def test_open_individual_with_mnvs(self):
''' test that open_individual works with MNVs
'''
vcf = make_vcf_header()
vcf.append(make_vcf_line(pos=1, cq='splice_region_variant',
extra='HGNC=ATRX;MAX_AF=0.0001'))
vcf.append(make_vcf_line(pos=2, cq='missense_variant',
extra='HGNC=ATRX;MAX_AF=0.0001'))
path = os.path.join(self.temp_dir, "temp.vcf.gz")
self.write_gzipped_vcf(path, vcf)
person = Person('fam_id', 'sample', 'dad', 'mom', 'F', '2', path)
args = {'chrom': "1", 'position': 1, 'id': ".", 'ref': "G", 'alts': "T",
'filter': "PASS", 'info': "CQ=splice_region_variant;HGNC=ATRX;MAX_AF=0.0001",
'format': "DP:GT", 'sample': "50:0/1", 'gender': "female",
'mnv_code': 'modified_protein_altering_mnv'}
var1 = SNV(**args)
args['position'] = 2
args['mnv_code'] = None
args['info'] = "CQ=missense_variant;HGNC=ATRX;MAX_AF=0.0001"
var2 = SNV(**args)
# by default only one variant passes
self.assertEqual(self.vcf_loader.open_individual(person), [var2])
# if we include MNVs, then the passing variants swap
self.assertEqual(self.vcf_loader.open_individual(person,
mnvs={('1', 1): 'modified_protein_altering_mnv',
('1', 2): 'modified_synonymous_mnv'}), [var1])
def test_open_individual_with_mnvs(self):
''' test that open_individual works with MNVs
'''
vcf = make_vcf_header()
vcf.append(make_vcf_line(pos=1, cq='splice_region_variant',
extra='HGNC=ATRX;MAX_AF=0.0001'))
vcf.append(make_vcf_line(pos=2, cq='missense_variant',
extra='HGNC=ATRX;MAX_AF=0.0001'))
path = os.path.join(self.temp_dir, "temp.vcf.gz")
write_gzipped_vcf(path, vcf)
person = Person('fam_id', 'sample', 'dad', 'mom', 'F', '2', path)
args = {'chrom': "1", 'position': 1, 'id': ".", 'ref': "G", 'alts': "T",
'filter': "PASS", 'info': "CQ=splice_region_variant;HGNC=ATRX;MAX_AF=0.0001",
'format': "DP:GT", 'sample': "50:0/1", 'gender': "female",
'mnv_code': 'modified_protein_altering_mnv', 'qual': '1000'}
var1 = SNV(**args)
args['position'] = 2
args['mnv_code'] = None
args['info'] = "CQ=missense_variant;HGNC=ATRX;MAX_AF=0.0001"
var2 = SNV(**args)
# by default only one variant passes
self.assertEqual(open_individual(person), [var2])
# if we include MNVs, then the passing variants swap
self.assertEqual(open_individual(person,
mnvs={('1', 1): 'modified_protein_altering_mnv',
('1', 2): 'modified_synonymous_mnv'}), [var1])
def test_open_individual(self):
''' test that open_individual() works correctly
'''
# missing individual returns empty list
self.assertEqual(open_individual(None), [])
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, "temp.vcf")
write_temp_vcf(path, vcf)
person = Person('fam_id', 'sample', 'dad', 'mom', 'F', '2', path)
var1 = SNV(chrom="1", position=1, id=".", ref="G", alts="T",
qual='1000', filter="PASS", info="CQ=missense_variant;HGNC=TEST;MAX_AF=0.0001",
format="DP:GT", sample="50:0/1", gender="female", mnv_code=None)
var2 = SNV(chrom="1", position=2, id=".", ref="G", alts="T",
qual='1000', filter="PASS", info="CQ=missense_variant;HGNC=ATRX;MAX_AF=0.0001",
format="DP:GT", sample="50:0/1", gender="female", mnv_code=None)
self.assertEqual(open_individual(person), [var2])
