Clarify documentation of DEG/DAR methods

[mukamel-lab]

Thank you for your great ongoing work on Seurat. The huge impact of this tool makes it a standard that sets the expectations for many best practices in the field. Unfortunately, I have noticed an increasing number of papers that use the FindMarkers to define differentially expressed genes and/or differential accessibility regions (DEGs/DARs) when comparing biological conditions rather than cell types. The Seurat DE vignette rightly points out the dangers of this approach, which treats every cell as an independent replicate without accounting for biological variability across samples. Unfortunately, the ease/convenience of the FindMarkers function seems to lead to inappropriate use of this tool for drawing conclusions that suffer from highly inflated significance.

It would be great to update Seurat with some very clear "warning labels" to reduce the danger of this inappropriate practice.

Thanks!
Eran Mukamel