From Ensembl Transcript Id To Go Term(S), Is There A Mapping?

Question

[Rvosa]

Dear BioStar,

I have a bunch of sequences from Ensembl, with FASTA definition lines like so:

>ENSP00000366697 pep:known chromosome:GRCh37:9:67926761:67969840:1 gene:ENSG00000196774 transcript:ENST00000377477

When I BLAST this sequence in Amigo it returns a record that has the transcript ID as one of its external references and that shows a GO term association:

http://amigo.geneontology.org/cgi-bin/amigo/gp-details.cgi?gp=UniProtKB:Q5TYW2&session_id=7161amigo1295567345

So now I'm looking for that black box where I can stick a transcript ID ("ENST00000377477") into it, and out come the GO term identifiers ("0005886", in this case). Is there an association file? It would be ideal if I can grep against such a file, because I have quite a few of these transcript IDs.

Thanks!

[Neilfws]

Very similar to a previous question How Do I Do Simple Go Term Lookup Given A Gene (Or Mrna) Identifier?

Answer 1

[Neilfws]

BioMart will do this.

Brief usage: select database (Ensembl) and dataset (human genes). Choose "Ensembl Transcript ID" as the filter and either paste or upload a file of IDs. Under "Attributes", choose the required "GO Term accessions" (CC/BP/MF). Then hit the "Results" button.

Here's a screenshot of the result:

There's also programmatic access to BioMart using e.g. biomaRt from Bioconductor. For more details of both, search this site: there are quite a few answers explaining usage in more detail (e.g. this one).

[Rvosa]

Yup, that's the ticket. Thanks for helping me find my black box so quickly. I kinda figured the answer would be BioMart, though I'd never used it. Sweet.

[Giulietta - Ensembl Helpdesk]

In case you have trouble using BioMart, Ensembl provides a tutorial video (it's a short one):

http://www.ensembl.org/Help/Movie?id=189

These types of questions can be directed to: [e-mail]