Is there too little variation to detect meaningful gene changes?

Question

[Megan]

Hi everyone,

I wanted to get some feedback on some DESEQ analysis I am currently doing. I have tried to run my pipeline and my PCA plot looks like this.

I have tried running it and I am getting little to no DEGS. Any advice would be greatly appericated.

[karl.stamm]

So there's no difference. Reasonable result. Triple check your protocol and sample handling, did they get mixed up? Did they get processed correctly? Do you have gene expression of a reasonable level on genes you do expect to have active? This would be evidence of good RNA extraction.

[Megan]

The RIN values for these samples ranged around 6.0-7.0, which I think would suggest RNA degradation. From what I am aware of, I don't think that the samples were mixed up but I will double check that.

[swbarnes2]

I doubt your RNA quality is the issue. It's not likely that poor quality would cause reads to align to the wrong gene. Its far more likely that your treatment has a very small effect.

Answer 1

[Istvan Albert]

The problem is not necessarily that there is too little variation - we can't tell how much variation is because the plot is about percentages

the problem seems to be that the variation between the groups is not consistent

Small variations can be easily detected as long as there is consistency between replicates.

H the problem may be that the intra-replicate variability is comparable to the inter-group variability