# define a set of variants to automatically pass, and check that these
# variants pass.
child_keys = set([('1', 1), ('1', 2)])
self.assertEqual(open_individual(person,
child_variants=child_keys), [var1, var2])
def test_open_individual(self):
''' test that open_individual() works correctly
'''
# missing individual returns empty list
self.assertEqual(self.vcf_loader.open_individual(None), [])
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, "temp.vcf")
self.write_temp_vcf(path, vcf)
person = Person('fam_id', 'sample', 'dad', 'mom', 'F', '2', path)
var1 = SNV(chrom="1", position=1, id=".", ref="G", alts="T",
filter="PASS", info="CQ=missense_variant;HGNC=TEST;MAX_AF=0.0001",
format="DP:GT", sample="50:0/1", gender="female", mnv_code=None)
var2 = SNV(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)
self.assertEqual(self.vcf_loader.open_individual(person), [var2])
# define a set of variants to automatically pass, and check that these
# variants pass.
self.vcf_loader.child_keys = set([('1', 1), ('1', 2)])
self.assertEqual(self.vcf_loader.open_individual(person,
child_variants=True), [var1, var2])
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
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))
self.write_gzipped_vcf(path, vcf)
return path
def test_get_mnv_candidates_catch_assertion_error(self):
''' check that get_mnv_candidates works correctly
'''
lines = make_vcf_header()
lines.append(make_vcf_line(chrom='1', pos=1, extra='Protein_position=1;Codons=aaT/aaG'))
lines.append(make_vcf_line(chrom='1', pos=2, extra='Protein_position=2;Codons=Att/Ctt'))
self.write_vcf(lines)
self.assertEqual(get_mnv_candidates(self.path), {})
def test_get_mnv_candidates(self):
''' check that get_mnv_candidates works correctly
'''
lines = make_vcf_header()
lines.append(make_vcf_line(chrom='1', pos=1, extra='Protein_position=1;Codons=aaT/aaG'))
lines.append(make_vcf_line(chrom='1', pos=2, extra='Protein_position=1;Codons=Aat/Cat'))
self.write_vcf(lines)
self.assertEqual(get_mnv_candidates(self.path), {
('1', 1): 'alternate_residue_mnv', ('1', 2): 'alternate_residue_mnv'})
def test_find_nearby_variants_separated(self):
''' test that find_nearby_variants() doesn't include vars far apart
'''
lines = make_vcf_header()
lines.append(make_vcf_line(pos=1))
lines.append(make_vcf_line(pos=4))
self.write_vcf(lines)
vcf = open_vcf(self.path)
exclude_header(vcf)
self.assertEqual(find_nearby_variants(vcf), [])
def test_find_nearby_variants(self):
''' test that find_nearby_variants() works correctly
'''
lines = make_vcf_header()
lines.append(make_vcf_line(pos=1))
lines.append(make_vcf_line(pos=2))
self.write_vcf(lines)
vcf = open_vcf(self.path)
exclude_header(vcf)
self.assertEqual(find_nearby_variants(vcf), [[('1', 1), ('1', 2)]])
def test_find_nearby_variants_different_chroms(self):
''' test that find_nearby_variants() works correctly with successive
variants on different chroms, but at the same position.
'''
# get the default two variants
lines = make_vcf_header()
lines.append(make_vcf_line(chrom='1', pos=1))
lines.append(make_vcf_line(chrom='2', pos=1))
vcf = open_vcf(self.path)
exclude_header(vcf)
self.assertEqual(find_nearby_variants(vcf), [])
def test_same_aa_different_positions(self):
''' check that same_aa() works correctly for different amino acids
'''
lines = make_vcf_header()
lines.append(make_vcf_line(pos=5, extra='Protein_position=2'))
lines.append(make_vcf_line(pos=7, extra='Protein_position=3'))
lines.append(make_vcf_line(pos=8, extra='Protein_position=4'))
self.write_vcf(lines)
vcf = tabix.open(self.path)
pairs = [[('1', 7), ('1', 8)]]
self.assertEqual(same_aa(vcf, pairs), [])
def test_find_nearby_variants_different_threshold(self):
''' test that find_nearby_variants() works correctly when we change the threshold distance.
