)
starts = [v.start for v in reader.iter_vrt()]
self.assertEqual(starts, sorted(starts))
starts2 = [
v.start for v in reader.find_vrt('chr15', 74719587, 74824401)
]
self.assertEqual(starts2, sorted(starts2))
def test_from_variants_with_attributes(self):
reader = VCFReader(pkg_file('genomvar.test', 'data/example1.vcf'))
vset = VariantSet.from_variants(list(reader.iter_vrt(parse_info=True)))
vrt = list(vset.find_vrt('chr24', 1200, 1210))
self.assertEqual(len(vrt), 2)
v1 = vrt[0]
self.assertEqual(v1.attrib['info']['NSV'], 1)
self.assertEqual(v1.attrib['id'], '5')
v2 = vrt[1]
self.assertEqual(v2.attrib['id'], None)
recs = vset.to_records()
self.assertEqual(recs[0]['attrib']['info']['NSV'], 2)
def test_zip_variants(self):
fl = pkg_file('genomvar.test', 'data/example3.vcf')
with open(fl, 'rt') as fh1:
with open(fl) as fh2:
matches = list(
zip_variants(
VCFReader(fh1).iter_vrt(),
VCFReader(fh2).iter_vrt()))
d1, v1, ovlp1 = matches[0]
self.assertEqual(d1, 0)
self.assertEqual(v1.alt, '')
self.assertEqual([v.ref for v in ovlp1], ['G', ''])
dlast, vlast, ovlplast = matches[-1]
self.assertEqual(vlast.alt, 'CC')
self.assertEqual([v.ref for v in ovlplast], ['TG'])
def test_from_variants_vcf(self):
vs0 = varset.VariantSet.from_vcf(pkg_file('genomvar.test',
'data/example1.vcf'),
parse_info=True)
variants1 = sorted(vs0.iter_vrt(), key=lambda v: v.key)
vs = VariantSet.from_variants(variants1)
_desc = 'Test for multinumber field'
info_spec_tuples = [('DP4', 4, 'Integer', _desc),
('NSV', 1, 'Integer')]
info_spec_dict = vs0.dtype['info']
for info_spec in (info_spec_tuples, info_spec_dict):
tf = tempfile.NamedTemporaryFile(suffix='.vcf')
with open(tf.name, 'wt') as fh:
vs.to_vcf(fh, info_spec=info_spec)
with open(tf.name, 'rt') as fh:
self.assertIn(
'##INFO=<ID=DP4,Number=4,Type=Integer,Description="{}">'\
.format(_desc),
fh.read().splitlines())
fh.seek(0)
# print(fh.read())
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_init(self):
reader = VCFReader(pkg_file('genomvar.test', 'data/example1.vcf'))
self.assertEqual(reader.header_len, 15)
dtype = reader._dtype
self.assertEqual(len(dtype['format']), 1)
self.assertTrue(issubclass(dtype['format']['GT']['dtype'], np.object_),
msg='Got type' + str(dtype['format']['GT']['type']))
def test_iter_chrom_rows(self):
reader = VCFReader(pkg_file('genomvar.test', 'data/example2.vcf.gz'))
chroms = set()
rows = {}
for chrom, it in reader.iter_rows_by_chrom():
chroms.add(chrom)
l = list(it)
rows[chrom] = len(l)
self.assertEqual(chroms, {'chr24', 'chr23'})
self.assertEqual(rows['chr23'], 5)
self.assertEqual(rows['chr24'], 4)
# same but without reading the chroms
chroms = set()
for chrom, it in reader.iter_rows_by_chrom():
chroms.add(chrom)
self.assertEqual(chroms, {'chr24', 'chr23'})
def test_from_vcf_missing_values(self):
buf = io.StringIO()
format_fields = (RESERVED_FORMAT.AD, RESERVED_FORMAT.DP,
RESERVED_FORMAT.GT)
header = vcf_header.render(
samples=['S1','S2'],
ctg_len={},
format=[{k:getattr(spec, k.upper()) for k in \
('name', 'number', 'type', 'description')} \
for spec in format_fields])
buf.write(header)
buf.write('chr15\t17017413\t.\tA\tG\t38\t.\t.\tGT\t./.\t0/1\n')
buf.write('chr15\t17017413\t.\tA\tG\t38\t.\t.\tGT:AD\t./.\t0/1:10,.\n')
buf.write('chr15\t17017413\t.\tA\tG\t38\t.\t.\tGT:DP\t./.\t1/1:.\n')
buf.seek(0)
buf.seek(0)
