Hello,

I have 6 RNA-Seq samples, 3 for each condition and am performing RNA-Seq analysis using EdgeR to filter & TMM-Normalize counts then using Limma-VoomwithQuality Weights. My MDS plot looks like this:

Red is treated and Black is untreated, The numbers "205,206,207" represent a different mouse. 205 seems show high variation from 207 & 206 across dimension 1 while expected variation is shown in Dimension 2. Should I do any more normalization or control for bath effects?

Thanks, Sara

The experiment is a paired comparison, which can be analysed by

design <- model.matrix(~ Mouse + Treatment)

This is a standard analysis. There's no need to change the normalization.