Does The Order Of Samples In Cuffdiff Matter?

Question

[skm770]

Hi,

We wanted to do fold change using cuffdiff on control_vs_treatment case. What should the order of samples be for input in cuffdiff. Its on the cuffdiff website that :-

FPKMx FPKM of the gene in sample x FPKMy FPKM of the gene in sample y log2(FPKMy/FPKMx) The (base 2) log of the fold change y/x

Currently I am giving cuffdiff with input of control sample and then treatment sample so the genes that are down regulated according to the result are actually upregulated. I just wanted to confirm if reversing the order of sample in cuffdiff would give me different results or not.

thanks

Answer 1

[Devon Ryan]

Yes, reversing the order of the groups will flip the sign of the log2 fold-change.