Mapping Custom Agilent Gene Ids To Go Terms

Question

[K. Wyatt Mcmahon]

Hello everyone,

I always come back to BioStar because the help is always useful and quick and the people are always so considerate. I'm sure this experience will be no different.

I am performing an analysis on someone's custom Agilent array design. I am trying to map each /many of the genes on the chip to a GO term (if such an annotation exists). This would be simple if all of the IDs were from one source (ENSEMBL, Genbank, etc). Instead, what I have is a mix of different identifier types for the chip.

I have gone to BioMart and - while hugely powerful - it seems to require that all of the IDs come from the same source. I can separate these out, but I'm wondering if there is a site that can recognize the type of identifier you have and then bring GO terms for you.

I do not have the files used to generate the arrays. I do, however, have the ProbeIDs, SystematicIDs, etc. from the array intensity files.

Thanks in advance for any help you can offer!

Wyatt

Answer 1

[Larry_Parnell]

Couldn't you relate each data source to a standard, unique identifier such as the EntrezGene ID? This would involve querying BioMart several times - once for each data type - but the results, taken together, would get the job done.

Or, because this is a custom array, the person/group who supplied Agilent with the data to build the array may even have this EntrezGene ID info. Perhaps not, and then that becomes a teaching point for them rather then, oh, Bioinformatics will just fix the problem.

[K. Wyatt Mcmahon]

Thanks for your advice, Larry. I'm thinking this is what I'll have to do; I was just wondering if there was some service that everyone knew about that I'm missing out on.

Answer 2

[Chris Evelo]

You could try [?]BridgeDB[?] (install locally as a webservice) the standard database provided has Agilent gene ID's (which in fact are contained in ENSEMBL as well, because that is where we have that DB from).

[Chris Evelo]

Ha, no, you are right. You may still want to use BridgeDB but you really need to stack a table with your custom IDs linked to some known ID on top of the database or mapping service you want to use. BridgeDB can do that kind of stacking.

[K. Wyatt Mcmahon]

Thanks for the suggestion, Chris. I actually looked into BridgeDB, and may still go that way. The problem is that this is a custom Agilent array, so it won't have the IDs from this array in BridgeDB.

[K. Wyatt Mcmahon]

Ah! I did not realize that it could do that! I'll look a little more into BridgeDB. Thanks, Chris!