'''
# get the default two variants
lines = make_vcf_header()
lines.append(make_vcf_line(pos=1))
lines.append(make_vcf_line(pos=2))
vcf = open_vcf(self.path)
exclude_header(vcf)
# using a lower threshold shouldn't allow any of the variants to pass
self.assertEqual(find_nearby_variants(vcf, threshold=0), [])
def test_same_aa(self):
''' check that same_aa() works correctly
'''
# get the VCF lines
lines = make_vcf_header()
lines.append(make_vcf_line(pos=2, extra='Protein_position=1'))
lines.append(make_vcf_line(pos=4, extra='Protein_position=1'))
self.write_vcf(lines)
vcf = tabix.open(self.path)
pairs = [[('1', 2), ('1', 4)]]
self.assertEqual(same_aa(vcf, pairs), [[('1', 2), ('1', 4)]])
def test_find_nearby_variants_duplicate_position(self):
''' test that find_nearby_variants() works correctly with a duplicate var
'''
# get the default two variants
lines = make_vcf_header()
lines.append(make_vcf_line(pos=1))
lines.append(make_vcf_line(pos=2))
# make a third variant, but at the same position as the second
lines.append(make_vcf_line(pos=2))
self.write_vcf(lines)
vcf = open_vcf(self.path)
exclude_header(vcf)
self.assertEqual(find_nearby_variants(vcf), [[('1', 1), ('1', 2)]])
def test_same_aa_missing_protein_positions(self):
''' check that same_aa() works correctly when the vars aren't in the CDS
'''
# if one of the variants in the pair does not have a protein position
# listed (i.e. residue number), that indicates the variant could be
# affecting the splice site, so we can't use the pair.
lines = make_vcf_header()
lines.append(make_vcf_line(pos=5))
lines.append(make_vcf_line(pos=7))
lines.append(make_vcf_line(pos=8, extra='Protein_position=4'))
self.write_vcf(lines)
vcf = tabix.open(self.path)
pairs = [[('1', 7), ('1', 8)]]
self.assertEqual(same_aa(vcf, pairs), [])
def test_screen_pairs_nonstandard_pair(self):
''' test that screen_pairs() works correctly
'''
# get the VCF lines
lines = make_vcf_header()
lines.append(make_vcf_line(pos=2))
lines.append(make_vcf_line(pos=4))
lines.append(make_vcf_line(pos=5))
lines.append(make_vcf_line(pos=7))
lines.append(make_vcf_line(pos=8))
self.write_vcf(lines)
vcf = tabix.open(self.path)
