Open alanbridge opened 2 years ago
Thanks Alan for the suggestion. The UniProt cross-references in ChEBI have not been updated for a few years now and its something we definitely need to update when we redevelop the ChEBI infrastructure. I will forward your suggestion to the new ChEBI developers and they can get in touch with you in due course when they update the automatic cross-reference section as part of this redevelopment. Any other suggestions for improvements are also very welcome.
Hi ChEBI,
ChEBI provides links to UniProt in the section "Automatic Xrefs". Would you consider changing the way these are generated, in order to make use of ChEBI annotations in UniProt?
As far as I know, these links are currently provided using some kind of automated text mining procedure. For example, the link
https://www.ebi.ac.uk/chebi/displayAutoXrefs.do?chebiId=CHEBI:29033
gives access to 4396 UniProtKB sequences.
We are now using ChEBI to extenstively annotate UniProt records (in our next release this will also cover the ligand binding features). You could therefore do a more precise search in UniProt, using ChEBI identifiers, e.g.
https://www.uniprot.org/uniprot/?query=CHEBI%3A29033+&sort=score
that gives over 2.4 million protein sequences from UniProt, of which 5,546 are found in Swiss-Prot.
Thanks for considering this alternative approach,
All the best, Alan