Dear all,

I am performing differential gene expression analysis of single-nuclei RNA-seq data with DESeq2 and pseudobulk counts per cell type. 3 groups, 4 samples per group.

The nuclei vary to some extent concerning their mitochondrial RNA counts. So, DESeq2 naturally produces the most significant genes to be mitochondrial.

Is there a way to include this as a covariate in the design formula or somewhere else and provide an additional column "percent.mitochondrial.RNA" in the colData?

Thank you, best, Chris

You could include the percent mitochondrial in the design, but also, if you aren't interested in these genes, you could filter them before running DESeq2.