def test_load_burden(self):
t = []
load_burden(B, t)
tl = [('CDS1', [
'FM211187:3910-3951',
]), ('CDS2', [
'FM211187:4006-4057',
])]
self.assertEqual(t, tl)
t = []
load_burden(BM, t)
tl = [('CDS1', [
'FM211187:3910-3951',
]), ('CDS2', [
'FM211187:4006-4057',
]), ('CDS3', [
'FM211187:3910-3951',
'FM211187:4006-4057',
])]
self.assertEqual(t, tl)
# no file found
with self.assertRaises(FileNotFoundError):
load_burden('nope', [])
# not the right file
with self.assertRaises(ValueError):
load_burden(P, [])
def test_read_variant_vcf(self):
p = pd.read_csv(P, index_col=0, sep='\t')['binary']
infile = VariantFile(VCF)
t = read_variant(infile, p, 'vcf', False, [], False, p.index, [])
eof, k, var_name, kstrains, nkstrains, af, missing = t
self.assertEqual(eof, False)
self.assertEqual(abs((k - np.zeros(50)).max()), 0.0)
self.assertEqual(var_name, 'FM211187_16_G_A')
self.assertEqual(kstrains, [])
self.assertEqual(nkstrains,
sorted(['sample_%d' % x for x in range(1, 51)]))
self.assertEqual(af, 0.0)
self.assertEqual(missing, 0.0)
# not providing samples
t = read_variant(infile, p, 'vcf', False, [], False, set(), [])
eof, k, var_name, kstrains, nkstrains, af, missing = t
self.assertEqual(eof, False)
self.assertEqual(k, None)
self.assertEqual(var_name, None)
self.assertEqual(kstrains, None)
self.assertEqual(nkstrains, None)
self.assertEqual(af, None)
# providing burden
burden_regions = deque([])
load_burden(B, burden_regions)
t = read_variant(infile, p.head(5), 'vcf', True, burden_regions, False,
p.head(5).index, [])
eof, k, var_name, kstrains, nkstrains, af, missing = t
self.assertEqual(eof, False)
self.assertTrue(abs((k - np.array([0, 0, 0, 0, 0])).max()) < 1E-7)
self.assertEqual(var_name, 'CDS1')
self.assertEqual(kstrains, [])
self.assertEqual(
nkstrains,
['sample_1', 'sample_2', 'sample_3', 'sample_4', 'sample_5'])
self.assertEqual(af, 0.0)
self.assertEqual(missing, 0)
# uncompressed option - no effect
t = read_variant(infile, p.head(5), 'vcf', False, [], True,
p.head(5).index, [])
eof, k, var_name, kstrains, nkstrains, af, missing = t
self.assertEqual(eof, False)
self.assertTrue(abs((k - np.array([0, 1, 0, 0, 0])).max()) < 1E-7)
self.assertEqual(var_name, 'FM211187_3982_C_A')
self.assertEqual(kstrains, ['sample_2'])
self.assertEqual(nkstrains,
['sample_1', 'sample_3', 'sample_4', 'sample_5'])
self.assertEqual(af, 0.2)
self.assertEqual(missing, 0.0)
# different type
with self.assertRaises(AttributeError):
t = read_variant(infile, p.head(5), 'kmers', False, [], True,
p.head(5).index, [])
with self.assertRaises(AttributeError):
t = read_variant(infile, p.head(5), 'Rtab', False, [], False,
p.head(5).index, [])
# read until exhaustion
while not t[0]:
t = read_variant(infile, p, 'vcf', False, [], False, p.index, [])
eof, k, var_name, kstrains, nkstrains, af, missing = t
self.assertEqual(eof, True)
self.assertEqual(k, None)
self.assertEqual(var_name, None)
self.assertEqual(kstrains, None)
self.assertEqual(nkstrains, None)
self.assertEqual(af, None)
self.assertEqual(missing, None)
# different file
infile = gzip.open(KMER)
with self.assertRaises(AttributeError):
t = read_variant(infile, p.head(5), 'vcf', False, [], False,
p.head(5).index, [])
infile = open(PRES)
with self.assertRaises(AttributeError):
t = read_variant(infile, p.head(5), 'vcf', False, [], False,
p.head(5).index, [])
def test_read_variant_vcf(self):
p = pd.read_csv(P,
index_col=0,
sep='\t')['binary']
infile = VariantFile(VCF)
t = read_variant(infile, p, 'vcf',
False, [], False,
p.index, [])
eof, k, var_name, kstrains, nkstrains, af, missing = t
self.assertEqual(eof, False)
self.assertEqual(abs((k - np.zeros(50)).max()), 0.0)
self.assertEqual(var_name,
'FM211187_16_G_A')
self.assertEqual(kstrains,
[])
self.assertEqual(nkstrains,
sorted(['sample_%d' % x
for x in range(1, 51)]))
self.assertEqual(af, 0.0)
self.assertEqual(missing, 0.0)
# not providing samples
t = read_variant(infile, p, 'vcf',
False, [], False,
set(), [])
eof, k, var_name, kstrains, nkstrains, af, missing = t
self.assertEqual(eof, False)
self.assertEqual(k, None)
