GSVA and Log2FC

Question

[Morris_Chair]

Hello Everyone, I have a question, is it possible to use the Log2FC for doing the GSVA? In the past I used to use the rlog values or the normalized read working with GSVA from the Broad Institute or other tools, but working with scRNA-seq I can't have it anymore

I appreciate the help

thank you

[jared.andrews07]

Why can't you use your RNA-seq counts?

[Morris_Chair]

Hi Jared, It's a single cell RNA-seq where I have an expression gene list from SEURAT of different clusters, I want to do a GSVA on this clusters but Seurat gives AvgLog2FC only

thanks

[jared.andrews07]

Why not just do it on individual cells and compare between clusters using whatever metric you want (average, median, etc)? I've been involved in the development of a new R package called escape that will be in the next Bioconductor release that has a number of functions to make running ssGSEA via GSVA on scRNA-seq quite easy. You may want to consider that if you want a plug and play solution with additional visualization and comparison functions.

[Morris_Chair]

Hi Jared, Thank you for your answer, definitely I ll take a look to escape :)