SE values for log-FC output of FindMarkers

[mirG1234]

Hi all,

First off, thanks for creating such a lovely and accessible package for scRNAseq analysis :)

My request may boil down to a question if there is some way to extract this info already. I'm looking to feed my differential expression results from FindMarkers into a downstream GSEA analysis, and want to be able to use the SNR (signal-to-noise ratio) as is described by the GSEA documentation and the default in their software. However, to do so I need a few additional values that are not present in the DE output, including mean expression for each group, as well and the SE in said the expression values. Alternatively, simply proving an SE for the log2-FC estimate would also be incredibly useful, as SNR could also be calculated as the ratio of LFC/SE.

I know that many of the tests available in the FindMarkers function already calculate the SE under the hood (such as DESeq2), so I was wondering if a) there is currently a way to access these values in seurat v5 or b) if I need to extract the raw matrix to perform tests manually. Furthermore, if this info could be included by default in future versions of FindMarkers I would really appreciate the added convenience!

M