def make_annotation(db="blah",
db_id="blah12345",
db_obj_symb="blah",
qualifier="",
negated=False,
goid="GO:1234567",
references="BLAH:54321",
evidence="IDA",
withfrom="BLAH:12345",
aspect="C",
db_obj_name="",
db_obj_syn="",
db_obj_type="blah",
taxon="taxon:12345",
date="20200330",
assigned_by="blah",
extension="",
gene_form_id="",
properties="",
version=None,
from_gaf=True):
if from_gaf:
qual_parse = assocparser.Qualifier2_1()
if version == "2.2":
qual_parse = assocparser.Qualifier2_2()
annotation = [
db, db_id, db_obj_symb, qualifier, goid, references, evidence,
withfrom, aspect, db_obj_name, db_obj_syn, db_obj_type, taxon,
date, assigned_by, extension, gene_form_id
]
return gafparser.to_association(annotation,
qualifier_parser=qual_parse)
else:
if version != "2.0":
# Default to 1.2 if we're anything but 2.0
annotation = [
db, db_id, qualifier, goid, references, evidence, withfrom,
taxon, date, assigned_by, extension, properties
]
return gpadparser.to_association(annotation, version="1.2")
else:
annotation = [
"{}:{}".format(db, db_id), "NOT" if negated else "", qualifier,
goid, references, evidence, withfrom, taxon, date, assigned_by,
extension, properties
]
return gpadparser.to_association(annotation, version="2.0")
def to_association(gaf_line: List[str], report=None, group="unknown", dataset="unknown", qualifier_parser=assocparser.Qualifier2_1(), bio_entities=None) -> assocparser.ParseResult:
report = Report(group=group, dataset=dataset) if report is None else report
bio_entities = collections.BioEntities(dict()) if bio_entities is None else bio_entities
source_line = "\t".join(gaf_line)
if source_line == "":
report.error(source_line, "Blank Line", "EMPTY", "Blank lines are not allowed", rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
if len(gaf_line) > 17:
# If we see more than 17 columns, we will just cut off the columns after column 17
report.warning(source_line, assocparser.Report.WRONG_NUMBER_OF_COLUMNS, "",
msg="There were more than 17 columns in this line. Proceeding by cutting off extra columns after column 17.",
rule=1)
gaf_line = gaf_line[:17]
if 17 > len(gaf_line) >= 15:
gaf_line += [""] * (17 - len(gaf_line))
if len(gaf_line) != 17:
report.error(source_line, assocparser.Report.WRONG_NUMBER_OF_COLUMNS, "",
msg="There were {columns} columns found in this line, and there should be 15 (for GAF v1) or 17 (for GAF v2)".format(columns=len(gaf_line)), rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
## check for missing columns
## We use indeces here because we run GO RULES before we split the vals into individual variables
DB_INDEX = 0
DB_OBJECT_INDEX = 1
TAXON_INDEX = 12
REFERENCE_INDEX = 5
if gaf_line[DB_INDEX] == "":
report.error(source_line, Report.INVALID_IDSPACE, "EMPTY", "col1 is empty", taxon=gaf_line[TAXON_INDEX], rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
if gaf_line[DB_OBJECT_INDEX] == "":
report.error(source_line, Report.INVALID_ID, "EMPTY", "col2 is empty", taxon=gaf_line[TAXON_INDEX], rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
if gaf_line[REFERENCE_INDEX] == "":
report.error(source_line, Report.INVALID_ID, "EMPTY", "reference column 6 is empty", taxon=gaf_line[TAXON_INDEX], rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
parsed_taxons_result = gaf_line_validators["taxon"].validate(gaf_line[TAXON_INDEX]) # type: assocparser.ValidateResult
if not parsed_taxons_result.valid:
report.error(source_line, Report.INVALID_TAXON, parsed_taxons_result.original, parsed_taxons_result.message, taxon=parsed_taxons_result.original, rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
taxon = parsed_taxons_result.parsed[0]
date = assocparser.parse_date(gaf_line[13], report, source_line)
if date is None:
return assocparser.ParseResult(source_line, [], True, report=report)
interacting_taxon = parsed_taxons_result.parsed[1] if len(parsed_taxons_result.parsed) == 2 else None
subject_curie = association.Curie(gaf_line[0], gaf_line[1])
subject = association.Subject(subject_curie, gaf_line[2], [gaf_line[9]], gaf_line[10].split("|"), [association.map_gp_type_label_to_curie(gaf_line[11])], taxon)
gpi_entity = bio_entities.get(subject_curie)
if gpi_entity is not None and subject != gpi_entity:
