Does someone can explain the curve shape of GSEA analysis enrichement profile? As control of my GSEA analysis with PreRanked data set following Log2FoldChange, I did the same analysis but with Log2FoldChange randomly attributed to genes of my data set. The consequence is that, in the interrogated genes sets, I have systematically lower ES values and much higher FDR (>25%) in the control GSEA analysis which is really consistent.
However, the curve shape of the enrichment profile is always (in all genes sets) with the peak offset to the left of the graph.. and this also for my control GSEA analysis... I cannot understand the basis of this, and why in the control one it does not acquire completely random shape (with peak in the middle for ex.)
Can you help me to make up my mind with this?
Thank you !