Fig. 5
From: Multi-factor regulatory network and different clusters in hypertrophic obstructive cardiomyopathy
Clusters of HOCM and identification of hub genes a K = 4 was selected as the optimal number of clusters since the K value is decreased by a negligible amount. b The tSNE algorithm provided each sample with a unique x- and y-coordinate (tSNE1 and tSNE2) according to each sample’s gene expression of 32 core genes. All HOCM samples were clearly divided into 4 clusters. c The expression of core genes in all HOCMs was shown in the heatmap. d The expression of core genes in 4 clusters was shown in the heatmap. e COMP, FMOD, AEBP1 and SULF1 showed significant expression in different clusters. f Receiver operating characteristic (ROC) curves of the model and 4 hub genes validated the classification performance of 4 genes