Fig. 1

The comprehensive assessment of the genetic and epigenetic heterogeneity identified DNA damage repair (DDR) alterations as a ubiquitous feature in multifocal hepatocellular carcinomas (mHCCs). a, b Heterogeneous mutational patterns of seven mHCCs. Bar plots showing proportions of mutations shared by any pair of lesions in each patient (a). Stacked bar plots showing the mutation proportions of each lesion accounted for by each of the six mutation types. For each patient, a Fisher exact test was applied to evaluate the difference in mutation spectra between each pair of lesions. Significant P values are denoted by asterisks (**P  < 0.01, *P < 0.05) (b). c, d Heterogeneous CNA patterns of seven mHCCs. CNA profiles of all tumor lesions from seven mHCC patients. Each track represents one tumor lesion. CNA regions are colored either in red (gain) or blue (loss) (c). Bar plots showing proportions of CNAs shared by any pair of lesions in each patient (d). e Phylogenetic trees of seven mHCC patients. For each patient, the phylogenetic tree was constructed from all somatic nonsynonymous mutations using the NJ algorithm. Truncal mutations and branch mutations are represented with brown and green lines, respectively. The branch lengths are proportional to the number of mutations. Putative HCC driver genes and DDR genes are indicated along with trees. f The detection of HBV integration sites based on WES.Grid indicates the presence (green) or absence (gray) of HBV integration sites detected. The HBV integration sites verified by VISDB database are indicated in red. g The relative contribution of each COSMIC v2 signature for mHCC patient. h, i Heterogeneous epigenetic features of seven mHCCs. Principal component analysis of the methylation profiles based on the averagemethylation values in 5 kb titling regions of all normal tissue and tumor lesions from seven mHCC patients (h). Methylation density plots of seven mHCC patients (i). For each column that represents one sample in the plot, colors were mapped to the density values in the corresponding distribution. The black dashed lines correspond to the five quantiles of the distributions (0%, 25%, 50%, 75%, and 100%) and the red dashed lines correspondto the mean value of the distributions. j Phyloepigenetic trees of seven mHCC patients. For each patient, the phyloepigenetic tree was constructed based on the NJ approach using CpGs with at least 20 reads. For direct comparison, phylogenetic trees from panel e were reproduced next to the matched phyloepigenetic trees. Pearson’s correlation coefficient was employed to calculate the similarity between the genetic and epigenetic distance matrices for each patient and it is indicated on the plot. k Pearson’s correlation analysis was performed between TMB and TNB for HCC2, HCC4, HCC5 andHCC6. l Pearson’s correlation analysis was performed between CD8+T cell densities and expressions of PD-L1 (CPS score) for HCC2, HCC4, HCC5 and HCC6