RSEM Groups and Pairwise testing

Question

[yrc0313]

I have 30 different sets of RNA-seq that I would like to run through RSEM. My question is whether it is better to split them into two random groups with 15 fastq sets in each group or create a group for each sample. In terms of the resulting FPKM values will there be a difference? Thank you!

Answer 1

[Charles Warden]

You will want to run RSEM per sample.

If you have biological groups to separate your samples, you can work directly with the expression values from RSEM or use a tool like EBSeq for differential expression. If you don't have biological groups for your samples, then that would make things trickier (but presumably you had some reason for running these samples).