I'm using edgeR to perform a a RNA-seq differential expression analysis with S. cerevisiae samples and then trying to use goana. My genes ids are from ensembl so I added a column with entrezgene ids to the DGELRT object, which is coming from the glmQLFTest function. I'm not sure where in the DGLRT object I have to add the column with entrezgene ids, in the manual says:

...or the name of the column of de$genes containing the Gene IDs.

But I cannot find which element in the DGELRT object is de$genes, so I added to de$table and I'm passing that to goana. I also tried to add the entregene ids to dgelist_qltest$geneid but goana fails in the same way

> class(dgelist_qltest)
[1] "DGELRT"
attr(,"package")
[1] "edgeR"

> goana.DGELRT(de=dgelist_qltest,  geneid=dgelist_qltest$table$entrezgene_id)
Error in goana.default(de = DEGenes, universe = universe, ...) : 
  No annotated genes found in universe 

> head(dgelist_qltest$table)
             logFC    logCPM          F       PValue entrezgene_id
YAL001C  0.2240038  6.943048 14.1162483 0.0022238429        851262
YAL002W  0.2994297  6.364025 18.9983099 0.0006987048        851261
YAL003W -0.3346512 10.272637 10.8792050 0.0054613104        851260
YAL004W  0.1465868  7.660092  0.8873887 0.3657633848            NA
YAL005C  0.2767830 11.574323  4.7249463 0.0479133341        851259
YAL007C -0.2657464  7.172642  8.8780937 0.0102093078        851226

#When I pass this to goana, fails with the same error: 
> head(dgelist_qltest$geneid)
[1] 851262 851261 851260     NA 851259 851226

Related to this, when I use goana with (or without) the geneid parameter and species=... I get this error:

> goana(de=dgelist_qltest,  geneid=dgelist_qltest$table$entrezgene_id, species='Sc')
Error in goana.default(de = DEGenes, universe = universe, ...) : 
  org.Sc.eg.db package required but not not installed (or can't be loaded)

Where the name of the db isn't found because in the case of S. cerevisiae is org.Sc.sgd.db , which does have "sgd" instead of the "eg" in the name.

Your DGELRT object dgelist_qltest is a list, with a genes item that you can put things in. In general when you make the DGEList to begin with, you would put all the annotation in at that point. You could do something like

## here I am assuming that your counts were in an object called 'counts'
## with SGD IDs as the row names, which seems to be a reasonable assumption
annot <- select(org.Sc.sgd.db, row.names(counts), "ENTREZID")

dgelist <- DGEList(counts, genes = annot)

And then in your call to goana you would just use the argument geneid = "ENTREZID"

What the error messages mean

It is good that you are reading the goana help page. You need to read the help pages ?goana.DGELRT and ?goana in conjunction. I will make some comments to try to help you interpret R help pages like this one.

goana.DGELRT(de=dgelist_qltest,

You never need to use goana.DGELRT() explicitly. Just use goana(dgelist_qltest, and the goana.DGELRT() function will be called automatically because of the class of dgelist_qltest.

No annotated genes found in universe

Well, this is because you tried to do a GO analysis for homo sapiens. Look at the usage line on the help page and you will see that the default setting for species is species = Hs. None of the row names in your object are human Entrez Gene Ids, hence the message.

I cannot find which element in the DGELRT object is de$genes

The help page tells you that de is a name for the first argument entered to goana and you have written in your own code that de=dgelist_qltest. So de$genes is dgelist_qltest$genes.

If dgelist_qltest$genes already exists, then you can add the Entrez Gene Ids by

dgelist_qltest$genes$GeneID <- geneid

or similar. If it doesn't exist, then you can add it by

dgelist_qltest$genes <- data.frame(GeneID=geneid)

This is just base R programming and follows from the fact that dge_qltest is a list.

so I added to de$table

Well, that won't work. You have to do what the help pages says.

org.Sc.eg.db package required but not not installed

As James as explained, goana only works for species for which a Bioconductor Entrez Id based organism package exists. goana can't use org.Sc.sgd.db.

Work around

As James has advised, a good solution is to use the goseq package instead of goana for this species.

Alternatively, you can always used kegga instead of goana for non-standard species. In this case you can use:

library("org.Sc.sgd.db")
ORF.GO <- toTable(org.Sc.sgdGO2ALLORFS)
library(GO.db)
GO.Name <- toTable(GOTERM)
GO.Name <- GO.Name[, c("go_id","Term")]
k <- kegga(dgelist_qltest, gene.pathway=ORF.GO, pathway.names=GO.Name)
topKEGG(k)

and the results will be exactly as if goana supported this species. Note that you don't need to convert geneids in this approach.