search

bridgedb / create-bridgedb-metabolites

Create BridgeDb identity mapping files from HMDB, ChEBI, and Wikidata
Other
4 stars 4 forks source link

Add Protein IDs for hormones? #23

Open DeniseSl22 opened 5 years ago

DeniseSl22 commented 5 years ago

@Chris-Evelo @egonw ;

Since several hormones/polypeptides are in the border region of being called a metabolite or a protein, we could think about including protein IDs for these types of compounds in the metabolite mapping files (derived from Wikidata, even though I also saw some examples in ChEBI linking to Uniprot).

What do you think?

Example: hGC wikidata and wikipedia; currently very relevant for me ;)

Chris-Evelo commented 5 years ago

I am not sure. Yes, you are right these things are gene products and thus will typically feature in the gene-product database as well. To make it even more confusing they could in fact be annotated in one pathway as a metabolite and in another as a gene product. In fact, we typically use insulin as the example to describe that problem, but you are right, there are many more. The question is how we would deal with that problem in real life. Would you detect the hormonal activity using a biochemistry assay and then combine that with gene expression regulation studie for instance? My gut feeling is that this would be a case where using the stacking feature of BridgeDb would work best. Have a separate, small, database that links a core metabolite ID to a protein ID and thus allow people to move from one world to another. But I would really like to see a real-world research question first and see what would be needed to solve that. Bioinformaticians have sometimes been identified as the people that solve problems that nobody has. And I would really like to prevent that here.