GO analysis custom reference?

Question

[Pit]

Hi all,

I have done a branch selection test on a group of marine mammal genomes, and now I'm trying to run the list of genes with evidence of positive selection through GO analysis. The problem is that my candidate genomes cover a wide range of species (representatives from pinnipeds, cetaceans, and sirenians), and I'm having a hard time setting my reference for the GO analysis. What is the most appropriate way to go about this? If I need to construct a custom reference, how should I do that?

Thanks.

[dthorbur]

How are you doing the GO analysis? Is it per species?

An approach of identifying outlier genes in each genome, then conducting a GO analysis on each species' gene set as valid. Another option would be to find a set of orthologs, but if the gene sets are quite different among species, I struggle to see the biological relevance of this gene set.