def __get_translated_filter_and_variant_annotation(
cls,
call: SimpleCall,
panel: Panel,
call_reference_assembly: ReferenceAssembly,
translated_reference_allele: Optional[str],
) -> Tuple[FullCallFilter, str]:
annotated_alleles = cls.__get_annotated_alleles(
call, call_reference_assembly, translated_reference_allele)
if panel.has_ref_seq_difference_annotation(call.gene,
call.start_coordinate,
call.reference_allele,
call_reference_assembly):
annotate_as_ref = all(
cls.
__is_ref_allele_to_opposite_assembly_due_to_ref_sequence_difference(
annotated_allele, call_reference_assembly)
for annotated_allele in annotated_alleles)
all_variants_are_ref_to_a_reference_assembly = all(
cls.__allele_is_ref_to_a_reference_assembly(
annotated_allele, call_reference_assembly)
for annotated_allele in annotated_alleles)
if annotate_as_ref:
translated_variant_annotation = REF_CALL_ANNOTATION_STRING
translated_filter = FullCallFilter.PASS
elif all_variants_are_ref_to_a_reference_assembly:
translated_variant_annotation = panel.get_ref_seq_difference_annotation(
call.gene, call.start_coordinate, call.reference_allele,
call_reference_assembly)
if call.is_pass():
translated_filter = FullCallFilter.PASS
else:
translated_filter = FullCallFilter.INFERRED_PASS
else:
translated_variant_annotation = call.variant_annotation + "?"
translated_filter = FullCallFilter.UNKNOWN
logging.warning(
f"Unexpected allele in ref seq difference location. Check whether annotation is correct: "
f"found alleles=({annotated_alleles[0]}, {annotated_alleles[1]}), "
f"annotation={translated_variant_annotation}")
elif panel.contains_rs_id_matching_call(call, call_reference_assembly):
# known variant and no ref seq differences involved
translated_variant_annotation = call.variant_annotation
if call.is_pass():
translated_filter = FullCallFilter.PASS
else:
translated_filter = FullCallFilter.NO_CALL
else:
# unknown variant, no ref seq difference involved
translated_variant_annotation = call.variant_annotation + "?"
translated_filter = FullCallFilter.UNKNOWN
logging.warning(
f"Unknown variant. Check whether annotation is correct: "
f"found alleles=({annotated_alleles[0]}, {annotated_alleles[1]}), "
f"annotation={translated_variant_annotation}")
return translated_filter, translated_variant_annotation
def get_genotype_tsv_text(cls, pgx_analysis: PgxAnalysis, panel: Panel,
version: str) -> str:
gene_to_haplotype_calls = pgx_analysis.get_gene_to_haplotype_calls()
genes_in_analysis = set(gene_to_haplotype_calls.keys())
assert genes_in_analysis == panel.get_genes(), (
f"Gene lists inconsistent.\n"
f"From analysis={sorted(list(genes_in_analysis))}\n"
f"From panel={sorted(list(panel.get_genes()))}")
gene_to_drug_info = {}
for gene_info in panel.get_gene_infos():
sorted_drugs = sorted([drug for drug in gene_info.drugs],
key=lambda info:
(info.name, info.url_prescription_info))
gene_to_drug_info[gene_info.gene] = (cls.DRUG_SEPARATOR.join(
[drug.name for drug in sorted_drugs]),
cls.DRUG_SEPARATOR.join([
drug.url_prescription_info
for drug in sorted_drugs
]))
header = cls.TSV_SEPARATOR.join(cls.GENOTYPE_TSV_COLUMNS)
lines = [header]
for gene in sorted(gene_to_haplotype_calls.keys()):
if gene_to_haplotype_calls[gene]:
for haplotype_call in sorted(
gene_to_haplotype_calls[gene],
key=lambda call: call.haplotype_name):
lines.append(
cls.TSV_SEPARATOR.join([
gene,
haplotype_call.haplotype_name,
cls.__get_zygosity(haplotype_call),
panel.get_haplotype_function(
gene, haplotype_call.haplotype_name),
gene_to_drug_info[gene][0],
gene_to_drug_info[gene][1],
panel.get_id(),
version,
]))
else:
lines.append(
cls.TSV_SEPARATOR.join([
gene,
cls.UNRESOLVED_HAPLOTYPE_STRING,
cls.NOT_APPLICABLE_ZYGOSITY_STRING,
UNKNOWN_FUNCTION_STRING,
gene_to_drug_info[gene][0],
gene_to_drug_info[gene][1],
panel.get_id(),
version,