# set up a list of 'pairs', where one 'pair' has three variants in it.
# we exclude 'pairs' where n != 2.
pairs = [[('1', 2), ('1', 4), ('1', 5)], [('1', 7), ('1', 8)]]
self.assertEqual(screen_pairs(vcf, pairs, is_not_indel), [[('1', 7), ('1', 8)]])
def test_open_individual_male_het_chrx(self):
""" test that open_individual() passes over hets in males on chrX
"""
# the sub-functions are all tested elsewhere, this test merely checks
# that valid variants are added to the variants list, and invalid
# variants are passed over without being added to the variants list
vcf = make_vcf_header()
vcf.append(make_vcf_line(chrom='X', pos=1, genotype='0/1',
extra='HGNC=TEST;MAX_AF=0.0001'))
path = os.path.join(self.temp_dir, "temp.vcf")
write_temp_vcf(path, vcf)
person = Person('fam_id', 'sample', 'dad', 'mom', 'M', '2', path)
self.assertEqual(open_individual(person), [])
def test_open_individual_male_het_chrx(self):
""" test that open_individual() passes over hets in males on chrX
"""
# the sub-functions are all tested elsewhere, this test merely checks
# that valid variants are added to the variants list, and invalid
# variants are passed over without being added to the variants list
vcf = make_vcf_header()
vcf.append(make_vcf_line(chrom='X', pos=1, genotype='0/1',
extra='HGNC=TEST;MAX_AF=0.0001'))
path = os.path.join(self.temp_dir, "temp.vcf")
write_temp_vcf(path, vcf)
person = Person('fam_id', 'sample', 'dad', 'mom', 'M', '2', path)
self.assertEqual(open_individual(person), [])
def test__write_vcf(self):
''' check that _write_vcf() works correctly
'''
path = tempfile.NamedTemporaryFile(suffix='.vcf.gz', dir=self.temp_dir,
delete=False)
lines = make_vcf_header() + ['X\t150\t.\tA\tG\t50\tPASS\tHGNC=TEST;'
'CQ=missense_variant;random_tag;EUR_AF=0.0005;'
'ClinicalFilterGeneInheritance=Monoallelic;'
'ClinicalFilterType=single_variant;'
'ClinicalFilterReportableHGNC=TEST\tGT:DP:INHERITANCE:'
'INHERITANCE_GENOTYPE\t0/1:50:deNovo:1,0,0\n']
_write_vcf(path.name, lines)
with gzip.open(path.name, 'r') as handle:
vcf = [ x.decode() for x in handle ]
self.assertEqual(lines, vcf)
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",
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",
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",
filter="PASS", info="CQ=missense_variant;HGNC=ARID1B",
format="DP:GT", sample="50:0/0", gender="male", mnv_code=None)),
['single_variant'], ['Monoallelic'], ['ARID1B'])])
def test_get_matches(self):
''' check that get_matches works correctly
'''
# get the VCF lines
lines = make_vcf_header()
lines.append(make_vcf_line(pos=1))
lines.append(make_vcf_line(pos=2))
lines.append(make_vcf_line(pos=4))
lines.append(make_vcf_line(pos=5))
self.write_vcf(lines)
vcf = tabix.open(self.path)
pair = [('1', 2), ('1', 4)]
# define the expected lines
var1 = parse_vcf_line(make_vcf_line(pos=2).split('\t'), self.Variant)
var2 = parse_vcf_line(make_vcf_line(pos=4).split('\t'), self.Variant)
self.assertEqual(list(get_matches(vcf, pair)), [var1, var2])
def test__write_vcf(self):
''' check that _write_vcf() works correctly
'''
path = tempfile.NamedTemporaryFile(suffix='.vcf.gz',
dir=self.temp_dir,
delete=False)
lines = make_vcf_header() + [
'X\t150\t.\tA\tG\t50\tPASS\tHGNC=TEST;'
'CQ=missense_variant;random_tag;EUR_AF=0.0005;'
'ClinicalFilterGeneInheritance=Monoallelic;'
'ClinicalFilterType=single_variant;'
'ClinicalFilterReportableHGNC=TEST\tGT:DP:INHERITANCE:'
