Start a new topic

Many theoretical pi results for target http://purl.uniprot.org/uniprot/Q9Y5Y9

https://explorer.openphacts.org/targets?uri=http%3A%2F%2Fpurl.uniprot.org%2Funiprot%2FQ9Y5Y9 has lots of theoretical pi results for http://bio2rdf.org/drugbank:BE0004901 . The API states that theoretical pi should only have 1 response but this one has an array. Is it right or wrong? There are also many cellular locations when the API states there should be only one.

1 Comment

If you use this uniprot url in the mapURL API call, you retrieve two drugbank targets: http://identifiers.org/drugbankv4.target/BE0000177 and http://identifiers.org/drugbankv4.target/BE0004901. The second one is a protein target group. So it seems the API returns the values for each of the proteins in this target group. That's why you have two main items for the drugbank dataset, and then one of them returns several values. I don't think the docs were written with the assumption that several targets could be returned here.
What we could do is to remove all mappings to these protein target groups from the IMS by default. Probably people would want to get the result only for the protein they specifically asked for, and not information on related proteins. Any other opinions on this?

 

Login or Signup to post a comment