Hi Manjunath
You should include all genes, not just those that pass an arbitrary p-value cutoff. GSEA needs the information from the non-significantly differentially expressed genes when scoring the set and set nulls.
As to which metric specifically to use, I would lean towards using either the test statistic (i.e. wald) or the log2FC, but it is our policy not to provide an official recommendation for GSEA Preranked metric selection.
-Anthony
Anthony S. Castanza, PhD
Curator, Molecular Signatures Database
Mesirov Lab, Department of Medicine
University of California, San Diego
--
You received this message because you are subscribed to the Google Groups "gsea-help" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gsea-help+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gsea-help/7d8b7c53-b73c-4f6e-bcb3-f98efad5723an%40googlegroups.com.