This has been bothering me AND I have seen conflicting information. When it comes to edgeR, how am I supposed to be analyzing the BCV plot to measure up my experimental design? Is it matter of what my BCV hovers around (according to the edgeR vignette, well-controlled human data have BCVs around .2-.4. ) Am I looking to the shape of the distribution along the trend line? What exactly is this telling me, and how is this helpful in improving my design? Appreciate the help.