Dear all, if have a question regarding a result that I obtained using the DEXSeq package. I analysed a set of five samples of an siRNA kd (3 control, 2 treatment). When plotting the estimated dispersion plot show below was generated. I've been searching for explanations and understanding for a while, but I guess I'm a bit lost by now and I would appreciate any hint. What puzzles me ist the fact that the fitted curve furcates for genes with low normalised counts and high dispersion. I have seen some unexpected dispersion plots during my search but nothing alike. In addition is my interpretation that a lot of genes seem to be very variable between replicates correct? I checked whether I've chosen the correct experimental design. 

Best,

Lennart