Fig. 2

Dissecting HB tumor heterogeneity through subclustering and functional pathway analysis. a. UMAP showing the Hepatocyte sub-clustering, with the upper-left UMAP colored in salmon to show which cells were selected for sub-clustering, and the larger right-hand UMAP showing the three sub-clusters that were derived using unsupervised clustering. b. Stacked proportional bar plot showing the prevalence of each Hepatocyte sub-cluster in background (left) and tumor (right) samples. All clusters add up to 100% for each condition. c. Dot plot showing average normalized expression (color) and the percentage of cells expressing the gene (dot size) for marker genes used to define the Hepatocyte sub-clusters. d. UMAP showing Cycling Hepatocyte sub-clustering, with the upper-left UMAP colored in gold to show which Hepatocyte sub-clusters were selected for further sub-clustering, and the larger right-hand UMAP showing the three resultant Cycling Hepatocyte sub-clusters. e. The same Cycling Hepatocyte sub-cluster UMAP plot colored by Seurat-derived S Score (top) and G₂M Score (bottom), with darker blue indicating higher expression. f. Gene Set Enrichment Analysis (GSEA) plots showing a positive enrichment score in tumor cells compared to background for the ‘HALLMARK_MITOTIC_SPINDLE’ gene set and a negative enrichment score for the ‘HALLMARK_IL6_JAK_STAT3_SIGNALING’ gene set.