At a meeting with @balhoff some of the participants are using GO and are confused by the fact they are getting assigned_by=CAFA annotations for human etc. They googled CAFA and learn it is about prediction and are even more confused since this should presumably be non-experimental evidence, and all the CAFA annotations we have are IDA and other experimental types.
I believe the story is that these annotations were funded by CAFA and done by normal uniprot-goa curators, used for evaluation. We retain CAFA as assigned by to give credit to the CAFA group for supporting the curation.
The link is technically problematic as the https certificate has expired, but more problematically it doesn't really describe where the annotations came from.
I suggest that we change to URL to a link on our wiki that explains explicitly the curation process and/or we include a comment field (may require extending the schema) with text. @vanaukenk @pgaudet seem reasonable?
Also checking with @kltm @dougli1sqrd that these get loaded into the rdf store
At a meeting with @balhoff some of the participants are using GO and are confused by the fact they are getting assigned_by=CAFA annotations for human etc. They googled CAFA and learn it is about prediction and are even more confused since this should presumably be non-experimental evidence, and all the CAFA annotations we have are IDA and other experimental types.
I believe the story is that these annotations were funded by CAFA and done by normal uniprot-goa curators, used for evaluation. We retain CAFA as assigned by to give credit to the CAFA group for supporting the curation.
This is our current entry for CAFA https://github.com/geneontology/go-site/blob/1a70a5d7d62305a9cb9ea3fe60b441061ce3dc6c/metadata/groups.yaml#L317-L320
The link is technically problematic as the https certificate has expired, but more problematically it doesn't really describe where the annotations came from.
I suggest that we change to URL to a link on our wiki that explains explicitly the curation process and/or we include a comment field (may require extending the schema) with text. @vanaukenk @pgaudet seem reasonable?
Also checking with @kltm @dougli1sqrd that these get loaded into the rdf store