GSEA for qPCR data

Question

[Juan Cordero]

I have computed the differentially expressed genes in a qPCR experiment with ~600 genes. Now I need to perform a gene set enrichment analysis (GO) but do not know which set of genes I should use as the whole set of genes: the 600 genes used the qPCR array or all the genes in the genome given by the annotation?

The model is R and I'm using bioconductor.

Answer 1

[WouterDeCoster]

You should limit the background/whole set to what was analyzed/assayed. If not you will heavily bias in favor of sets you already selected multiple genes from and bias against sets you only selected 1 or 0.