Open wshayes opened 6 years ago
Are you bringing in any chemicals (abundances) via MESH? I think these were the reason we needed the string match for equivalencing.
Good point - I am. tools/namespaces/mesh.py also needs to have equivalences processed for abundances.
From Natalie Looking here - https://github.com/OpenBEL/resource-generator/blob/master/rdf.py - in the resource generator code, it appears that I was only building the string match equivalences for abundances and biological processes. The function used was “get_close_matches”. We used a different skos relationship (closeMatch) for these than for the other equivlences (exactMatch, except when it was not a 1:1).
Abundance namespaces (ChEBI and ChEMBL) have equivalences (ChEMBL -> ChEBI and from both to INCHI_KEY).
Biological processes need to be equivalenced - the old OpenBEL RDF approach was to use lowercased names/synonyms and do exact string-match on them.