Drawbacks of Custom DE Methods #19
Description
In Chapter 6 Marker detection, redux there is recommendation against limma-style analysis. I wonder if that applies to today's times. One of the reasons is computational efficiency compared to Cohen's d. I often see data sets with fifty patients and if the gene abundances were averaged per gene and per cell group, perhaps limma would suit.