Confronting false discoveries in single-cell differential expression

TLDR

Single‑cell differential expression analysis dissects cell‑type‑specific responses, yet the distinguishing principles and performance of available statistical methods remain unclear. The study demonstrates that method performance depends on accounting for biological‑replicate variation. The authors illustrate this by analyzing differential expression in injured mouse spinal cord, revealing true and false discoveries. Methods that ignore biological‑replicate variation are biased, leading to false discoveries, with popular methods detecting hundreds of genes even when no biological differences exist.

Abstract

Differential expression analysis in single-cell transcriptomics enables the dissection of cell-type-specific responses to perturbations such as disease, trauma, or experimental manipulations. While many statistical methods are available to identify differentially expressed genes, the principles that distinguish these methods and their performance remain unclear. Here, we show that the relative performance of these methods is contingent on their ability to account for variation between biological replicates. Methods that ignore this inevitable variation are biased and prone to false discoveries. Indeed, the most widely used methods can discover hundreds of differentially expressed genes in the absence of biological differences. To exemplify these principles, we exposed true and false discoveries of differentially expressed genes in the injured mouse spinal cord.