]))
text = "\n".join(lines) + "\n"
return text
def __get_translated_reference_allele(
cls, call: SimpleCall, panel: Panel,
call_reference_assembly: ReferenceAssembly) -> Optional[str]:
if panel.contains_rs_id_matching_call(call, call_reference_assembly):
rs_id_info = panel.get_matching_rs_id_info(
call.start_coordinate, call.reference_allele,
call_reference_assembly)
cls.__assert_rs_id_call_matches_info(call.rs_ids,
(rs_id_info.rs_id, ))
return rs_id_info.get_reference_allele(
call_reference_assembly.opposite())
else:
# unknown variant
return None
def __get_example_panel(cls) -> Panel:
dpyd_two_a_variant = Variant("rs3918290", "T")
dpyd_two_b_variant = Variant("rs1801159", "C")
dpyd_three_variant = Variant("rs72549303", "TG")
fake_variant = Variant("rs1212125", "C")
fake2_variant = Variant("rs1212127", "C")
dpyd_haplotypes = frozenset({
Haplotype("*2A", "No Function", frozenset({dpyd_two_a_variant})),
Haplotype("*2B", "No Function", frozenset({dpyd_two_a_variant, dpyd_two_b_variant})),
Haplotype("*3", "Normal Function", frozenset({dpyd_three_variant})),
})
dpyd_rs_id_infos = frozenset({
RsIdInfo("rs3918290", "C", "C", GeneCoordinate("1", 97915614), GeneCoordinate("chr1", 97450058)),
RsIdInfo("rs72549309", "GATGA", "GATGA", GeneCoordinate("1", 98205966), GeneCoordinate("chr1", 97740410)),
RsIdInfo("rs1801159", "T", "T", GeneCoordinate("1", 97981395), GeneCoordinate("chr1", 97515839)),
RsIdInfo("rs72549303", "TG", "TC", GeneCoordinate("1", 97915621), GeneCoordinate("chr1", 97450065)),
})
dpyd_drugs = frozenset({
DrugInfo("5-Fluorouracil", "https://www.pharmgkb.org/chemical/PA128406956/guidelineAnnotation/PA166104939"),
DrugInfo("Capecitabine", "https://www.pharmgkb.org/chemical/PA448771/guidelineAnnotation/PA166104963"),
})
dpyd_rs_id_to_difference_annotations = {
"rs72549303": Annotation("6744CA>GA", "6744GA>CA"),
}
fake_haplotypes = frozenset({
Haplotype("*4A", "Reduced Function", frozenset({fake_variant})),
})
fake_rs_id_infos = frozenset({
RsIdInfo("rs1212125", "T", "T", GeneCoordinate("5", 97915617), GeneCoordinate("chr5", 97450060)),
})
fake_drugs = frozenset({
DrugInfo("Aspirin", "https://www.pharmgkb.org/some_other_url"),
})
fake_rs_id_to_difference_annotations: Dict[str, Annotation] = {}
fake2_haplotypes = frozenset({
Haplotype("*4A", "Reduced Function", frozenset({fake2_variant})),
})
fake2_rs_id_infos = frozenset({
RsIdInfo("rs1212127", "C", "T", GeneCoordinate("16", 97915617), GeneCoordinate("chr16", 97450060)),
})
fake2_drugs = frozenset({
DrugInfo("Aspirin", "https://www.pharmgkb.org/some_other_url"),
})
fake2_rs_id_to_difference_annotations = {"rs1212127": Annotation("1324C>T", "1324T>C")}
gene_infos = frozenset({
GeneInfo("DPYD", "*1", dpyd_haplotypes, dpyd_rs_id_infos,
dpyd_drugs, dpyd_rs_id_to_difference_annotations),
GeneInfo("FAKE", "*1", fake_haplotypes, fake_rs_id_infos,
fake_drugs, fake_rs_id_to_difference_annotations),
GeneInfo("FAKE2", "*1", fake2_haplotypes, fake2_rs_id_infos,
fake2_drugs, fake2_rs_id_to_difference_annotations),
})
name = "Panel"
version = "0.2"
return Panel(name, version, gene_infos)
def test_haplotype_reporter_empty(self) -> None:
pgx_analysis = PgxAnalysis(FullCallData(frozenset()), {})
panel = Panel("EmptyPanel", "0.3", frozenset())
version = "V1"
result = HaplotypeReporter.get_genotype_tsv_text(pgx_analysis, panel, version)
result_expected = "gene\thaplotype\tzygosity\tfunction\tlinked_drugs\turl_prescription_info\tpanel_version\trepo_version\n"
self.assertEqual(result_expected, result)
def load_panel(panel_path: str) -> Panel:
""" Load manually annotated JSON panel file """
try:
with open(panel_path, "r+", encoding="utf-8") as json_file:
return Panel.from_json(json.load(json_file))
except IOError:
raise FileNotFoundError(
f"Panel file {panel_path} not found or cannot be opened.")
