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 _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 _name_to_id(self, probe_name):
names = []
params = get_params(probe_name)
if params.get("mut"):
names.append("_".join([params.get("gene"), params.get("mut")]))
var_name = params.get("var_name")
else:
var_name = probe_name.split("?")[0].split("-")[1]
names.append(var_name)
return "-".join(names)
def _get_names(self, allele_name):
names = []
params = get_params(allele_name)
if params.get("mut"):
names.append("_".join([params.get("gene"), params.get("mut")]))
allele_name_split = allele_name.split('?')[0].split('-')
if len(allele_name_split) > 1:
names.append(allele_name_split[1])
else:
names.append(allele_name_split[0])
return names
def _update_lineage_calls_dict(self, probe_name, call):
if self.lineage_variants is None:
return
params = get_params(probe_name)
try:
var_name = params["var_name"]
lineage = self.lineage_variants[var_name]
except KeyError:
return
if lineage["name"] not in self.lineage_calls_dict:
self.lineage_calls_dict[lineage["name"]] = {}
self.lineage_calls_dict[lineage["name"]][var_name] = call
def _get_names(self, allele_name):
names = []
params = get_params(allele_name)
if params.get("mut"):
names.append("_".join([params.get("gene"), params.get("mut")]))
allele_name_split = allele_name.split('?')[0].split('-')
if len(allele_name_split) > 1:
names.append(allele_name_split[1])
else:
names.append(allele_name_split[0])
return names
def _create_variant(self, probe_name):
names = []
params = get_params(probe_name)
if params.get("mut"):
names.append("_".join([params.get("gene"), params.get("mut")]))
var_name = probe_name.split('?')[0].split('-')[1]
names.append(var_name)
try:
# If it's a variant panel we can create a variant
ref, start, alt = split_var_name(var_name)
return Variant.create(start=start,
reference_bases=ref,
alternate_bases=[alt],
names=names,
info=params)
except AttributeError:
return None
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 _update_lineage_calls_dict(self, call, probe_name=None, var_name=None):
if self.lineage_variants is None:
return
if probe_name is not None:
# probe_name is expected be of the form eg:
# ref-K43R?var_name=AAG781686AGA&num_alts=1&ref=NC_000962.3&enum=0&gene=rpsL&mut=K43R
# and we want the var_name entry from that
params = get_params(probe_name)
try:
var_name = params["var_name"]
lineage = self.lineage_variants[var_name]
except KeyError:
return
elif var_name is not None:
# We've been provided the var_name part of a probe_name, so no need to extract
# from a full probe_name
lineage = self.lineage_variants[var_name]
else:
raise Exception("Must provide probe_name or var_name")
if lineage["name"] not in self.lineage_calls_dict:
self.lineage_calls_dict[lineage["name"]] = {}
self.lineage_calls_dict[lineage["name"]][var_name] = call