'INHERITANCE_GENOTYPE\t0/1:50:deNovo:1,0,0\n'
]
_write_vcf(path.name, lines)
with gzip.open(path.name, 'r') as handle:
vcf = [x.decode() for x in handle]
self.assertEqual(lines, vcf)
def test_screen_pairs(self):
''' test that screen_pairs() works correctly
'''
# get the VCF lines
lines = make_vcf_header()
lines.append(make_vcf_line(pos=1))
lines.append(make_vcf_line(pos=2))
lines.append(make_vcf_line(pos=4))
lines.append(make_vcf_line(pos=5))
lines.append(make_vcf_line(pos=7))
lines.append(make_vcf_line(pos=8))
self.write_vcf(lines)
vcf = tabix.open(self.path)
pairs = [[('1', 2), ('1', 4)], [('1', 7), ('1', 8)]]
self.assertEqual(screen_pairs(vcf, pairs, is_not_indel), pairs)
# check that the other filter function also works cleanly
self.assertEqual(screen_pairs(vcf, pairs, is_coding), pairs)
def test__get_vcf_lines(self):
""" check that _get_vcf_lines() works correctly
"""
# write VFs for the trio members, in order to be able to pick up the
# VCF provenance information
family = self.trio
for member in [family.child, family.mother, family.father]:
header = make_vcf_header()
with open(member.get_path(), 'w') as handle:
handle.writelines(header)
# define what the header will become
vcf_lines = [
"##fileformat=VCFv4.1\n", '##fileDate=2014-01-01\n',
'##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">\n',
'##INFO=<ID=ClinicalFilterType,Number=.,Type=String,Description="The '
'type of clinical filter that passed this variant.">\n',
'##INFO=<ID=ClinicalFilterGeneInheritance,Number=.,Type=String,'
'Description="The inheritance mode (Monoallelic, Biallelic etc) '
'under which the variant was found.">\n',
'##INFO=<ID=ClinicalFilterReportableHGNC,Number=.,Type=String,'
'Description="The HGNC symbol which the variant was identified '
'as being reportable for.">\n',
'##INFO=<ID=CANDIDATE_MNV,Number=.,Type=String,'
'Description="Code for candidate multinucleotide variants. '
'Field is only included if the translated MNV differs from '
'both of the SNV translations. There are five possibilities: '
'alternate_residue_mnv=MNV translates to a residue not in SNVs, '
'masked_stop_gain_mnv=MNV masks a stop gain, '
'modified_stop_gained_mnv=MNV introduces a stop gain, '
'modified_synonymous_mnv=MNV reverts to synonymous, '
'modified_protein_altering_mnv=synonymous SNVs but missense '
'MNV.">\n',
'##FORMAT=<ID=INHERITANCE_GENOTYPE,Number=.,Type=String,'
'Description="The 012 coded genotypes for a trio (child, '
'mother, father).">\n',
'##FORMAT=<ID=INHERITANCE,Number=.,Type=String,Description="'
'The inheritance of the variant in the trio (biparental, '
'paternal, maternal, deNovo).">\n',
"##ClinicalFilterRunDate={0}\n".format(
datetime.date.today()), "##ClinicalFilterVersion={}\n".format(
clinicalfilter.__version__),
"##ClinicalFilterHistory=single_variant,compound_het\n",
"##UberVCF_proband_Id=child\n",
"##UberVCF_proband_Checksum=0a35d4e98a0f07153584f255ead8a2014ebacad0\n",
"##UberVCF_proband_Basename=child.vcf\n",
"##UberVCF_proband_Date=2014-01-01\n",
"##UberVCF_maternal_Id=mother\n",
"##UberVCF_maternal_Checksum=0a35d4e98a0f07153584f255ead8a2014ebacad0\n",
"##UberVCF_maternal_Basename=mother.vcf\n",
"##UberVCF_maternal_Date=2014-01-01\n",
"##UberVCF_paternal_Id=father\n",
"##UberVCF_paternal_Checksum=0a35d4e98a0f07153584f255ead8a2014ebacad0\n",
"##UberVCF_paternal_Basename=father.vcf\n",
"##UberVCF_paternal_Date=2014-01-01\n",
