Fig. 3From: BERMUDA: a novel deep transfer learning method for single-cell RNA sequencing batch correction reveals hidden high-resolution cellular subtypesRemoving batch effects in scRNA-seq data of pancreas cells. a UMAP visualizations of batch effect removal results for Experiment all on pancreas dataset. Identified alpha and beta cell subpopulations in the Baron batch are highlighted with dashed circles. b Expression patterns of differently expressed genes within alpha and beta cells colored by log-transformed TPM values. Statistical significance of differential expression analysis is listed in Additional file 1: Table S1-S2Back to article page