subject = gpi_entity
# column 4 is qualifiers -> index 3
# For allowed, see http://geneontology.org/docs/go-annotations/#annotation-qualifiers
# We use the below validate to check validaty if qualifiers, not as much to *parse* them into the GoAssociation object.
# For GoAssociation we will use the above qualifiers list. This is fine because the above does not include `NOT`, etc
# This is confusing, and we can fix later on by consolidating qualifier and relation in GoAssociation.
parsed_qualifiers = qualifier_parser.validate(gaf_line[3])
if not parsed_qualifiers.valid:
report.error(source_line, Report.INVALID_QUALIFIER, parsed_qualifiers.original, parsed_qualifiers.message, taxon=gaf_line[TAXON_INDEX], rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
aspect = gaf_line[8]
negated, relation_label, qualifiers = assocparser._parse_qualifier(gaf_line[3], aspect)
# Note: Relation label is grabbed from qualifiers, if any exist in _parse_qualifier
qualifiers = [association.Curie.from_str(curie_util.contract_uri(relations.lookup_label(q), strict=False)[0]) for q in qualifiers]
object = association.Term(association.Curie.from_str(gaf_line[4]), taxon)
if isinstance(object, association.Error):
report.error(source_line, Report.INVALID_SYMBOL, gaf_line[4], "Problem parsing GO Term", taxon=gaf_line[TAXON_INDEX], rule=1)
# References
references = [association.Curie.from_str(e) for e in gaf_line[5].split("|") if e]
for r in references:
if isinstance(r, association.Error):
report.error(source_line, Report.INVALID_SYMBOL, gaf_line[5], "Problem parsing references", taxon=gaf_line[TAXON_INDEX], rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
gorefs = [ref for ref in references if ref.namespace == "GO_REF"] + [None]
eco_curie = ecomap.coderef_to_ecoclass(gaf_line[6], reference=gorefs[0])
if eco_curie is None:
report.error(source_line, Report.UNKNOWN_EVIDENCE_CLASS, gaf_line[6], msg="Expecting a known ECO GAF code, e.g ISS", rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
withfroms = association.ConjunctiveSet.str_to_conjunctions(gaf_line[7])
if isinstance(withfroms, association.Error):
report.error(source_line, Report.INVALID_SYMBOL, gaf_line[7], "Problem parsing with/from", taxon=gaf_line[TAXON_INDEX], rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
evidence_type = association.Curie.from_str(eco_curie)
if isinstance(evidence_type, association.Error):
report.error(source_line, Report.INVALID_SYMBOL, gaf_line[6], "Problem parsing evidence type", taxon=gaf_line[TAXON_INDEX], rule=1)
evidence = association.Evidence(association.Curie.from_str(eco_curie), references, withfroms)
if any([isinstance(e, association.Error) for e in evidence.has_supporting_reference]):
first_error = [e for e in evidence.has_supporting_reference if isinstance(e, association.Error)][0]
report.error(source_line, Report.INVALID_SYMBOL, gaf_line[5], first_error.info, taxon=str(taxon), rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
subject_extensions = []
if gaf_line[16]:
subject_filler = association.Curie.from_str(gaf_line[16])
if isinstance(subject_filler, association.Error):
report.error(source_line, assocparser.Report.INVALID_ID, gaf_line[16], subject_filler.info, taxon=str(taxon), rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
# filler is not an Error, so keep moving
subject_extensions.append(association.ExtensionUnit(association.Curie.from_str("rdfs:subClassOf"), subject_filler))
conjunctions = []
if gaf_line[15]:
conjunctions = association.ConjunctiveSet.str_to_conjunctions(
gaf_line[15],
conjunct_element_builder=lambda el: association.ExtensionUnit.from_str(el))
if isinstance(conjunctions, association.Error):
report.error(source_line, Report.EXTENSION_SYNTAX_ERROR, conjunctions.info, "extensions should be relation(curie) and relation should have corresponding URI", taxon=str(taxon), rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
relation_uri = relations.lookup_label(relation_label)
if relation_uri is None:
report.error(source_line, assocparser.Report.INVALID_QUALIFIER, relation_label, "Could not find CURIE for relation `{}`".format(relation_label), taxon=str(taxon), rule=1)
return assocparser.ParseResult(source_line, [], True, report=report)
# We don't have to check that this is well formed because we're grabbing it from the known relations URI map.
relation_curie = association.Curie.from_str(curie_util.contract_uri(relation_uri)[0])
a = association.GoAssociation(
source_line="\t".join(gaf_line),
subject=subject,
relation=relation_curie,
object=object,
negated=negated,
qualifiers=qualifiers,
aspect=aspect,
interacting_taxon=interacting_taxon,
evidence=evidence,
subject_extensions=subject_extensions,
object_extensions=conjunctions,
provided_by=gaf_line[14],
date=date,
properties={})
return assocparser.ParseResult(source_line, [a], False, report=report)
def qualifier_parser(self) -> assocparser.ColumnValidator:
if self.gaf_version() == "2.2":
return assocparser.Qualifier2_2()
return assocparser.Qualifier2_1()
from ontobio.rdfgen import relations
from ontobio.ontol import Ontology
import dateutil.parser
import functools
import click
logger = logging.getLogger(__name__)
gaf_line_validators = {
"default": assocparser.ColumnValidator(),
"qualifier2_1": assocparser.Qualifier2_1(),
"qualifier2_2": assocparser.Qualifier2_2(),
"curie": assocparser.CurieValidator(),
"taxon": assocparser.TaxonValidator()
}
def protein_complex_sublcass_closure(ontology: Ontology) -> Set[str]:
protein_containing_complex = association.Curie(namespace="GO", identity="0032991")
children_of_complexes = set(ontology.descendants(str(protein_containing_complex), relations=["subClassOf"], reflexive=True))
return children_of_complexes
class GafParser(assocparser.AssocParser):
"""
Parser for GO GAF format
