def test_low_coverage(self):
pc = ProbeCoverage(min_depth=100,
percent_coverage=16,
median_depth=16,
k_count=16,
klen=31)
s = SequenceProbeCoverage(name="A123T",
probe_coverage=pc,
percent_coverage_threshold=80
)
call = self.pt_10.type(s)
assert call.genotype == [0, 0]
pc = ProbeCoverage(min_depth=100,
percent_coverage=80,
median_depth=16,
k_count=16,
klen=31)
s = SequenceProbeCoverage(name="A123T",
probe_coverage=pc,
percent_coverage_threshold=80
)
call = self.pt_10.type(s)
assert call.genotype == [1, 1]
def _parse_variant_panel(self, row):
allele, median_depth, min_depth, percent_coverage, k_count = self._parse_summary_covgs_row(
row)
params = get_params(allele)
if 'var_name' in params:
var_name = params.get('var_name')
else:
var_name = allele.split('?')[0].split('-')[1]
num_alts = int(params.get("num_alts", 0))
reference_coverages = [
ProbeCoverage(percent_coverage=percent_coverage,
median_depth=median_depth,
min_depth=min_depth,
k_count=k_count)
]
alt_or_ref = 'ref'
alternate_coverages = []
for i in range(num_alts - 1):
row = next(self.reader)
ref_allele, median_depth, min_depth, percent_coverage, k_count = self._parse_summary_covgs_row(
row)
if ref_allele.split('-')[0] != 'ref':
logger.warning("Fewer ref alleles than alt alleles for %s" %
ref_allele)
alternate_coverages.append(
ProbeCoverage(min_depth=min_depth,
k_count=k_count,
percent_coverage=percent_coverage,
median_depth=median_depth))
num_alts -= 1
break
assert ref_allele.split('-')[0] == 'ref'
reference_coverages.append(
ProbeCoverage(percent_coverage=percent_coverage,
median_depth=median_depth,
min_depth=min_depth,
k_count=k_count))
for i in range(num_alts):
row = next(self.reader)
alt_allele, median_depth, min_depth, percent_coverage, k_count = self._parse_summary_covgs_row(
row)
assert alt_allele.split('-')[0] == 'alt'
alternate_coverages.append(
ProbeCoverage(min_depth=min_depth,
k_count=k_count,
percent_coverage=percent_coverage,
median_depth=median_depth))
variant_probe_coverage = VariantProbeCoverage(
reference_coverages=reference_coverages,
alternate_coverages=alternate_coverages,
var_name=var_name,
params=params)
try:
self.variant_covgs[allele].append(variant_probe_coverage)
except KeyError:
self.variant_covgs[allele] = [variant_probe_coverage]
def test_genotyping_gene_11(self):
pt_no_contaim = PresenceTyper(expected_depths=[20])
pt_contaim = PresenceTyper(
expected_depths=[20],
contamination_depths=[10])
pc = ProbeCoverage(min_depth=10,
percent_coverage=100,
median_depth=10,
k_count=100,
klen=31)
s = SequenceProbeCoverage(name="A123T",
probe_coverage=pc,
percent_coverage_threshold=80
)
call = pt_no_contaim.type(s)
assert call.genotype == [1, 1]
call = pt_contaim.type(s)
assert call.genotype == [0, 0]
pc = ProbeCoverage(min_depth=10,
percent_coverage=100,
median_depth=30,
k_count=30,
klen=31)
s = SequenceProbeCoverage(name="A123T",
probe_coverage=pc,
percent_coverage_threshold=80
)
call = pt_no_contaim.type(s)
assert call.genotype == [1, 1]
call = pt_contaim.type(s)
assert call.genotype == [1, 1]
pc = ProbeCoverage(min_depth=10,
percent_coverage=100,
median_depth=20,
k_count=20,
klen=31)
s = SequenceProbeCoverage(name="A123T",
probe_coverage=pc,
percent_coverage_threshold=80
)
call = pt_no_contaim.type(s)
assert call.genotype == [1, 1]
call = pt_contaim.type(s)
assert call.genotype == [1, 1]
def test_alt_vars(self):
reference_coverage = ProbeCoverage(min_depth=100,
percent_coverage=3,
median_depth=100,
k_count=3)
alternate_coverages = [
ProbeCoverage(min_depth=100,
percent_coverage=100,
median_depth=100,
k_count=100)
]
v1 = VariantProbeCoverage(var_name="A123T",
reference_coverages=[reference_coverage],
alternate_coverages=alternate_coverages)
call = self.vt.type([v1])
assert call['genotype'] == [1, 1]
def test_2(self):
reference_coverage = ProbeCoverage(min_depth=131,
percent_coverage=95.2381,
median_depth=155,
k_count=131)
alt1 = ProbeCoverage(min_depth=1,
percent_coverage=100,
median_depth=1,
k_count=1)
