def test_format_depths(genotype, alleles, expected):
"""
Tests the function which formats depths for the MAF.
"""
idx = Extractors.VariantAlleleIndexExtractor.extract(genotype)
gt, depths = Extractors.GenotypeAndDepthsExtractor.extract(
idx, genotype, alleles)
dp, ref_ct, alt_ct = Formatters.format_depths(genotype=gt,
depths=depths,
var_allele_idx=idx,
default_total_dp=0)
assert (dp, ref_ct, alt_ct) == expected
def transform(self, vcf_record, data, is_tumor_only, line_number=None):
"""
Transform into maf record.
"""
# Generic data
collection = InputCollection()
keys = itemgetter("selected_effect", itemgetter("Hugo_Symbol"))
collection.add(
column="Hugo_Symbol",
value=data["selected_effect"].get("Hugo_Symbol"),
default="Unknown",
)
collection.add(
column="Entrez_Gene_Id", value=data["selected_effect"]["Entrez_Gene_Id"]
)
collection.add(column="Center", value=self.options["maf_center"])
collection.add(column="NCBI_Build", value="GRCh38")
collection.add(column="Chromosome", value=vcf_record.chrom)
collection.add(column="Start_Position", value=data["location_data"]["start"])
collection.add(column="End_Position", value=data["location_data"]["stop"])
collection.add(column="Strand", value="+")
collection.add(column="Variant_Classification", value=data["variant_class"])
collection.add(column="Variant_Type", value=data["location_data"]["var_type"])
collection.add(
column="Reference_Allele", value=data["location_data"]["ref_allele"]
)
for k, v in zip(
["Tumor_Seq_Allele1", "Tumor_Seq_Allele2"],
format_alleles(
genotype=data["tumor_gt"],
alleles=data["location_data"]["alleles"],
defaults=[
data["location_data"]["ref_allele"],
data["location_data"]["var_allele"],
],
),
):
collection.add(column=k, value=v)
if not is_tumor_only:
for k, v in zip(
["Match_Norm_Seq_Allele1", "Match_Norm_Seq_Allele2"],
format_alleles(
genotype=data["normal_gt"],
alleles=data["location_data"]["alleles"],
defaults=[
data["location_data"]["ref_allele"],
data["location_data"]["ref_allele"],
],
),
):
collection.add(column=k, value=v)
else:
for k in ["Match_Norm_Seq_Allele1", "Match_Norm_Seq_Allele2"]:
collection.add(column=k, value="")
collection.add(column="dbSNP_RS", value=data["selected_effect"]["dbSNP_RS"])
collection.add(
column="Tumor_Sample_Barcode", value=self.options["tumor_submitter_id"]
)
collection.add(
column="Matched_Norm_Sample_Barcode",
value=self.options["normal_submitter_id"],
default="",
)
collection.add(column="Sequencer", value=self.options["sequencer"], default="")
collection.add(
column="Tumor_Sample_UUID", value=self.options["tumor_aliquot_uuid"]
)
collection.add(
column="Matched_Norm_Sample_UUID",
value=self.options["normal_aliquot_uuid"],
default="",
)
collection.add(column="all_effects", value=";".join(data["effects"]))
for k, v in zip(
["t_depth", "t_ref_count", "t_alt_count"],
format_depths(
genotype=data["tumor_gt"],
depths=data["tumor_depths"],
var_allele_idx=data["var_allele_idx"],
default_total_dp=0,
),
):
collection.add(column=k, value=v)
if not is_tumor_only:
for k, v in zip(
["n_depth", "n_ref_count", "n_alt_count"],
format_depths(
genotype=data["normal_gt"],
depths=data["normal_depths"],
var_allele_idx=data["var_allele_idx"],
),
):
collection.add(column=k, value=v)
else:
for k in ["n_depth", "n_ref_count", "n_alt_count"]:
collection.add(column=k, value=None)
for k in data["selected_effect"]:
if k in self._colset and k not in collection._colset:
collection.add(column=k, value=data["selected_effect"][k])
# Set other uuids
collection.add(column="src_vcf_id", value=self.options["src_vcf_uuid"])
collection.add(column="tumor_bam_uuid", value=self.options["tumor_bam_uuid"])
collection.add(column="normal_bam_uuid", value=self.options["normal_bam_uuid"])
collection.add(column="case_id", value=self.options["case_uuid"])
# VCF columns
collection.add(column="FILTER", value=";".join(sorted(list(vcf_record.filter))))
collection.add(column="vcf_region", value=data["vcf_columns"]["vcf_region"])
