Fig. 5From: Compressing gene expression data using multiple latent space dimensionalities learns complementary biological representationsDifferent latent dimensionalities implicate different tissue types. a Sample Pearson correlation for all data in the testing data partition for The Cancer Genome Atlas (TCGA). The different algorithms follow the legend provided in panel d. b Mean Pearson correlation for select cancer types in the testing data partition. Pearson correlation gain between sequential latent dimensionalities for c select cancer types in TCGA and d select tissue types in GTExBack to article page