Skip to main content
Fig. 3 | BMC Bioinformatics

Fig. 3

From: Hierarchical classification-based pan-cancer methylation analysis to classify primary cancer

Fig. 3

CHCT can identify the cancers well using the screen-out methylation signature. a Heat map depicts the methylation level of retained probes of first layer classifier, reflecting clear distinction among the 12 cancer groups. Upper annotation indicates sample types of 12 cancer groups. Left-sided annotation indicates 12 cancer-specific methylation panels. b ROC curve shows the high AUC in predicting 12 cancer types using the methylation signature. c UMAP plot shows a high degree of heterogeneity between the same type samples from TCGA and GEO

Back to article page