Thank you for your advice!
So you think DESeq2 and edgeR are more accurate than Cuffdiff?
They are faster than Cuffdiff.
But Cole Trapnell (Cuffdiff author) believes DESeq and edgeR (the tools that test raw gene-level counts) are less accurate than Cuffdiff, since they assume that isoforms have the same length.
So are there some known issues with Cuffdiff?
Thanks a lot!