"#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\tFORMAT\tsample\n"
]
# define what the default variant vcf line will become
line = [
'X\t150\t.\tA\tG\t50\tPASS\tCQ=missense_variant;'
'ClinicalFilterGeneInheritance=Monoallelic;'
'ClinicalFilterReportableHGNC=TEST;ClinicalFilterType=single_variant;'
'DENOVO-SNP;HGNC=TEST;HGNC_ID=1001;MAX_AF=0.0005\tGT:DP:INHERITANCE:'
'INHERITANCE_GENOTYPE\t0/1:50:deNovo:1,0,0\n'
]
# check that a list of one variant produces the correct VCF output. Note
# that we haven't checked against CNVs, which can change the
# INHERITANCE_GENOTYPE flag, nor have we tested a larger list of variants
var = (self.variants[0], ["single_variant"], ["Monoallelic"], ["TEST"])
var[0].child.add_vcf_line([
'X', '150', '.', 'A', 'G', '50', 'PASS',
'HGNC=TEST;HGNC_ID=1001;CQ=missense_variant;EUR_AF=0.0005',
'GT:DP', '0/1:50'
])
self.assertEqual(list(_get_vcf_lines([var], family)), vcf_lines + line)
def test__make_vcf_header(self):
""" check that _make_vcf_header() works correctly
"""
# define the intial header lines
header = make_vcf_header()
# define the VCF provenances
provenance = [("checksum", "proband.calls.date.vcf.gz", "2014-01-01"),
("checksum", "mother.calls.date.vcf.gz", "2014-01-02"),
("checksum", "father.calls.date.vcf.gz", "2014-01-03")]
processed_header = [
"##fileformat=VCFv4.1\n", '##fileDate=2014-01-01\n',
"##FORMAT=<ID=GT,Number=1,Type=String,Description=\"Genotype\">\n",
'##INFO=<ID=ClinicalFilterType,Number=.,Type=String,'
'Description="The type of clinical filter that passed this '
'variant.">\n',
'##INFO=<ID=ClinicalFilterGeneInheritance,Number=.,Type=String,'
'Description="The inheritance mode (Monoallelic, Biallelic '
'etc) under which the variant was found.">\n',
'##INFO=<ID=ClinicalFilterReportableHGNC,Number=.,Type=String,'
'Description="The HGNC symbol which the variant was identified '
'as being reportable for.">\n',
'##INFO=<ID=CANDIDATE_MNV,Number=.,Type=String,'
'Description="Code for candidate multinucleotide variants. '
'Field is only included if the translated MNV differs from '
'both of the SNV translations. There are five possibilities: '
'alternate_residue_mnv=MNV translates to a residue not in SNVs, '
'masked_stop_gain_mnv=MNV masks a stop gain, '
'modified_stop_gained_mnv=MNV introduces a stop gain, '
'modified_synonymous_mnv=MNV reverts to synonymous, '
'modified_protein_altering_mnv=synonymous SNVs but missense '
'MNV.">\n',
'##FORMAT=<ID=INHERITANCE_GENOTYPE,Number=.,Type=String,'
'Description="The 012 coded genotypes for a trio (child, '
'mother, father).">\n',
'##FORMAT=<ID=INHERITANCE,Number=.,Type=String,Description="The '
'inheritance of the variant in the trio (biparental, paternal, '
'maternal, deNovo).">\n', "##ClinicalFilterRunDate={0}\n".format(
datetime.date.today()), "##ClinicalFilterVersion={}\n".format(
clinicalfilter.__version__),
"##ClinicalFilterHistory=single_variant,compound_het\n",
"##UberVCF_proband_Id=proband\n",
"##UberVCF_proband_Checksum=checksum\n",
"##UberVCF_proband_Basename=proband.calls.date.vcf.gz\n",
"##UberVCF_proband_Date=2014-01-01\n",
"##UberVCF_maternal_Id=mother\n",
"##UberVCF_maternal_Checksum=checksum\n",
"##UberVCF_maternal_Basename=mother.calls.date.vcf.gz\n",
"##UberVCF_maternal_Date=2014-01-02\n",
"##UberVCF_paternal_Id=father\n",
"##UberVCF_paternal_Checksum=checksum\n",
"##UberVCF_paternal_Basename=father.calls.date.vcf.gz\n",
