Fig. 3From: Benchmarking differential abundance methods for finding condition-specific prototypical cells in multi-sample single-cell datasetsPerformance of the six DA testing methods for DA prediction on the three synthetic datasets (linear (A), branch (B), and cluster (C)) with a range of DA ratios (0.75, 0.85, and 0.95) in the target DA cell population. The boxplots represent the distributions of AUROC and AUPRC scores for various target DA cell populations across multiple random seeds. #points represents the number of points for each boxplotBack to article page