Entering edit mode
I'm going to perform GO enrichment test on my data set with topGO package. But i have some number of obsolete GO terms both in my gene universe and the list of genes of interest. Will they are just invisible for the software or they will somehow contribute to the analysis?
If I am not mistaken, topGO uses a named vector of p-values as input, where the names are some ID. As an example, the example data are like this:
> head(geneList)
1095_s_at 1130_at 1196_at 1329_s_at 1340_s_at 1342_g_at
1.0000000 1.0000000 0.6223795 0.5412240 1.0000000 1.0000000
And the GO mappings come from the GO.db package. So where are your obsolete GO terms coming from?
Hi, thanks for your response! From your example i can guess you deal with some DE experiment data and model organism. But both are not the case for me. Please check the "4.3 Custom annotations" and "4.4 Predefined list of interesting genes" sections of topGO manual. That's what i do. I checked the total GOs amount i have in current version of annotation of my organism by means of GO.db package and found that some part out of them is obsolete. This why i worry about that fact.
In that case, have you read the second to last paragraph in section 4.4? It seems to answer your question, doesn't it?
"One important point to notice is that not all the genes that are provided by geneList, the initial gene universe, can be annotated to the GO. This can be seen by comparing the number of all available genes, the genes present in geneList, with the number of feasible genes. We are therefore forced at this point to restrict the gene universe to the set of feasible genes for the rest of the analysis."
Yes, of course, i've read the mentioned above section and even posted a couple of related questions a while ago (A: Understanding of "feasible genes" conception and Absence of GO terms for some genes of interest in topGO package). Since that time, for me was clear that genes without GO annotation at all are invisible for the software, but still there was some hope that topGO somehow could handle genes that have GO terms, but they are obsolete ones. Now it's seems clear, that unfortunately it is not the case. Thank you for your assistance.