self.assertEqual(var_name, None)
self.assertEqual(kstrains, None)
self.assertEqual(nkstrains, None)
self.assertEqual(af, None)
# providing burden
burden_regions = deque([])
load_burden(B, burden_regions)
t = read_variant(infile, p.head(5), 'vcf',
True, burden_regions, False,
p.head(5).index, [])
eof, k, var_name, kstrains, nkstrains, af, missing = t
self.assertEqual(eof, False)
self.assertTrue(abs((k -
np.array([0, 0, 0, 0, 0])).max()) < 1E-7)
self.assertEqual(var_name,
'CDS1')
self.assertEqual(kstrains,
[])
self.assertEqual(nkstrains,
['sample_1', 'sample_2', 'sample_3',
'sample_4', 'sample_5'])
self.assertEqual(af, 0.0)
self.assertEqual(missing, 0)
# providing burden
burden_regions = deque([])
load_burden(BM, burden_regions)
# last one has multiple regions
burden_regions.reverse()
t = read_variant(infile, p.head(5), 'vcf',
True, burden_regions, False,
p.head(5).index, [])
eof, k, var_name, kstrains, nkstrains, af, missing = t
self.assertEqual(eof, False)
self.assertTrue(abs((k -
np.array([0, 0, 0, 0, 0])).max()) < 1E-7)
self.assertEqual(var_name,
'CDS3')
self.assertEqual(kstrains,
[])
self.assertEqual(nkstrains,
['sample_1', 'sample_2', 'sample_3',
'sample_4', 'sample_5'])
self.assertEqual(af, 0.0)
self.assertEqual(missing, 0)
# uncompressed option - no effect
infile = VariantFile(VCF)
t = read_variant(infile, p.head(5), 'vcf',
False, [], True,
p.head(5).index, [])
eof, k, var_name, kstrains, nkstrains, af, missing = t
self.assertEqual(eof, False)
self.assertTrue(abs((k -
np.array([0, 0, 0, 0, 0])).max()) < 1E-7)
self.assertEqual(var_name,
'FM211187_16_G_A')
self.assertEqual(kstrains,
[])
self.assertEqual(nkstrains,
['sample_1', 'sample_2', 'sample_3',
'sample_4', 'sample_5'])
self.assertEqual(af, 0.0)
self.assertEqual(missing, 0.0)
# different type
with self.assertRaises(AttributeError):
t = read_variant(infile, p.head(5), 'kmers',
False, [], True,
p.head(5).index, [])
with self.assertRaises(AttributeError):
t = read_variant(infile, p.head(5), 'Rtab',
False, [], False,
p.head(5).index, [])
# read until exhaustion
while not t[0]:
t = read_variant(infile, p, 'vcf',
False, [], False,
p.index, [])
eof, k, var_name, kstrains, nkstrains, af, missing = t
self.assertEqual(eof, True)
self.assertEqual(k, None)
self.assertEqual(var_name, None)
self.assertEqual(kstrains, None)
self.assertEqual(nkstrains, None)
self.assertEqual(af, None)
self.assertEqual(missing, None)
# different file
infile = gzip.open(KMER)
with self.assertRaises(AttributeError):
t = read_variant(infile, p.head(5), 'vcf',
False, [], False,
p.head(5).index, [])
infile = open(PRES)
with self.assertRaises(AttributeError):
t = read_variant(infile, p.head(5), 'vcf',
False, [], False,
p.head(5).index, [])
# burden with missing genotypes in last read variant
# issue #90
p = pd.read_csv(P,
index_col=0,
sep='\t')['binary']
infile = VariantFile(VCFMISSING)
burden_regions = deque([])
load_burden(BMISSING, burden_regions)
t = read_variant(infile, p.head(5), 'vcf',
True, burden_regions, False,
p.head(5).index, [])
eof, k, var_name, kstrains, nkstrains, af, missing = t
self.assertEqual(eof, False)
self.assertTrue(abs((k -
np.array([1, 1, 0, 0, 0])).max()) < 1E-7)
self.assertEqual(var_name,
'CDS1')
self.assertEqual(kstrains,
['sample_1', 'sample_2'])
self.assertEqual(nkstrains,
['sample_3', 'sample_4', 'sample_5'])
self.assertEqual(af, 0.4)
self.assertEqual(missing, 0)
# check that missing variants are properly missed
# issue #120
p = pd.read_csv(P,
index_col=0,
sep='\t')['binary']
infile = VariantFile(VCFMISSING)
variant = next(infile)
total = 0
missing = 0
samples = set()
for sample, call in variant.samples.items():
if sample not in p.index:
continue
for haplotype in call.get('GT', [None]):
if haplotype is None or haplotype == '.':
missing += 1
total += 1
samples.add(sample)
pysam_missing = missing / float(total)
infile = VariantFile(VCFMISSING)
t = read_variant(infile, p, 'vcf', False, [], False, p.index, [])
eof, k, var_name, kstrains, nkstrains, af, missing = t
self.assertEqual(pysam_missing, missing)