Fig. 5From: Benchmarking differential abundance methods for finding condition-specific prototypical cells in multi-sample single-cell datasetsPerformance of the six DA testing methods in identifying differentially abundant cell-types in three synthetic datasets (linear (left), branch (middle), and cluster (right)) with batch effects of varying magnitudes (from 0 to 1.5). When batchSd=0, no batch effects are present. The boxplots represent the AUROC scores for different target DA cell populations, DA ratios, and random seeds. #points represents the number of points for each boxplotBack to article page