It is know that some enrichment methods (Over representation analysis) methods do not take into account that downregulated genes in a pathway could result in the pathway being more active (see this answer for example). I'm curious if there is a better way to address the problem: that pathway might have upregulated and down-regulated genes.
The Functional Class Score (FCS) methods, such as GSVA and GSEA, are methods that take into account the expression value one way or another and might be better suited for this.
I think GSVA compares a gene set to the other remaining genes.
Is there any method similar to it that compare gene sets to random sets of the same number of genes (I think GSVA doesn't) and use that to compare between gene sets scores? I couldn't find any so far.
I would expect that most gene sets are not differently regulated thus a good score to compare gene sets would result in a normal distribution. Helping up picking up gene sets that are over represented and really active one way or another.
