GO analysis

[mparida]

Hi All
I am trying to choose between the following:
a) using all the genes in my genome as a background for my GO analysis
or
b) using only those genes that have at least 1 GO term as a background for my GO analysis

Option a) gives me more (32) enriched GO terms when compared to a set of differentially expressed genes and enhanced q-values
option b) gives me less (18) enriched GO terms when compared to a set of differentially expressed genes and relatively higher q-values than option a.

Any help is appreciated. FYI, I am using GOSEQ for my analysis.

[N311V]

Personally I think option a) is easier to interpret.

Given the genome as a background, there are 32 enriched GO terms in this set of differentially expressed genes.

If you've applied an expression threshold perhaps the background could be expressed genes.