Mapping Genes To Metabolites

Question

[alexhli]

If I have a large list of human genes, and a large list of human metabolites (with KEGG and HMDB IDs), what is the best way to determine if any of these genes and metabolites have a known biological relationship?

i.e., suppose metabolite XXX is a byproduct of fatty acid metabolism, which gene YYY is involved in. That would be interesting and I would like a way to flag these relationships.

I am certain all the information I need is in an existing database, I just am not sure how to go about such a query. These data are all currently stored as flat files.

Thanks!

Answer 1

[aidan-budd]

Hi alexhli

You may find what you need in STITCH, a resource put together to study interactions between metabolites and genes/proteins. It's a bit like STRING with chemicals.

You can query it with a list of proteins and a list of metabolites, although not sure which IDs work in there (although they have a good synonym mapper as part of the software, so I hope/expect they're pretty good at matching many of the IDs out there

[alexhli]

STITCH looks promising, I only played with it for a few minutes and I like the figures it produces. I will need to try out the download version to see if I can batch query my data. Thanks

[aidan-budd]

there are a bunch of different ways you can interact with STITCH; for example, you can use the "multiple names" query tab to upload a long(ish) list of gene and chemical names, in which case you get a view of the interactions between the entities you upload (and to other entities not included in your list). If you want to automate your queries, then you may be able to get help/advice from the developers about how to do it - you can also download the entire dataset and play with it on your own local machine, if that'd help