I am a veterinarian working on a breed-specific disease with limited number of samples sent for RNA-Sequencing. As I was told that I am a biologist instead of bioinformatician, I was directed to use the commercial product StrandNGS to run analysis on my sequencing results.
There is a tutorial available by StrandNGS on this, but the tutorial is extremely vague and I literally had to figure multiple things by myself; so here's the gist of what I have done
1) Total RNA purified from colon samples of a) Control Beagles x3 b) Diseased Mucosa of paired Dachshund x3 c) Normal Mucosa of paired Dachshund was sent for sequencing using illumine HiSeq 2500 with single-read, 50bp library
2))RIN, cDNA, sequencing, and quality check was performed by the sequencing company
3) Average reads were at 20million reads
4) Reads were mapped onto the CamFam library using StrandNGS
5) Reads were filtered according to the user manual, and the samples were grouped into Beagle and Dachshund as well as Control, Diseased and Normal Mucosa
6) Normalization was done via TMM
7) Normalized reads were ran in the R editor within the built in script of StrandNGS
8) Script ran was "DEG in edgeR"
I was wondering if what I did was correct as most literature suggest edgeR to detect DEGS when replicate numbers are less than 12, and I don't think it would be correct to run the reads using normal T-test or paired T-test.
I managed to get a list, with multiple columns where I have log cpm, FC, log FC, p-value, corrected p-value, regulation, and the combination of comparison amongst each of 3 groups.
What I am a little confused is while the comparisons between each of 3 groups gave me up and down regulation, the concluded log cpm, FC, log FC, p-value, corrected p-value and regulation only gave me up regulation or not significant without a down regulation. In this case, is it safe to say that I actually have a list of DEGs, but I can only see there up and down regulation using the data from the comparison between each of 3 groups?
and how would I be able to make use of the log cpm when it is all positive?
It's very confusing, even for me because it is a very unusual results I'm looking at.
Please do enlighten me if there's anyone out there who is familiar with the combined use of edgeR and StrandNGS.