def test_panel_with_repeat_gene_names(self) -> None:
"""Error when panel has info for gene multiple times"""
name = "FakePanel"
version = "1.0"
gene1 = "FAKE"
gene2 = "OTHER"
reference_haplotype_name = "*1"
other_reference_haplotype_name = "*1_something else"
haplotypes: FrozenSet[Haplotype] = frozenset()
rs_id_infos: FrozenSet[RsIdInfo] = frozenset()
drugs: FrozenSet[DrugInfo] = frozenset()
rs_id_to_ref_seq_difference_annotation: Dict[str, Annotation] = dict()
gene_info1 = GeneInfo(
gene1,
reference_haplotype_name,
haplotypes,
rs_id_infos,
drugs,
rs_id_to_ref_seq_difference_annotation,
)
gene_info2 = GeneInfo(
gene2,
reference_haplotype_name,
haplotypes,
rs_id_infos,
drugs,
rs_id_to_ref_seq_difference_annotation,
)
gene_info3 = GeneInfo(
gene1,
other_reference_haplotype_name,
haplotypes,
rs_id_infos,
drugs,
rs_id_to_ref_seq_difference_annotation,
)
gene_infos_without_repeat = frozenset([gene_info1, gene_info2])
gene_infos_with_repeat = frozenset([gene_info1, gene_info3])
Panel(name, version, gene_infos_without_repeat)
with self.assertRaises(ValueError):
Panel(name, version, gene_infos_with_repeat)
def get_gene_to_haplotypes_call(
cls, full_call_data: FullCallData,
panel: Panel) -> Dict[str, Set[HaplotypeCall]]:
gene_to_haplotype_calls = {}
for gene_info in panel.get_gene_infos():
logging.info(f"Calling haplotypes for {gene_info.gene}")
gene_to_haplotype_calls[
gene_info.gene] = cls.__get_haplotypes_call(
full_call_data, gene_info)
return gene_to_haplotype_calls
def __coordinates_of_call_overlap_with_panel_coordinates(
cls, call_index: int, panel: Panel, variants: Dict[str, Any],
vcf_reference_assembly: ReferenceAssembly) -> bool:
chromosome = cls.__get_chromosome_from_variants(call_index, variants)
position = cls.__get_position_from_variants(call_index, variants)
reference_allele = cls.__get_reference_allele_from_variants(
call_index, variants)
relevant_coordinates = cls.__get_relevant_coordinates(
chromosome, position, reference_allele)
coordinate_match_to_panel_exists = any(
panel.contains_rs_id_with_coordinate(coord, vcf_reference_assembly)
for coord in relevant_coordinates)
return coordinate_match_to_panel_exists
def __fill_in_rs_ids_if_needed(
cls, call: SimpleCall, panel: Panel,
reference_assembly: ReferenceAssembly) -> SimpleCall:
rs_ids: Tuple[str, ...]
if call.rs_ids == (".", ) and panel.contains_rs_id_matching_call(
call, reference_assembly):
rs_id_info = panel.get_matching_rs_id_info(call.start_coordinate,
call.reference_allele,
reference_assembly)
rs_ids = (rs_id_info.rs_id, )
new_simple_call = SimpleCall(
call.start_coordinate,
call.reference_allele,
call.alleles,
call.gene,
rs_ids,
call.variant_annotation,
call.filter,
)
return new_simple_call
else:
return call
def __get_calls_for_panel_variants_without_calls(
cls, simple_call_data: SimpleCallData,
panel: Panel) -> FrozenSet[SimpleCall]:
# assume ref call when no call is found. Set filter to NO_CALL
reference_assembly = simple_call_data.reference_assembly
rs_ids_found_in_patient = {
rs_id
for call in simple_call_data.calls for rs_id in call.rs_ids
if rs_id != "."