collection.add(column="vcf_info", value=data["vcf_columns"]["vcf_info"])
collection.add(column="vcf_format", value=data["vcf_columns"]["vcf_format"])
collection.add(column="vcf_tumor_gt", value=data["vcf_columns"]["vcf_tumor_gt"])
collection.add(
column="vcf_normal_gt", value=data["vcf_columns"].get("vcf_normal_gt")
)
# Set the other columns to none
collection.add(column="Score", value="")
collection.add(column="BAM_File", value="")
collection.add(column="Sequencing_Phase", value="")
anno_set = ("dbSNP_Val_Status", "COSMIC", "CONTEXT", "Mutation_Status")
for i in self._colset - set(collection.columns()):
if i not in anno_set:
collection.add(column=i, value=None)
collection.transform(self._scheme)
# Generate maf record
maf_record = init_empty_maf_record(line_number=line_number)
for i in collection:
maf_record += i.transformed
# Annotations
if self.annotators["dbsnp_priority_db"]:
maf_record = self.annotators["dbsnp_priority_db"].annotate(maf_record)
else:
maf_record["dbSNP_Val_Status"] = get_builder(
"dbSNP_Val_Status", self._scheme, value=None
)
if self.annotators["cosmic_id"]:
maf_record = self.annotators["cosmic_id"].annotate(maf_record, vcf_record)
else:
maf_record["COSMIC"] = get_builder("COSMIC", self._scheme, value=None)
if self.annotators["non_tcga_exac"]:
maf_record = self.annotators["non_tcga_exac"].annotate(
maf_record, vcf_record, var_allele_idx=data["var_allele_idx"]
)
if self.annotators["hotspots"]:
maf_record = self.annotators["hotspots"].annotate(maf_record)
else:
maf_record["hotspot"] = get_builder("hotspot", self._scheme, value=None)
maf_record = self.annotators["reference_context"].annotate(
maf_record, vcf_record
)
maf_record = self.annotators["mutation_status"].annotate(
maf_record, vcf_record, self.options["tumor_vcf_id"]
)
# Filters
gdc_filters = []
for filt_key in self.filters:
filt_obj = self.filters[filt_key]
if filt_obj and filt_obj.filter(maf_record):
gdc_filters.extend(filt_obj.tags)
maf_record["GDC_FILTER"] = get_builder(
"GDC_FILTER", self._scheme, value=";".join(sorted(gdc_filters))
)
return maf_record
def transform(self, vcf_record, data, is_tumor_only, line_number=None):
"""
Transform into maf record.
"""
# Generic data
collection = InputCollection()
collection.add(
column="Hugo_Symbol",
value=data["selected_effect"].get("Hugo_Symbol"),
default="Unknown",
)
collection.add(column="Center", value=self.options["maf_center"])
collection.add(column="NCBI_Build", value="GRCh38")
collection.add(column="Chromosome", value=vcf_record.chrom)
collection.add(column="Start_Position",
value=data["location_data"]["start"])
collection.add(column="End_Position",
value=data["location_data"]["stop"])
collection.add(column="Strand", value="+")
collection.add(column="Variant_Classification",
value=data["variant_class"])
collection.add(column="Variant_Type",
value=data["location_data"]["var_type"])
collection.add(column="Reference_Allele",
value=data["location_data"]["ref_allele"])
for k, v in zip(
["Tumor_Seq_Allele1", "Tumor_Seq_Allele2"],
format_alleles(
genotype=data["tumor_gt"],
alleles=data["location_data"]["alleles"],
defaults=[
data["location_data"]["ref_allele"],
data["location_data"]["var_allele"],
],
),
):
collection.add(column=k, value=v)
if not is_tumor_only:
for k, v in zip(
["Match_Norm_Seq_Allele1", "Match_Norm_Seq_Allele2"],
format_alleles(
genotype=data["normal_gt"],
alleles=data["location_data"]["alleles"],
defaults=[
data["location_data"]["ref_allele"],
data["location_data"]["ref_allele"],
],
),
):
collection.add(column=k, value=v)
else:
for k in ["Match_Norm_Seq_Allele1", "Match_Norm_Seq_Allele2"]:
collection.add(column=k, value="")
collection.add(column="dbSNP_RS",
value=data["selected_effect"]["dbSNP_RS"])
collection.add(column="Tumor_Sample_Barcode",
value=self.options["tumor_submitter_id"])
collection.add(
column="Matched_Norm_Sample_Barcode",
value=self.options["normal_submitter_id"],
default="",
)
collection.add(column="Sequencer",
value=self.options["sequencer"],
default="")