"##UberVCF_paternal_Date=2014-01-03\n",
"#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\tFORMAT\tsample\n"
]
# check that the standard function returns the expected value. Note that
# I haven't checked the output if self.known_genes_date is not None, nor
# have I checked if the _clinicalFilterVersion is available
self.assertEqual(_make_vcf_header(header, provenance),
processed_header)
def test__make_vcf_header(self):
""" check that _make_vcf_header() works correctly
"""
# define the intial header lines
header = make_vcf_header()
# define the VCF provenances
provenance = [("checksum", "proband.calls.date.vcf.gz", "2014-01-01"),
("checksum", "mother.calls.date.vcf.gz", "2014-01-02"),
("checksum", "father.calls.date.vcf.gz", "2014-01-03")]
processed_header = ["##fileformat=VCFv4.1\n",
'##fileDate=2014-01-01\n',
"##FORMAT=<ID=GT,Number=1,Type=String,Description=\"Genotype\">\n",
'##INFO=<ID=ClinicalFilterType,Number=.,Type=String,'
'Description="The type of clinical filter that passed this '
'variant.">\n',
'##INFO=<ID=ClinicalFilterGeneInheritance,Number=.,Type=String,'
'Description="The inheritance mode (Monoallelic, Biallelic '
'etc) under which the variant was found.">\n',
'##INFO=<ID=ClinicalFilterReportableHGNC,Number=.,Type=String,'
'Description="The HGNC symbol which the variant was identified '
'as being reportable for.">\n',
'##INFO=<ID=CANDIDATE_MNV,Number=.,Type=String,'
'Description="Code for candidate multinucleotide variants. '
'Field is only included if the translated MNV differs from '
'both of the SNV translations. There are five possibilities: '
'alternate_residue_mnv=MNV translates to a residue not in SNVs, '
'masked_stop_gain_mnv=MNV masks a stop gain, '
'modified_stop_gained_mnv=MNV introduces a stop gain, '
'modified_synonymous_mnv=MNV reverts to synonymous, '
'modified_protein_altering_mnv=synonymous SNVs but missense '
'MNV.">\n',
'##FORMAT=<ID=INHERITANCE_GENOTYPE,Number=.,Type=String,'
'Description="The 012 coded genotypes for a trio (child, '
'mother, father).">\n',
'##FORMAT=<ID=INHERITANCE,Number=.,Type=String,Description="The '
'inheritance of the variant in the trio (biparental, paternal, '
'maternal, deNovo).">\n',
"##ClinicalFilterRunDate={0}\n".format(datetime.date.today()),
"##ClinicalFilterVersion={}\n".format(clinicalfilter.__version__),
"##ClinicalFilterHistory=single_variant,compound_het\n",
"##UberVCF_proband_Id=proband\n",
"##UberVCF_proband_Checksum=checksum\n",
"##UberVCF_proband_Basename=proband.calls.date.vcf.gz\n",
"##UberVCF_proband_Date=2014-01-01\n",
"##UberVCF_maternal_Id=mother\n",
"##UberVCF_maternal_Checksum=checksum\n",
"##UberVCF_maternal_Basename=mother.calls.date.vcf.gz\n",
"##UberVCF_maternal_Date=2014-01-02\n",
"##UberVCF_paternal_Id=father\n",
"##UberVCF_paternal_Checksum=checksum\n",
"##UberVCF_paternal_Basename=father.calls.date.vcf.gz\n",
"##UberVCF_paternal_Date=2014-01-03\n",
"#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\tFORMAT\tsample\n"]
# check that the standard function returns the expected value. Note that
# I haven't checked the output if self.known_genes_date is not None, nor
# have I checked if the _clinicalFilterVersion is available
self.assertEqual(_make_vcf_header(header, provenance),
processed_header)
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__get_vcf_lines(self):
""" check that _get_vcf_lines() works correctly
"""
