We currently have mappings to UMLS CUIs for genes (see #18). For example, on the entry for NOS3 http://mygene.info/v3/gene/4846, we have:
"umls": {
"cui": "C1328053"
},
But since mygene effectively merges gene and protein objects, we should probably also add mappings to UMLS CUIs for proteins. For NOS3, that identifier would be https://uts.nlm.nih.gov/metathesaurus.html?cui=C0669365, which looks like it can be joined to our gene entries via Swiss-Prot identifiers in MRSAT.RRF (which I believe is a mapping file provided by UMLS)
(RefSeq also appears to be a common identifier that could be used for mapping...)
If that's all correct, then I'd suggest adding those mappings, perhaps in a structure like this:
We currently have mappings to UMLS CUIs for genes (see #18). For example, on the entry for NOS3 http://mygene.info/v3/gene/4846, we have:
But since mygene effectively merges gene and protein objects, we should probably also add mappings to UMLS CUIs for proteins. For NOS3, that identifier would be https://uts.nlm.nih.gov/metathesaurus.html?cui=C0669365, which looks like it can be joined to our gene entries via Swiss-Prot identifiers in MRSAT.RRF (which I believe is a mapping file provided by UMLS)
(RefSeq also appears to be a common identifier that could be used for mapping...)
If that's all correct, then I'd suggest adding those mappings, perhaps in a structure like this: