Hi all,

I can't find the answer with Google, hope an expert here can help. If two conditions don't have hundreds of differential express genes (in some conditions such as peripheral blood Alzheimer vs normal), is it possible with gene set variation analysis (GSVA), can I find pathways behave different between two conditions? ChatGPT4 said changes may be subtle at the gene level but more pronounced when looking at coordinated changes across sets of genes within pathways. Gordon comments in previous post above made me rethink about my current method. I try to find if there are papers using this method in non DEG sample and find interesting things at pathway level. In a pathway, some genes up regulated and some genes down regulated. So how can we know overall this pathway score will higher or lower when comparing with control? Thank you so much!

Update: I got answer for one question here: Why p value inconsistent when using different number of pathway score?