topGO factor with 2 levels

[tinguzman]

Hi,

I'm trying to do GO enrichment analysis using topgo, but i've encountered a problem, read below. what does factor w/ level 1 mean? it says that i should have 2 level factors.

geneNames contains my universal set of genes with GO annotation, then myInterestingGenes contains list of differentially expressed genes.


> myInterestingGenes <- as.vector(AvsB$genes)
> geneList <- factor(as.integer(geneNames %in% myInterestingGenes))
> names(geneList) <- geneNames
> names
function (x) .Primitive("names")
> str(geneList)
Factor w/ 1 level "0": 1 1 1 1 1 1 1 1 1 1 ...
- attr(*, "names")= chr [1:18756] "Genes" "c10020_g1" "c10035_g1" "c10052_g1" ...


Error in .local(.Object, ...) : allGenes must be a factor with 2 levels


how do i fix this?


Thanks

[dpryan]

"geneList" is just a bunch of 1s, which means that everything is differentially expressed. What topGO and similar programs expect is a sequence of 0s and 1s, which is then two levels (a sequence of 1s being just one level/value). Typically a value of 0 denotes "not differentially expressed" and 1 "differentially expressed", though you'll have to double check the topGO documentation to ensure it's not the other way around.