collection.add(column="Tumor_Sample_UUID",
value=self.options["tumor_aliquot_uuid"])
collection.add(
column="Matched_Norm_Sample_UUID",
value=self.options["normal_aliquot_uuid"],
default="",
)
collection.add(column="all_effects", value=";".join(data["effects"]))
for k, v in zip(
["t_depth", "t_ref_count", "t_alt_count"],
format_depths(
genotype=data["tumor_gt"],
depths=data["tumor_depths"],
var_allele_idx=data["var_allele_idx"],
default_total_dp=0,
),
):
collection.add(column=k, value=v)
if not is_tumor_only:
for k, v in zip(
["n_depth", "n_ref_count", "n_alt_count"],
format_depths(
genotype=data["normal_gt"],
depths=data["normal_depths"],
var_allele_idx=data["var_allele_idx"],
),
):
collection.add(column=k, value=v)
else:
for k in ["n_depth", "n_ref_count", "n_alt_count"]:
collection.add(column=k, value=None)
# Add other columns from selected_effect if they are canonical in the schema
for k in data["selected_effect"]:
if k in self._colset and k not in collection._colset:
collection.add(column=k, value=data["selected_effect"][k])
# Set other uuids
collection.add(column="src_vcf_id", value=self.options["src_vcf_uuid"])
collection.add(column="tumor_bam_uuid",
value=self.options["tumor_bam_uuid"])
collection.add(column="normal_bam_uuid",
value=self.options["normal_bam_uuid"])
collection.add(column="case_id", value=self.options["case_uuid"])
# VCF columns
collection.add(column="FILTER",
value=";".join(sorted(list(vcf_record.filter))))
collection.add(column="vcf_region",
value=data["vcf_columns"]["vcf_region"])
collection.add(column="vcf_info",
value=data["vcf_columns"]["vcf_info"])
collection.add(column="vcf_format",
value=data["vcf_columns"]["vcf_format"])
collection.add(column="vcf_tumor_gt",
value=data["vcf_columns"]["vcf_tumor_gt"])
collection.add(column="vcf_normal_gt",
value=data["vcf_columns"].get("vcf_normal_gt"))
# Set the other columns to none
collection.add(column="Score", value="")
collection.add(column="BAM_File", value="")
collection.add(column="Sequencing_Phase", value="")
# I don't think this is needed, all of these are created at initialization from the schema columns
# dbSNP_Val_Status needs to remain as column but will always be empty
anno_set = ("COSMIC", "CONTEXT", "Mutation_Status")
for i in self._colset - set(collection.columns()):
if i not in anno_set:
collection.add(column=i, value=None)
collection.transform(self._scheme)
# Generate maf record
maf_record = init_empty_maf_record(line_number=line_number)
for i in collection:
maf_record += i.transformed
# # check if None introduced before here <------ DEBUG
# foo = set(maf_record.keys()) - set(self._columns)
# if len(foo) != 0:
# raise KeyError("Unexpected keys found: {}".format(foo))
# Annotations
maf_record["dbSNP_Val_Status"] = get_builder("dbSNP_Val_Status",
self._scheme,
value=None)
if self.annotators["cosmic_id"]:
maf_record = self.annotators["cosmic_id"].annotate(
maf_record, vcf_record)
else:
maf_record["COSMIC"] = get_builder("COSMIC",
self._scheme,
value=None)
if self.annotators["hotspots"]:
maf_record = self.annotators["hotspots"].annotate(maf_record)
else:
maf_record["hotspot"] = get_builder("hotspot",
self._scheme,
value=None)
if self.annotators["entrez_gene_id"]:
maf_record = self.annotators["entrez_gene_id"].annotate(maf_record)
else:
maf_record["Entrez_Gene_Id"] = get_builder("entrez_gene_id",
self._scheme,
value=0)
if self.annotators["gnomad_noncancer"]:
maf_record = self.annotators["gnomad_noncancer"].annotate(
maf_record, vcf_record, data["var_allele_idx"])
maf_record = self.annotators["reference_context"].annotate(
maf_record, vcf_record)
maf_record = self.annotators["mutation_status"].annotate(
maf_record, vcf_record, self.options["tumor_vcf_id"])
# Filters
gdc_filters = []
for filt_key in self.filters:
filt_obj = self.filters[filt_key]
if filt_obj and filt_obj.filter(maf_record):
gdc_filters.extend(filt_obj.tags)
maf_record["GDC_FILTER"] = get_builder("GDC_FILTER",
self._scheme,
value=";".join(
sorted(gdc_filters)))
return maf_record