Hi! We performed RNA-Seq on bacteria that carry multiple plasmids and we want to focus the analysis of the effects of the experimental condition on the chromosomal genes. I was wondering if it's statistically valid to run DESeq2 on chromosome and plasmids separately or if it's more correct running the differential expression analysis on all genes together and then analyze chromosomal and plasmids genes separately from the results. Thanks!

I would run in combined mode. Normalization and dispersion estimation benefit from having many genes available.