Gene Ontology -- Getting Complete Gene And Gene Ontology Term Associations

Question

[jsporter]

Hello, I want a way to generate a gene ontology term (GO term) and gene id association file. On the gene ontology website, I think I can get gene id and GO term associations, but only the most specific GO term is associated with the gene id. The web app Funcassociate2 (http://llama.mshri.on.ca/funcassociate_client/html/) allows downloading a file (from the Download Go Associations button) that has the same format.

However, what I am looking for is a file that gives all GO terms associated with a gene higher up in the hierarchy as well. That is, if a GO term "A" annotates a gene, then the file should give that all parents "A" annotate the gene as well. I'm looking for a tab separated file (or similar like XML) similar to the file format from Funcassociate2 that lists each of these GO terms and gene associations.

I know that I could write a program myself to do this with the information that I have, but what I really want to know is if there is an existing tool that will produce the file in the format that I am looking for. If a program could produce this information from a given set of genes, that'd be great.

Thanks.

Answer 1

[Richard Smith]

For human or mouse data you can use metabolicMine. The parent relationships are pre-calculated (I think they follow both is_a and part_of relationships in the ontology).

This pre-defined query will fetch GO terms for a gene and in the final column shows the parents of each term: http://metabolicmine.org/beta/template.do?name=Gene_GO. You can export in various formats.

If you click on the 'Lists' tab at the top you can upload and save a list of genes, if you return to the query page you will be able to use that list in the query and export a table for all genes at once.

Answer 2

[jsporter]

Thanks for the response. The metabolicmine.org website did what I wanted.

I found the program process-go from the Biorithim software package from T.M. Murali's lab that does the job as well. (http://bioinformatics.cs.vt.edu/~murali/software/biorithm/process-go.html)