Fig. 1

Hierarchical clustering analyses of the promoter methylomes of 8 pairs of HCC and adjacent non-tumor samples. a Clustering of the average DNA methylation levels of all promoters were used in the “Pvclust” algorithm. Two types of P values (%) on the edge of the cluster are provided: approximately unbiased (AU) P value and bootstrap probability (BP) P value, which indicate how well the cluster is supported by the data. b Clustering of the top 1000 CGIs containing highly variable methylations that were selected based on P values from a chi-square analysis. The methylation ratio was calculated as sequenced reads number of C/sequenced reads number of C + T. Red color indicates high methylation ratio, black color indicates moderate methylation ratio, and green color indicates low methylation ratio. _T denotes tumor tissue samples, and _N denotes non-tumor tissue samples