}
coordinates_covered_by_found_calls = {
coordinate
for call in simple_call_data.calls
for coordinate in call.get_relevant_coordinates()
}
uncalled_calls = set()
for gene_info in panel.get_gene_infos():
for rs_id_info in gene_info.rs_id_infos:
coordinates_partially_handled = bool(
rs_id_info.get_relevant_coordinates(reference_assembly).
intersection(coordinates_covered_by_found_calls))
if rs_id_info.rs_id not in rs_ids_found_in_patient and not coordinates_partially_handled:
# Assuming REF/REF relative to reference assembly
start_coordinate = rs_id_info.get_start_coordinate(
reference_assembly)
reference_allele = rs_id_info.get_reference_allele(
reference_assembly)
uncalled_ref_call = SimpleCall(
start_coordinate,
reference_allele,
(reference_allele, reference_allele),
gene_info.gene,
(rs_id_info.rs_id, ),
REF_CALL_ANNOTATION_STRING,
SimpleCallFilter.NO_CALL,
)
uncalled_calls.add(uncalled_ref_call)
return frozenset(uncalled_calls)
def __rs_id_exists_in_panel(cls, call_index: int, panel: Panel,
variants: Dict[str, Any]) -> bool:
rs_ids = cls.__get_rs_ids_from_variants(call_index, variants)
rs_id_match_to_panel_exists = any(
panel.contains_rs_id(rs_id) for rs_id in rs_ids)
return rs_id_match_to_panel_exists
def __assert_gene_in_panel(cls, gene: str, panel: Panel) -> None:
if gene not in panel.get_genes():
error_msg = f"Call for unknown gene:\ngene={gene}"
raise ValueError(error_msg)
def test_panel_with_overlapping_rs_id_infos_for_different_genes(
self) -> None:
"""Error when panel has overlapping rs id infos for different genes, but not when they are exactly the same"""
name = "FakePanel"
version = "1.0"
gene1 = "FAKE"
gene2 = "OTHER"
chromosome_v37 = "X"
chromosome_v38 = "chrX"
reference_haplotype_name = "*1"
haplotypes: FrozenSet[Haplotype] = frozenset()
drugs: FrozenSet[DrugInfo] = frozenset()
rs_id_to_ref_seq_difference_annotation: Dict[str, Annotation] = dict()
rs_id_info1 = RsIdInfo(
"rs294924",
"AT",
"AT",
GeneCoordinate(chromosome_v37, 499593),
GeneCoordinate(chromosome_v38, 399483),
)
rs_id_info2 = RsIdInfo(
"rs3949923",
"C",
"C",
GeneCoordinate(chromosome_v37, 293993),
GeneCoordinate(chromosome_v38, 1388323),
)
rs_id_info3 = RsIdInfo(
"rs12993",
"GG",
"GG",
GeneCoordinate(chromosome_v37, 499592),
GeneCoordinate(chromosome_v38, 399482),
)
rs_id_infos1 = frozenset([rs_id_info1])
rs_id_infos2 = frozenset([rs_id_info1, rs_id_info2])
rs_id_infos3 = frozenset([rs_id_info3])
gene_info1 = GeneInfo(
gene1,
reference_haplotype_name,
haplotypes,
rs_id_infos1,
drugs,
rs_id_to_ref_seq_difference_annotation,
)
gene_info2 = GeneInfo(
gene2,
reference_haplotype_name,
haplotypes,
rs_id_infos2,
drugs,
rs_id_to_ref_seq_difference_annotation,
)
gene_info3 = GeneInfo(
gene2,
reference_haplotype_name,
haplotypes,
rs_id_infos3,
drugs,
rs_id_to_ref_seq_difference_annotation,
)
Panel(name, version, frozenset([gene_info1, gene_info2]))
with self.assertRaises(ValueError):
Panel(name, version, frozenset([gene_info1, gene_info3]))
