def __init__(self,
gene_id: str,
module: Module,
data_manager: DataManager,
config: GenedescConfigParser,
limit_to_group: str = None,
humans: bool = False):
"""initialize sentence generator object
Args:
config (GenedescConfigParser): an optional config object from which to read the options
limit_to_group (str): limit the evidence codes to the specified group
"""
self.ontology = data_manager.get_ontology(
get_data_type_from_module(module))
self.config = config
self.module = module
self.terms_already_covered = set()
self.terms_groups = defaultdict(lambda: defaultdict(set))
self.evidence_groups_priority_list = config.get_evidence_groups_priority_list(
module=module)
self.prepostfix_sentences_map = config.get_prepostfix_sentence_map(
module=module, humans=humans)
self.gene_annots = data_manager.get_annotations_for_gene(
gene_id=gene_id,
annot_type=get_data_type_from_module(module),
priority_list=config.get_annotations_priority(module=module))
self.trimmer = CONF_TO_TRIMMING_CLASS[config.get_module_property(
module=module, prop=ConfigModuleProperty.TRIMMING_ALGORITHM)](
ontology=self.ontology,
annotations=data_manager.get_associations(
get_data_type_from_module(module)),
nodeids_blacklist=config.get_module_property(
module=module, prop=ConfigModuleProperty.EXCLUDE_TERMS),
slim_terms_ic_bonus_perc=config.get_module_property(
module=module, prop=ConfigModuleProperty.SLIM_BONUS_PERC),
slim_set=data_manager.get_slim(module=module))
self.set_terms_groups(module, config, limit_to_group, humans)
def __init__(self,
gene_id: str,
module: Module,
data_manager: DataManager,
config: GenedescConfigParser,
limit_to_group: str = None,
humans: bool = False):
"""initialize sentence generator object
Args:
config (GenedescConfigParser): an optional config object from which to read the options
limit_to_group (str): limit the evidence codes to the specified group
"""
annot_type = None
if module == Module.DO_ORTHOLOGY or module == Module.DO_EXPERIMENTAL or module == module.DO_BIOMARKER:
self.ontology = data_manager.do_ontology
annot_type = DataType.DO
elif module == Module.GO:
self.ontology = data_manager.go_ontology
annot_type = DataType.GO
elif module == Module.EXPRESSION:
self.ontology = data_manager.expression_ontology
annot_type = DataType.EXPR
self.evidence_groups_priority_list = config.get_evidence_groups_priority_list(
module=module)
self.prepostfix_sentences_map = config.get_prepostfix_sentence_map(
module=module, humans=humans)
self.terms_groups = defaultdict(lambda: defaultdict(set))
ev_codes_groups_maps = config.get_evidence_codes_groups_map(
module=module)
annotations = data_manager.get_annotations_for_gene(
gene_id=gene_id,
annot_type=annot_type,
priority_list=config.get_annotations_priority(module=module))
self.annotations = annotations
self.module = module
self.data_manager = data_manager
self.annot_type = annot_type
evidence_codes_groups_map = {
evcode: group
for evcode, group in ev_codes_groups_maps.items()
if limit_to_group is None
or limit_to_group in ev_codes_groups_maps[evcode]
}
prepostfix_special_cases_sent_map = config.get_prepostfix_sentence_map(
module=module, special_cases_only=True, humans=humans)
if len(annotations) > 0:
for annotation in annotations:
if annotation["evidence"]["type"] in evidence_codes_groups_map:
aspect = annotation["aspect"]
ev_group = evidence_codes_groups_map[annotation["evidence"]
["type"]]
qualifier = "_".join(
sorted(annotation["qualifiers"]
)) if "qualifiers" in annotation else ""
if prepostfix_special_cases_sent_map and (aspect, ev_group, qualifier) in \
prepostfix_special_cases_sent_map:
for special_case in prepostfix_special_cases_sent_map[(
aspect, ev_group, qualifier)]:
if re.match(
re.escape(special_case[1]),
self.ontology.label(
annotation["object"]["id"],
id_if_null=True)):
ev_group = evidence_codes_groups_map[annotation["evidence"]["type"]] + \
str(special_case[0])
if ev_group not in self.evidence_groups_priority_list:
self.evidence_groups_priority_list.insert(
self.evidence_groups_priority_list.
index(evidence_codes_groups_map[
annotation["evidence"]["type"]]) +
1, ev_group)
break
self.terms_groups[(aspect, qualifier)][ev_group].add(
annotation["object"]["id"])