Hi,

I've been playing with Deseq2 to normalize and analyze my datasets, initially only out of curiosity. The results look more reasonably to me than results from other approaches that I used for the same dataset (e.g. normalizing with universal single-copy genes or genome equivalents). However, it is unclear to me if you would actually encourage the use of Deseq2 for DNA sequencing (since it was developed for the analysis of RNA-seq data), and if it makes sense to use Deseq2 for comparing functional potential of two different environments (rather than two different experimental conditions). My samples come from two different environments that are exposed to similar environmental stressors, and vary strongly in community composition.

Thanks for your help! Josephine

I think the methods will work to the degree that there are features that will work as the bulk of no-or-lowly affected genes work in RNA-seq. In some microbiome or metagenomics datasets it’s not clear there are anything like “housekeeping” features universally present and not greatly effected by condition. But if you say you have these, you can choose these as controlGenes during normalization.