def test_parse_broken_qualifier(self):
# test if wrong separator between GO ID and description (semicolon instead of colon) is caught
broken_qualifier = [
"GO:0004871: signal transducer activity",
"GO:0007165; signal transduction"
]
with self.assertRaisesRegex(ValueError, "Cannot parse qualifier"):
GOQualifier.from_biopython(broken_qualifier)
def test_biopython_to_and_from(self):
original = GOQualifier({
'GO:0004871': 'signal transducer activity',
'GO:0007165': 'signal transduction',
'GO:0016020': 'membrane'
})
new = GOQualifier.from_biopython(original.to_biopython())
assert original.go_entries == new.go_entries
def from_biopython(bio_feature: SeqFeature, feature: "PFAMDomain" = None, # type: ignore
leftovers: Dict[str, List[str]] = None) -> "PFAMDomain":
if leftovers is None:
leftovers = Feature.make_qualifiers_copy(bio_feature)
# grab mandatory qualifiers and create the class
description = leftovers.pop("description")[0]
p_start = int(leftovers.pop("protein_start")[0])
p_end = int(leftovers.pop("protein_end")[0])
xref = leftovers.get("db_xref", []) # only remove the interesting part
name = None
for i, ref in enumerate(xref):
if ref.startswith("PF"):
name = ref
xref.pop(i)
break
if name is None:
raise SecmetInvalidInputError("PFAMDomain missing identifier")
tool = leftovers.pop("aSTool")[0]
feature = PFAMDomain(bio_feature.location, description, p_start, p_end,
identifier=name, tool=tool)
# grab optional qualifiers
feature.gene_ontologies = GOQualifier.from_biopython(leftovers.pop("gene_ontologies", []))
if "probability" in leftovers:
feature.probability = float(leftovers["probability"][0])
# grab parent optional qualifiers
updated = super(PFAMDomain, feature).from_biopython(bio_feature, feature=feature, leftovers=leftovers)
assert isinstance(updated, PFAMDomain)
return updated
def test_pfam_domain(self):
original = PFAMDomain(FeatureLocation(2, 5),
description="test",
protein_start=5,
protein_end=10,
identifier="PF00002.17",
domain="p450",
tool="toolname")
original.domain_id = "domain_id"
original.database = "db"
original.detection = "someprogram"
original.evalue = 1e-5
original.score = 5.
original.locus_tag = "locus"
original.label = "somelabel"
original.translation = "ARNDCQ"
original.gene_ontologies = GOQualifier({
'GO:0004871': 'signal transducer activity',
'GO:0007165': 'signal transduction',
'GO:0016020': 'membrane'
})
new = PFAMDomain.from_biopython(original.to_biopython()[0])
for slot in [
"tool", "domain_id", "database", "detection", "evalue",
"score", "locus_tag", "label", "translation", "domain",
"protein_start", "protein_end", "identifier", "version"
]:
assert getattr(original, slot) == getattr(new, slot)
assert original.gene_ontologies.go_entries == new.gene_ontologies.go_entries
assert original.full_identifier == new.full_identifier
def test_go_ids(self):
original_go_entries = {
'GO:0004871': 'signal transducer activity',
'GO:0007165': 'signal transduction',
'GO:0016020': 'membrane'
}
go_qualifier = GOQualifier(original_go_entries)
assert set(go_qualifier.ids) == set(original_go_entries)
def add_to_record(self, record: Record) -> None:
"""Add GeneOntologies objects to the respective PFAMDomains.
Arguments:
record: Record to which to add GeneOntologies
"""
if record.id != self.record_id:
raise ValueError("Record to store in and record analysed don't match")
for domain in self.pfam_domains_with_gos:
domain.gene_ontologies = GOQualifier(self.get_all_gos(domain))
def from_biopython(cls: Type[T],
bio_feature: SeqFeature,
feature: T = None,
leftovers: Dict[str, List[str]] = None,
record: Any = None) -> T:
if leftovers is None:
leftovers = Feature.make_qualifiers_copy(bio_feature)
# grab mandatory qualifiers and create the class
description = leftovers.pop("description")[0]
p_start = int(leftovers.pop("protein_start")[0])
p_end = int(leftovers.pop("protein_end")[0])
xref = leftovers.get("db_xref", []) # only remove the interesting part
name = None
for i, ref in enumerate(xref):
if ref.startswith("PF"):
name = ref
xref.pop(i)
break
if name is None:
raise SecmetInvalidInputError("PFAMDomain missing identifier")
tool = leftovers.pop("aSTool")[0]
locus_tag = leftovers.pop("locus_tag", ["(unknown)"])[0]
feature = cls(bio_feature.location,
description,
FeatureLocation(p_start, p_end),
identifier=name,
tool=tool,
locus_tag=locus_tag)
# grab optional qualifiers
feature.gene_ontologies = GOQualifier.from_biopython(
leftovers.pop("gene_ontologies", []))
# grab parent optional qualifiers
updated = super().from_biopython(bio_feature,
feature=feature,
leftovers=leftovers,
record=record)
assert isinstance(updated, PFAMDomain)
return updated