# write VFs for the trio members, in order to be able to pick up the
# VCF provenance information
family = self.trio
for member in [family.child, family.mother, family.father]:
header = make_vcf_header()
with open(member.get_path(), 'w') as handle:
handle.writelines(header)
# define what the header will become
vcf_lines = ["##fileformat=VCFv4.1\n",
'##fileDate=2014-01-01\n',
'##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">\n',
'##INFO=<ID=ClinicalFilterType,Number=.,Type=String,Description="The '
'type of clinical filter that passed this variant.">\n',
'##INFO=<ID=ClinicalFilterGeneInheritance,Number=.,Type=String,'
'Description="The inheritance mode (Monoallelic, Biallelic etc) '
'under which the variant was found.">\n',
'##INFO=<ID=ClinicalFilterReportableHGNC,Number=.,Type=String,'
'Description="The HGNC symbol which the variant was identified '
'as being reportable for.">\n',
'##INFO=<ID=CANDIDATE_MNV,Number=.,Type=String,'
'Description="Code for candidate multinucleotide variants. '
'Field is only included if the translated MNV differs from '
'both of the SNV translations. There are five possibilities: '
'alternate_residue_mnv=MNV translates to a residue not in SNVs, '
'masked_stop_gain_mnv=MNV masks a stop gain, '
'modified_stop_gained_mnv=MNV introduces a stop gain, '
'modified_synonymous_mnv=MNV reverts to synonymous, '
'modified_protein_altering_mnv=synonymous SNVs but missense '
'MNV.">\n',
'##FORMAT=<ID=INHERITANCE_GENOTYPE,Number=.,Type=String,'
'Description="The 012 coded genotypes for a trio (child, '
'mother, father).">\n',
'##FORMAT=<ID=INHERITANCE,Number=.,Type=String,Description="'
'The inheritance of the variant in the trio (biparental, '
'paternal, maternal, deNovo).">\n',
"##ClinicalFilterRunDate={0}\n".format(datetime.date.today()),
"##ClinicalFilterVersion={}\n".format(clinicalfilter.__version__),
"##ClinicalFilterHistory=single_variant,compound_het\n",
"##UberVCF_proband_Id=child\n",
"##UberVCF_proband_Checksum=0a35d4e98a0f07153584f255ead8a2014ebacad0\n",
"##UberVCF_proband_Basename=child.vcf\n",
"##UberVCF_proband_Date=2014-01-01\n",
"##UberVCF_maternal_Id=mother\n",
"##UberVCF_maternal_Checksum=0a35d4e98a0f07153584f255ead8a2014ebacad0\n",
"##UberVCF_maternal_Basename=mother.vcf\n",
"##UberVCF_maternal_Date=2014-01-01\n",
"##UberVCF_paternal_Id=father\n",
"##UberVCF_paternal_Checksum=0a35d4e98a0f07153584f255ead8a2014ebacad0\n",
"##UberVCF_paternal_Basename=father.vcf\n",
"##UberVCF_paternal_Date=2014-01-01\n",
"#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\tFORMAT\tsample\n"]
# define what the default variant vcf line will become
line = ['X\t150\t.\tA\tG\t50\tPASS\tCQ=missense_variant;'
'ClinicalFilterGeneInheritance=Monoallelic;'
'ClinicalFilterReportableHGNC=TEST;ClinicalFilterType=single_variant;'
'DENOVO-SNP;HGNC=TEST;HGNC_ID=1001;MAX_AF=0.0005\tGT:DP:INHERITANCE:'
'INHERITANCE_GENOTYPE\t0/1:50:deNovo:1,0,0\n']
# check that a list of one variant produces the correct VCF output. Note
# that we haven't checked against CNVs, which can change the
# INHERITANCE_GENOTYPE flag, nor have we tested a larger list of variants
var = (self.variants[0], ["single_variant"], ["Monoallelic"], ["TEST"])
var[0].child.add_vcf_line(['X', '150', '.', 'A', 'G', '50',
'PASS', 'HGNC=TEST;HGNC_ID=1001;CQ=missense_variant;EUR_AF=0.0005',
'GT:DP', '0/1:50'])
self.assertEqual(list(_get_vcf_lines([var], family)), vcf_lines + line)
