How do you determine which a gene annotation is correct?

[blairw]

Using the program Artemis, I have to get the annotation of one specific gene. I then have to do a blast search on other databases (Ecocyc, KEGG, etc.) and compare those results to Artemis (to see if the Artemis annotation is correct). If the databases do not give the same annotations for that gene, how do I decide which is correct (Artemis vs. the others)?

Sorry if I used any of the incorrect terms, I'm an undergrad and a lot of this is new to me.

[mastal]

I'm not sure that you can determine which annotation is 'correct' without doing loads of additional work.

You would have to consider the source of the annotations, and which you would consider to be more reliable - whether it is from a manually curated database, or from a computational gene-prediction pipeline, or a mixture of both, for example.

[GenoMax]

It is always better to start with protein sequence when you are working with an unknown gene sequence. Start with Swissprot/Uniprot/RefSeq (http://www.uniprot.org/blast/) to see if you are able to find a conserved domain (or something that looks like one) even if the rest of the sequence match is not great.