Dear all,
I have carried out differential expression of RNA-seq data from a non-model organism using the limma+voom pipeline. Next, I wish to look for functional enrichment of DE genes using GOseq.
I successfully used GOseq on my data set by following the GOseq vignette, using custom gene lengths and GO mappings. However, I note that following the vignette leads to all DE genes being tested for enrichment regardless of the direction of differential expression (i.e., regardless of whether upregulated or downregulated). I have seen suggestions that upregulated and downregulated genes should be analysed separately for functional enrichment (e.g., https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3899863/).
To investigate function enrichment of (e.g.) only upregulated genes in GOseq, would it be as simple as modifying the DE genes vector so that all upregulated genes have a value of 1, all non-DE genes have a value of 0, and all downregulated genes have a value of 0? Intuitively this seems like the solution, but I just wanted to check before introducing any unwitting biases/errors into my analyses.
Thanks!