vs = VCFReader(buf)
v = list(vs.iter_vrt(parse_samples=True))[0]
self.assertEqual(v.attrib['samples']['S1']['GT'], (None, None))
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_iter_vrt_example1(self):
reader = VCFReader(pkg_file('genomvar.test', 'data/example1.vcf'))
self.assertEqual(reader.samples, ['SAMP1'])
vrts = list(reader.iter_vrt(parse_info=True, parse_samples=True))
vrt1, vrt2 = vrts[:2]
# Test Ref and Alt
self.assertEqual([vrt1.start, vrt1.ref, vrt1.alt], [23, 'G', ''])
self.assertEqual([vrt2.start, vrt2.ref, vrt2.alt], [24, '', 'G'])
# Check row numbers
self.assertEqual(vrt1.attrib['vcf_notation']['row'], 0)
self.assertEqual(vrt2.attrib['vcf_notation']['row'], 0)
self.assertEqual(vrt1.attrib['allele_num'], 0)
self.assertEqual(vrt2.attrib['allele_num'], 1)
# Test INFO
self.assertEqual(vrt1.attrib['info']['AF'], 0.5)
self.assertEqual(vrt2.attrib['info']['AF'], 0.5)
# Test SAMPLES fields
self.assertEqual(vrt1.attrib['samples']['SAMP1']['GT'], (0, 1, 0))
self.assertEqual(vrt2.attrib['samples']['SAMP1']['GT'], (0, 0, 1))
def test_from_variants_to_vcf_with_info(self):
variants1 = sorted(VCFReader(
pkg_file('genomvar.test',
'data/example1.vcf')).iter_vrt(parse_info=True),
key=lambda v: v.key)
vs = VariantSet.from_variants(variants1)
tf = tempfile.NamedTemporaryFile(suffix='.vcf')
# Test Invalid specs
invalid_specs = [('NSV', ), ('NSV', 1, 'Integedr'),
('NSV', 'C', 'Integer', 'Number of Simple Variants')]
_buf = io.StringIO()
for spec in invalid_specs:
with self.assertRaises(ValueError) as cm:
vs.to_vcf(_buf, info_spec=[spec])
exc = cm.exception
self.assertTrue('INFO spec' in exc.args[0])
with open(tf.name, 'wt') as fh:
vs.to_vcf(fh,
info_spec=[
('NSV', 1, 'Integer', 'Number of Simple Variants'),
('AF', 'A', 'Float', '', 'source', 'version')
])
with open(tf.name, 'rt') as fh:
self.assertIn(
'##INFO=<ID=NSV,Number=1,Type=Integer,'\
+'Description="Number of Simple Variants">',
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_change_of_attributes(self):
reader = VCFReader(
pkg_file('genomvar.test','data/example1.vcf'))
vrt = list(reader.iter_vrt())[0]
self.assertEqual(str(self.writer.get_row(vrt)),
'chr24\t23\t1\tAG\tA\t100\tPASS\t.')
vrt2 = copy.deepcopy(vrt)
vrt2.attrib['id'] = '.'
vrt2.attrib['qual'] = '.'
vrt2.attrib['filter'] = '.'
self.assertEqual(str(self.writer.get_row(vrt2)),
'chr24\t23\t.\tAG\tA\t.\t.\t.')
self.assertEqual(str(self.writer.get_row(vrt2, id='.', qual='.', filter='.')),
'chr24\t23\t.\tAG\tA\t.\t.\t.')
vrt3 = copy.deepcopy(vrt)
vrt3.attrib['id'] = None
vrt3.attrib['qual'] = None
vrt3.attrib['filter'] = None
self.assertEqual(str(self.writer.get_row(vrt3)),
'chr24\t23\t.\tAG\tA\t.\t.\t.')
reader.close()
reader.close()
def test_to_vcf_row_from_file(self):
def _split_multiallelic(rows):
for row in rows:
for alt in row.ALT.split(','):
kwds = {f:getattr(row,f) for f in VCF_FIELDS}
kwds['ALT'] = alt
kwds['INFO'] = '.'
kwds['FORMAT'] = None
kwds['SAMPLES'] = None
yield str(VCFRow(**kwds))
reader = VCFReader(pkg_file('genomvar.test','data/example1.vcf'))
variants = list(reader.iter_vrt(
parse_info=False,parse_samples=False))
rows = [str(self.writer.get_row(v)) for v in variants]
for r1, r2 in zip(
_split_multiallelic(reader.iter_rows()), rows):
if 'AG\tAGG' in r1: # stripping
continue
self.assertEqual(r1,r2)
reader.close()