BUG: node normalization not handling case of one PubChem ID mapping to multiple UMLS IDs

[andrewsu]

In NCATSTranslator/testing#249, the input query used PUBCHEM.COMPOUND:23994 as the input ID. Node normalizer returns two UMLS IDs (UMLS:C0043481 and UMLS:C0795680), both of which should be used in queries by BTE. It looks like the PK posted in that issue (72895d13-60c5-497f-a635-a702c0b4d06f) only used the latter UMLS ID. This PK (f74cf91f-3853-4961-b29e-b9966b2966a3) is for issuing the same query explicitly using the former UMLS ID. The intended behavior is that BTE returns the union of those two PKs.

More results examples:

• query results when using PUBCHEM.COMPOUND:23994: https://api.bte.ncats.io/v1/check_query_status/j3tfOxsz3u
• query results when using UMLS:C0043481: https://api.bte.ncats.io/v1/check_query_status/7z8rNQqEoM
• query results when using UMLS:C0795680: https://api.bte.ncats.io/v1/check_query_status/12iWR8AplS

[tokebe]

This appears to be an issue between NodeNorms's Production and Development instances.

Querying the production server, UMLS:C0043481 is not returned.

Querying the development server, UMLS:C0043481 is returned.

BTE uses the Production server.

[andrewsu]

Thanks @tokebe for the more careful detective work! Closing this issue...