def test_load_trio(self):
''' test that load_trio() works correctly
'''
def make_vcf(person):
# make a VCF, where one line would pass the default filtering
vcf = make_vcf_header()
vcf.append(make_vcf_line(pos=1, extra='HGNC=TEST;MAX_AF=0.0001'))
vcf.append(make_vcf_line(pos=2, extra='HGNC=ATRX;MAX_AF=0.0001'))
path = os.path.join(self.temp_dir, "{}.vcf.gz".format(person))
write_gzipped_vcf(path, vcf)
return path
child_path = make_vcf('child')
mother_path = make_vcf('mother')
father_path = make_vcf('father')
family = Family('fam_id')
family.add_child('sample', 'mother_id', 'father_id', 'female', '2', child_path)
family.add_mother('mother_id', '0', '0', 'female', '1', mother_path)
family.add_father('father_id', '0', '0', 'male', '1', father_path)
family.set_child()
sum_x_lr2_proband = 0
# define the parameters and values for the SNV class
args = {'chrom': "1", 'position': 2, 'id': ".", 'ref': "G", 'alts': "T",
'filter': "PASS", 'info': "CQ=missense_variant;HGNC=ATRX;MAX_AF=0.0001",
'format': "DP:GT", 'sample': "50:0/1", 'gender': "female",
'mnv_code': None, 'qual': '1000'}
dad_args = copy.deepcopy(args)
dad_args['gender'] = 'male'
self.assertEqual(load_trio(family, sum_x_lr2_proband),
[TrioGenotypes(chrom="1", pos=2, child=SNV(**args),
mother=SNV(**args), father=SNV(**dad_args)) ])
def test_load_variants(self):
''' test that load_variants() works correctly. Mainly checks variables
are set correctly.
'''
vcf = make_minimal_vcf()
path = os.path.join(self.temp_dir, "temp.vcf.gz")
write_gzipped_vcf(path, vcf)
sum_x_lr2 = {}
fam = Family('fam', children=[Person('fam', 'child', '0', '0', 'f', '2', path)])
variants = load_variants(fam, 0.9, ['AFR_AF'], self.known_genes, set(), sum_x_lr2)
self.assertEqual(SNV.known_genes, self.known_genes)
self.assertEqual(CNV.known_genes, self.known_genes)
self.assertEqual(Info.populations, ['AFR_AF'])
self.assertEqual(Info.last_base, set())
# and check that the
variants = load_variants(fam, 0.9, [], None, set([('1', 100)]), sum_x_lr2)
self.assertIsNone(SNV.known_genes, self.known_genes)
self.assertIsNone(CNV.known_genes, self.known_genes)
self.assertEqual(Info.populations, [])
self.assertEqual(Info.last_base, set([('1', 100)]))
def test_analyse_trio(self):
''' test that analyse_trio() works correctly
'''
# construct the VCFs for the trio members
paths = {}
for member in ['child', 'mom', 'dad']:
vcf = make_vcf_header()
geno, pp_dnm = '0/0', ''
if member == 'child':
geno, pp_dnm = '0/1', ';DENOVO-SNP;PP_DNM=1'
vcf.append(
make_vcf_line(genotype=geno, extra='HGNC=ARID1B' + pp_dnm))
# write the VCF data to a file
handle = tempfile.NamedTemporaryFile(dir=self.temp_dir,
delete=False,
suffix='.vcf')
for x in vcf:
handle.write(x.encode('utf8'))
handle.flush()
paths[member] = handle.name
# create a Family object, so we can load the data from the trio's VCFs
fam_id = 'fam01'
child = Person(fam_id, 'child', 'dad', 'mom', 'female', '2',
paths['child'])
mom = Person(fam_id, 'mom', '0', '0', 'female', '1', paths['mom'])
dad = Person(fam_id, 'dad', '0', '0', 'male', '1', paths['dad'])
family = Family(fam_id, [child], mom, dad)
self.assertEqual(self.finder.analyse_trio(family), [(TrioGenotypes(
chrom="1",
pos=1,
child=SNV(
chrom="1",
position=1,
id=".",
ref="G",
alts="T",
qual='1000',
filter="PASS",
info="CQ=missense_variant;DENOVO-SNP;HGNC=ARID1B;PP_DNM=1",
format="DP:GT",
sample="50:0/1",
gender="female",
mnv_code=None),
mother=SNV(chrom="1",
position=1,
id=".",
ref="G",
alts="T",
qual='1000',
filter="PASS",
info="CQ=missense_variant;HGNC=ARID1B",
format="DP:GT",
sample="50:0/0",
gender="female",
mnv_code=None),
father=SNV(chrom="1",
position=1,
id=".",
ref="G",
alts="T",
qual='1000',
filter="PASS",
info="CQ=missense_variant;HGNC=ARID1B",
format="DP:GT",
sample="50:0/0",
gender="male",
mnv_code=None)), ['single_variant'], [
'Monoallelic', 'Mosaic'
], ['ARID1B'])])
def test_find_variants(self):
""" test that find_variants() works correctly
"""
# define the trio, so that we can know whether the parents are affected.
# The child also needs to be included and set, so that we can get the
# child ID for logging purposes.
family = Family("famID")
family.add_child("child_id", 'dad_id', 'mom_id', 'f', '2', "/vcf/path")
family.add_father("dad_id", '0', '0', 'm', '1', "/vcf/path")
family.add_mother("mom_id", '0', '0', 'f', '1', "/vcf/path")
family.set_child()
# create variants that cover various scenarios
snv1 = create_variant("F", "missense_variant|missense_variant",
"TEST1|TEST2")
snv2 = create_variant("F", "missense_variant|synonymous_variant",
"OTHER1|OTHER2")
snv3 = create_variant("F", "missense_variant", "")
snv4 = create_variant("F", "missense_variant", "TESTX", chrom="X")
self.finder.known_genes = {
"TEST1": {
"inh": ["Monoallelic"]
},
"OTHER1": {
"inh": ["Monoallelic"]
},
"OTHER2": {
"inh": ["Monoallelic"]
},
"TESTX": {
"inh": ["X-linked dominant"]
}
}
# check the simplest case, a variant in a known gene
self.assertEqual(
self.finder.find_variants([snv1], "TEST1", family),
[(snv1, ["single_variant"], ["Monoallelic"], ["TEST1"])])
# check that a gene not in a known gene does not pass
self.assertEqual(self.finder.find_variants([snv1], "TEST2", family),
[])
# check a variant where the gene is known, but the consequence for that
# gene is not functional, does not pass
self.assertEqual(self.finder.find_variants([snv2], "OTHER2", family),
[])
# check that intergenic variants (which lack HGNC symbols) do not pass
self.assertEqual(self.finder.find_variants([snv3], None, family), [])
# check that a variant on chrX passes through the allosomal instance
self.assertEqual(
self.finder.find_variants([snv4], "TESTX", family),
[(snv4, ["single_variant"], ["X-linked dominant"], ["TESTX"])])
# remove the known genes, so that the variants in unknown genes pass
self.finder.known_genes = None
self.assertEqual(
sorted(self.finder.find_variants([snv1], "TEST2", family)),
[(snv1, ["single_variant"], ["Monoallelic"], ["TEST2"]),
(snv1, ["single_variant"], ["Mosaic"], ["TEST2"])])
# but variants without gene symbols still are excluded
self.assertEqual(self.finder.find_variants([snv3], None, family), [])
