Normalized enrichment Score (NES) cut-off

[Mahanish Jung Thapa]

One of my friends did GSEA and used NES cut-off value of equal or greater than 2 for upregulated gene sets. The cut-off value is arbitrary. He further did GO analysis of the enriched upregulated gene sets and got good results.

1. What would be the possible reasons for taking NES cut-off value of equal or greater than 2 for upregulated gene sets?

2. Why can't a cut-off value of 1.9 or 2.1 be considered?

3. Is there any research paper where the author used the same cut-off value 2?

[Anthony Castanza]

Using an arbitrary NES cutoff is not something we advocate. There isn't a specific scientific justification to do so.
A better approach is to select sets based on nom pValue and FDR cutoffs.

If you want a research paper to use to justify this arbitrary approach, I'd suggest asking your friend.

-Anthony

Anthony S. Castanza, PhD

Curator, Molecular Signatures Database

Mesirov Lab, Department of Medicine

University of California, San Diego