Some knowledge representations (e.g. simple RDF statements) do not tag their concepts (URI resources) with category. It may be a helpful heuristic to be able to resolve possible category values for URI's (CURIE's) based on their namespace listing under the id_prefixes slot of Biolink categoryclasses. Of course, if the mappings of id_prefixes is one-namespace-to-many, this heuristic may occasionally fail. Nonetheless, if a given namespace is associated with more than one category, one could at least return the list of all such category classes then leave it to the caller to decide how to capture use them in the tagged concept they are processing.
Whether or not only the most specific category class(es) are returned or the parent classes as well, is left as a separate decision (besides, BMT has a distinct 'ancester' tracing solution for that latter use case...)
Some knowledge representations (e.g. simple RDF statements) do not tag their concepts (URI resources) with
category
. It may be a helpful heuristic to be able to resolve possible category values for URI's (CURIE's) based on their namespace listing under theid_prefixes
slot of Biolinkcategory
classes. Of course, if the mappings ofid_prefixes
is one-namespace-to-many, this heuristic may occasionally fail. Nonetheless, if a given namespace is associated with more than onecategory
, one could at least return the list of all suchcategory
classes then leave it to the caller to decide how to capture use them in the tagged concept they are processing.Whether or not only the most specific
category
class(es) are returned or the parent classes as well, is left as a separate decision (besides, BMT has a distinct 'ancester' tracing solution for that latter use case...)