I'm a begginer in GSEA field, however when I performed GSEA analysis comparing my homemade data set sorted by log2FoldChange criteria with genes data set from very different sources, all theses GSEA analysis provides me with NES value above 0,5 ( at least ! And most of the time above 0,6 or 0,7). This even though some of the genes set that I compared to are not at all relevants.
My first question is then how can I be confident with GSEA analysis with gene set that are supposed to be relevant if even non-relevant gene sets are already positives?
Secondly, There is quite hudge differences in the analysis i following the ranked method used (log2FoldChange, T-test, Differences of classes, ratio of classes etc..). How can we choose the good criteria for the analysis?
I hope that It's clear enough,
Thank you !