Fig. 3From: ScInfoVAE: interpretable dimensional reduction of single cell transcription data with variational autoencoders and extended mutual information regularizationVisualization of identified clusters. The partitions identified with scDeepCluster, scziDesk, scGNN, ScInfoVAE(RSC) and ScInfoVAE(KM)) on five datasets (Muraro, Quake Limb Muscle, Quake Smart seq2 Trachea ,Young, Adam). The plots illustrate the t-SNE 2D projections of the created embeddings. All selected methods start by producing an embedding for the cells, which is clustered in a second phase. The quality of the method depends on both the created embedding and the clustering algorithm. Both our methods clustered the same embedding, produced by ScInfoVAEBack to article page