def test_load_panel(self) -> None:
"""Load panel from json"""
panel_path = get_test_resource("test_panel.json")
panel = load_panel(str(panel_path))
dpyd_two_a_variant = Variant("rs3918290", "T")
dpyd_two_b_variant = Variant("rs1801159", "C")
dpyd_three_variant = Variant("rs72549303", "TG")
fake_variant = Variant("rs1212125", "C")
fake2_variant = Variant("rs1212127", "C")
dpyd_haplotypes_expected = frozenset({
Haplotype("*2A", "No Function", frozenset({dpyd_two_a_variant})),
Haplotype("*2B", "No Function",
frozenset({dpyd_two_a_variant, dpyd_two_b_variant})),
Haplotype("*3", "Normal Function",
frozenset({dpyd_three_variant})),
})
dpyd_rs_id_infos_expected = frozenset({
RsIdInfo("rs3918290", "C", "C", GeneCoordinate("1", 97915614),
GeneCoordinate("chr1", 97450058)),
RsIdInfo("rs72549309", "GATGA", "GATGA",
GeneCoordinate("1", 98205966),
GeneCoordinate("chr1", 97740410)),
RsIdInfo("rs1801159", "T", "T", GeneCoordinate("1", 97981395),
GeneCoordinate("chr1", 97515839)),
RsIdInfo("rs72549303", "TG", "TC", GeneCoordinate("1", 97915621),
GeneCoordinate("chr1", 97450065)),
RsIdInfo("rs1801265", "G", "A", GeneCoordinate("1", 98348885),
GeneCoordinate("chr1", 97883329)),
})
dpyd_drugs_expected = frozenset({
DrugInfo("5-Fluorouracil",
"https://www.source_url.org/5-Fluorouracil"),
DrugInfo("Capecitabine",
"https://www.source_url.org/Capecitabine"),
})
dpyd_rs_id_to_difference_annotations = {
"rs72549303": Annotation("6744CA>GA", "6744GA>CA"),
"rs1801265": Annotation("85C>T", "85T>C"),
}
fake_haplotypes_expected = frozenset({
Haplotype("*4A", "Reduced Function", frozenset({fake_variant})),
})
fake_rs_id_infos_expected = frozenset({
RsIdInfo("rs1212125", "T", "T", GeneCoordinate("5", 97915617),
GeneCoordinate("chr5", 97450060)),
})
fake_drugs_expected = frozenset({
DrugInfo("Aspirin", "https://www.source_url.org/Aspirin"),
})
fake_rs_id_to_difference_annotations: Dict[str, Annotation] = {}
fake2_haplotypes_expected = frozenset({
Haplotype("*4A", "Reduced Function", frozenset({fake2_variant})),
})
fake2_rs_id_infos_expected = frozenset({
RsIdInfo("rs1212127", "C", "T", GeneCoordinate("16", 97915617),
GeneCoordinate("chr16", 97450060)),
})
fake2_drugs_expected = frozenset({
DrugInfo("Aspirin", "https://www.source_url.org/Aspirin"),
})
fake2_rs_id_to_difference_annotations = {
"rs1212127": Annotation("1324C>T", "1324T>C")
}
gene_infos_expected = frozenset({
GeneInfo("DPYD", "*1", dpyd_haplotypes_expected,
dpyd_rs_id_infos_expected, dpyd_drugs_expected,
dpyd_rs_id_to_difference_annotations),
GeneInfo("FAKE", "*1", fake_haplotypes_expected,
fake_rs_id_infos_expected, fake_drugs_expected,
fake_rs_id_to_difference_annotations),
GeneInfo("FAKE2", "*1", fake2_haplotypes_expected,
fake2_rs_id_infos_expected, fake2_drugs_expected,
fake2_rs_id_to_difference_annotations),
})
name_expected = "fake_panel"
version_expected = "0.3"
panel_expected = Panel(name_expected, version_expected,
gene_infos_expected)
self.assertEqual(panel_expected, panel)