alternate_coverages = [alt1]
v1 = VariantProbeCoverage(var_name="A123T",
reference_coverages=[reference_coverage],
alternate_coverages=alternate_coverages)
call = self.vt_no_contaim.type(v1)
assert call['genotype'] == [0, 0]
def _parse_seq_panel(self, row):
(
allele,
median_depth,
min_depth,
percent_coverage,
k_count,
klen,
) = self._parse_summary_covgs_row(row)
probe_coverage = ProbeCoverage(
percent_coverage=percent_coverage,
median_depth=median_depth,
min_depth=min_depth,
k_count=k_count,
klen=klen,
)
allele_name = allele.split("?")[0]
params = get_params(allele)
panel_type = params.get("panel_type", "presence")
name = params.get("name")
version = params.get("version", "1")
if panel_type in ["variant", "presence"]:
sequence_probe_coverage = SequenceProbeCoverage(
name=name,
probe_coverage=probe_coverage,
version=version,
length=params.get("length"),
)
try:
self.covgs[panel_type][name][version] = sequence_probe_coverage
except KeyError:
self.covgs[panel_type][name] = {}
self.covgs[panel_type][name][version] = sequence_probe_coverage
else:
# Species panels are treated differently
l = int(params.get("length", -1))
try:
self.covgs[panel_type][name]["total_bases"] += l
if percent_coverage > 75 and median_depth > 0:
self.covgs[panel_type][name]["percent_coverage"].append(
percent_coverage)
self.covgs[panel_type][name]["length"].append(l)
self.covgs[panel_type][name]["median"].append(median_depth)
except KeyError:
if panel_type not in self.covgs:
self.covgs[panel_type] = {}
self.covgs[panel_type][name] = {}
self.covgs[panel_type][name]["total_bases"] = l
if percent_coverage > 75 and median_depth > 0:
self.covgs[panel_type][name]["percent_coverage"] = [
percent_coverage
]
self.covgs[panel_type][name]["length"] = [l]
self.covgs[panel_type][name]["median"] = [median_depth]
else:
self.covgs[panel_type][name]["percent_coverage"] = []
self.covgs[panel_type][name]["length"] = []
self.covgs[panel_type][name]["median"] = []
def test_mixed_vars2(self):
reference_coverage = ProbeCoverage(min_depth=11,
percent_coverage=100,
median_depth=42,
k_count=42)
alternate_coverages = [
ProbeCoverage(min_depth=94,
percent_coverage=100,
median_depth=102,
k_count=94)
]
v1 = VariantProbeCoverage(var_name="A123T",
reference_coverages=[reference_coverage],
alternate_coverages=alternate_coverages)
call = self.vt.type(v1)
assert call['genotype'] == [0, 1]
def test_3(self):
reference_coverage = ProbeCoverage(min_depth=2,
percent_coverage=59.52,
median_depth=2,
k_count=60)
alt1 = ProbeCoverage(min_depth=1,
percent_coverage=83.33,
median_depth=1,
k_count=83)
alternate_coverages = [alt1]
v1 = VariantProbeCoverage(var_name="A123T",
reference_coverages=[reference_coverage],
alternate_coverages=alternate_coverages)
call = self.vt2_no_contaim.type(v1)
assert call['genotype'] == [1, 1]
assert call["info"]["conf"] < 100
def test_wt_vars(self):
reference_coverage = ProbeCoverage(min_depth=100,
percent_coverage=100,
median_depth=100,
k_count=100)
alternate_coverages = [
ProbeCoverage(min_depth=100,
percent_coverage=3,
median_depth=100,
k_count=3)
]
v1 = VariantProbeCoverage(var_name="A123T",
reference_coverages=[reference_coverage],
alternate_coverages=alternate_coverages)
call = self.vt.type([v1])
assert call['genotype'] == [0, 0]
assert call["info"].get('expected_depths') == [100]
def test_base_case_no_coverage(self):
pc = ProbeCoverage(min_depth=0,
percent_coverage=0,
median_depth=0,
k_count=0)
s1 = SequenceProbeCoverage(name="A123T", probe_coverage=pc)
call = self.pt.type(s1)
assert call.genotype == [0, 0]
def test_simple_case(self):
reference_coverage = ProbeCoverage(min_depth=100,
percent_coverage=100,
median_depth=80,
k_count=80)
alternate_coverages = [
ProbeCoverage(min_depth=100,
percent_coverage=100,
median_depth=20,
k_count=40)
]
v1 = VariantProbeCoverage(var_name="A123T",
reference_coverages=[reference_coverage],
alternate_coverages=alternate_coverages)
call = self.vt_no_contaim.type(v1)
assert call['genotype'] == [0, 1]
