[BioC] GOStats GOHyperG question

[Hayes, Neil (David)]

I am interested in using the function GOhyperG in the package GOStats but
have the following challenge.

I am using data from a u95av2 chip, but only about 25% of the genes are
candidates for my analysis.  Starting with these 3000 genes, I select those
associated with my outcome of interest.  To help with the interpretation of
a relatively long list, I want to use function GOHyperG, as shown below.

GOHyperG(GeneList, lib="hgu95av2", what="MF")

I suspect I need to limit the GO terms I pass into the function to those
associated with the 25% of genes that where the starting point of my
analysis, not the full set for the 95av2 chip.  I also suspect that this is
not that hard for those who are well versed in manipulating the "affy chip
packages."  I have not been able to do this however.

Is there an obvious easy solution that I am missing, or do I need to make an
annotation package for my starting gene set from scratch?

Neil Hayes


********************** 
Confidentiality Notice\ **********************\      The inf...{{dropped}}