I got list of "highly variable genes" between Groups A x B using variance in edgeR (ref: highly variable genes) Then performed standard edgeR analysis to find DE genes. ref: edgeR
PROBLEM: Why none of the DE genes between A x B are included in "highly variable genes"
My questions 1. What is the reason for the problem ? 2. Considering the problem, what is the use of "highly variable genes" in publication ? Can I simply exclude it and go with DE gene analysis only.
1 answer
Highly variable genes are useful for clustering samples and seeing if you've done things like swapped a sample label or have outlier samples. One should not expect them to be differentially expressed. You're looking at a quality control step, how you proceed will depend entirely on what it looks like.
Log in to answer this question.
did you determine your "highly variable genes" as it is indicated in your first reference? If so, these are not highly variable genes between groups A and B, as you call them, but rather genes with the highest variance in your data. That means they can change a lot independently of the group. For this reason you shouldn't expect an overlap with your AvsB DEGs.