How can we validate RNAseq differential gene expressions?

Question

[Sharon]

How can we validate RNAseq differential gene expressions?

I used Tophat and cuffdiff to get my differential expressions. Then I redid using Salmon and edgeR. I got the same expressions which is doubted by my professor. He found some genes up expressed or down expressed which he did not expect. So how can we say true or wrong? and how can I validate my results?

Thanks

Answer 1

[Devon Ryan]

The standard practice is to do qPCR with a few transcripts on other samples. It's rather important that other samples be used, as otherwise you're just testing whether the tools produced the correct output, when what you really care about is whether the output holds in other samples (regardless of what your prof might whine about when you tell him this).

[Sharon]

Got it, thank you Devon !