Hi, I have a dataset from RNAseq. I want to compared differential expression on GO terms. Expression of each GO terms is sum of gene expression in this GO. Can I use edgeR to compare GO expression between two conditions?

Thanks!

UPDATES:

Thank you all for suggestions and comments.

Sum of gene expression in a GO term results from HUMAnN2. I think it still provides an aspect on expression of cellular function. I am comparing expression between two conditions not comparing GO terms within each condition.

I was trying to use deferential expressed genes to do GO enrichment analysis. But my project is about metatranscriptome. There are many different known and unknown organisms. Is it still possible to do GO enrichment?

I don't think the "sum of gene expression" for all genes in a GO term is a meaningful measure of anything, especially if you are summing raw counts. So I would not recommend this approach. Just use the standard gene set testing approaches already available in Bioconductor.

You can use camera() or fry(), both of which do more or less what you want.

Suppose you have human data. If y is your DGEList and design is the design matrix, then you might do:

index <- as.list(org.Hs.eg.db:::org.Hs.egGO2ALLEGS)
ca <- camera(y, index, design, contrast=whatever)

or

fr <- fry(y, index, design, contrast=whatever)

You may use gage package for GO and pathway analysis.

http://bioconductor.org/packages/release/bioc/html/gage.html
http://bioconductor.org/packages/release/bioc/vignettes/gage/inst/doc/RNA-seqWorkflow.pdf