Question

Issues in generating workable rank file for GSEA

0

Entering edit mode

csijst • 0

@csijst-15102

Last seen 5.2 years ago

Singapore/National University of Singap…

Hi,

I'm trying to generate a ranking file for my differential expressed genes that was generated via DESeq2. I saw some websites like Genome Spot that taught me how to use the p-values and log2foldchange to rank the genes. But in DESeq2 output, some of them appear as "NA". And this will reappear in the ranking file. Apparently, pre-ranked GSEA either java Desktop application (from Broad Institute) or GenePattern are not able to process the NA entries.

As such, what should I do for genes that display NA? As in what is the usual procedure in handling such information? Remove them?

Thank you.

pre-ranked GSEA DESeq2 gseabase GSEA • 1.7k views

ADD COMMENT • link updated 5.9 years ago by Michael Love 41k • written 5.9 years ago by csijst • 0

score 0 · Answer 1 · 2018-06-18

0

Entering edit mode

alserg ▴ 240

@assaron

Last seen 7 days ago

St Louis, MO

Hi,

In DESeq2 you can use t-statistic for ranking, there are no NA-s. If you still want to use p-values, set DESeq2 option cooksCutoff=FALSE for the results function.

ADD COMMENT • link 5.9 years ago alserg ▴ 240

0

Entering edit mode

Hi,

I tried the command,

dDif_res <- results(dLRT, contrast=c("treatment", "shRNA", "ctrl"), alpha=0.05, altHypothesis="greaterAbs", cooksCutoff=FALSE)

head(dDif_res)

But the output still shows NA for some of the genes. These genes do not have any baseMean as well.

ADD REPLY • link 5.9 years ago csijst • 0

score 0 · Answer 2 · 2018-06-18

0

Entering edit mode

Michael Love 41k

@mikelove

Last seen 1 day ago

United States

You can also set NA to 1. There’s example code for this in the vignette FAQ

ADD COMMENT • link 5.9 years ago Michael Love 41k