def test_resistotype_gene_at_high_CN(self):
pc = ProbeCoverage(min_depth=100,
percent_coverage=100,
median_depth=1000,
k_count=100)
s = SequenceProbeCoverage(name="A123T",
probe_coverage=pc,
percent_coverage_threshold=80)
call = self.pt.type(s)
assert call.genotype == [1, 1]
def test_genotyping_gene_01(self):
pc = ProbeCoverage(min_depth=100,
percent_coverage=82,
median_depth=2,
k_count=82)
s = SequenceProbeCoverage(name="A123T",
probe_coverage=pc,
percent_coverage_threshold=80)
call = self.pt.type(s)
assert call.genotype == [0, 1]
def test_4(self):
vt = VariantTyper(expected_depths=[6],
contamination_depths=[],
confidence_threshold=3)
reference_coverage = ProbeCoverage(min_depth=1,
percent_coverage=100,
median_depth=2,
k_count=2)
alt1 = ProbeCoverage(min_depth=1,
percent_coverage=100,
median_depth=1,
k_count=1)
alternate_coverages = [alt1]
v1 = VariantProbeCoverage(var_name="A123T",
reference_coverages=[reference_coverage],
alternate_coverages=alternate_coverages)
call = vt.type(v1)
assert call['genotype'] == [0, 1]
print(call["info"])
assert call["info"]["conf"] < 100
def test_simple_case(self):
reference_coverage = ProbeCoverage(min_depth=100,
percent_coverage=70,
median_depth=80,
k_count=80)
alt1 = ProbeCoverage(min_depth=100,
percent_coverage=70,
median_depth=20,
k_count=20)
alt2 = ProbeCoverage(min_depth=100,
percent_coverage=100,
median_depth=80,
k_count=80)
alternate_coverages = [alt1, alt2]
v1 = VariantProbeCoverage(var_name="A123T",
reference_coverages=[reference_coverage],
alternate_coverages=alternate_coverages)
assert v1._choose_best_alternate_coverage() == alt2
call = self.vt_no_contaim.type(v1)
assert call['genotype'] == [1, 1]
def _parse_variant_panel(self, row):
(
probe,
median_depth,
min_depth,
percent_coverage,
k_count,
klen,
) = self._parse_summary_covgs_row(row)
params = get_params(probe)
probe_type = probe.split("-")[0]
if "var_name" in params:
var_name = (params.get("gene", "") + "_" + params.get("mut", "") +
"-" + params.get("var_name", ""))
else:
var_name = allele.split("?")[0].split("-")[1]
if not var_name in self.variant_covgs:
variant_probe_coverage = VariantProbeCoverage(
reference_coverages=[],
alternate_coverages=[],
var_name=probe,
params=params,
)
self.variant_covgs[var_name] = variant_probe_coverage
probe_coverage = ProbeCoverage(
min_depth=min_depth,
k_count=k_count,
percent_coverage=percent_coverage,
median_depth=median_depth,
klen=klen,
)
if probe_type == "ref":
self.variant_covgs[var_name].reference_coverages.append(
probe_coverage)
self.variant_covgs[
var_name].best_reference_coverage = self.variant_covgs[
var_name]._choose_best_reference_coverage()
elif probe_type == "alt":
self.variant_covgs[var_name].alternate_coverages.append(
probe_coverage)
self.variant_covgs[
var_name].best_alternate_coverage = self.variant_covgs[
var_name]._choose_best_alternate_coverage()
else:
raise ValueError("probe_type must be ref or alt")
def test_simple_case(self):
reference_coverage = ProbeCoverage(min_depth=100,
percent_coverage=80,
median_depth=80,
k_count=80,
klen=31)
alt1 = ProbeCoverage(min_depth=100,
percent_coverage=50,
median_depth=20,
k_count=20,
klen=31)
alt2 = ProbeCoverage(min_depth=100,
percent_coverage=40,
median_depth=80,
k_count=30,
klen=31)
alternate_coverages = [alt1, alt2]
v1 = VariantProbeCoverage(var_name="A123T",
reference_coverages=[reference_coverage],
alternate_coverages=alternate_coverages)
reference_coverage = ProbeCoverage(min_depth=100,
percent_coverage=80,
median_depth=80,
k_count=20,
klen=31)
alt1 = ProbeCoverage(min_depth=100,
percent_coverage=50,
median_depth=20,
k_count=20,
klen=31)
alt2 = ProbeCoverage(min_depth=100,
percent_coverage=100,
median_depth=80,
k_count=100,
klen=31)
alternate_coverages = [alt1, alt2]
v2 = VariantProbeCoverage(var_name="A123T",
reference_coverages=[reference_coverage],
alternate_coverages=alternate_coverages)
call = self.vt_no_contaim.type([v1, v2])
assert call['genotype'] == [1, 1]