fix!: ConceptMapping.coding.code should be a CURIE
#390
Conversation
close #383 * Use `concept_id` as the CURIE
| code_ = concept_id.upper() if source == NamespacePrefix.HGNC else source_id | ||
|
|
||
| return ConceptMapping( | ||
| coding=Coding(code=code(code_), system=system), relation=relation | ||
| coding=Coding(code=code(concept_id), system=system), relation=relation |
There was a problem hiding this comment.
@jsstevenson is it fine to just use the concept_id as the CURIE? Technically, HGNC uses uppercase. I might be overthinking things with mappings
There was a problem hiding this comment.
Oh, maybe we just need to look at updating our NamespacePrefix member values in the future?
There was a problem hiding this comment.
Yeah, I think that'd be simplest.
Realistically there is cannot be a "true" CURIE prefix for a given source (identifiers.org was supposed to be the solution, but then it has its own problems so people made the bioregistry, but it has problems and so on and so on), which is why Matt was talking about some kind of local prefix expansions document. Some sources obviously have their own personal opinions about what prefixes you use for them, but then it begs the question of what to do if there's a community standard which differs from that. We should definitely look at updating our own prefix mapping where it makes sense though.
There was a problem hiding this comment.
@jsstevenson Ah okay, maybe we wait for Matt (and others) to provide guidance then and update our NamespacePrefix values whenever that discussion happens?
close #383
concept_idas the CURIE