Parameters

[Arko Samad]

I'm running a N=1 sample (1 WT, 1 KO)
Are these the right parameters I should be using because almost all of my results, enrichment in WT and enrichment in KO, from a completed analysis with C7 Immuno Sigs possess a FDR greater than 25%

[Anthony Castanza]

You would need to make sure you're running gene_set permutation mode, those parameters are cut off from your screenshot.

I would probably advocate trying to use log2_ratio_of_classes as the ranking metric instead of diff-of-classes. If that errors for whatever reason (e.g. divide by zeros), you can try adding a small pseudocount (e.g. 0.001 or similar) to your data. If you do get NaN values like that you can also try computing that ranking outside GSEA, filtering or adjusting the NaNs, and using the GSEA Preranked mode.

That said we very much don't recommend running N=1 experiments and it's difficult to give advice in scenarios like this.

-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 visit https://groups.google.com/d/msgid/gsea-help/8281a359-3bf8-435d-bbe7-5c9b5f477ec5n%40googlegroups.com.