What to merge with cuffmerge?

Question

[sgtstars]

Hi,

I'm doing an RNA-seq experiment, with an affected sample and a control sample, for each of two different tissues. I plan to use cuffdiff to look for deferentially expressed genes in each of the tissues separately. My question is which of these samples should I be merging with cuffmerge? I can either merge all four samples to create the one .gtf, or merge just the samples from the same tissue together to create two .gtfs, since I'm going to be analysing the tissues separately with cuffdiff anyway. Not sure which is the more appropriate approach to use?

Thanks in advance for any help

Answer 1

[Satyajeet Khare]

Hi,

You can merge all samples that you want compare with each other using Cuffdiff into one single GTF file. If you do not plan to compare tissue 1 with tissue 2, then you do not need to merge them. If you plan to compare control of tissue 1 with treatment of tissue 1, you need to merge only these two conditions.

Having said that, I do not see any serious issue even if you merge all together.

Best,