Fig. 4: Heatmaps displaying the clustering results of PHet and the top three methods are presented for six single cell transcriptomics datasets. | Nature Communications

Fig. 4: Heatmaps displaying the clustering results of PHet and the top three methods are presented for six single cell transcriptomics datasets.

From: Heterogeneity-preserving discriminative feature selection for disease-specific subtype discovery

Fig. 4

The datasets (N = 6) are: Baron (a–d), Camp (e–h), Darmanis (i–l), Li (m–p), Patel (q–t), and Yan (u–x). For each method, the selected features are used. The bold font ARI score indicates the best-performing method for the corresponding data in the comparison. All values are provided as a Source Data file.

Back to article page