Fig 1.
Overview of single cells from 10 HCC samples.
(A) t-SNE of HCC samples. (B) t-SNE of the 20 clusters. (C) t-SNE of cell types. (D) Dot plot showing the top 1 marker gene expression across the 7 cell types. (E) Heat map showing the top 5 marker gene expression. (F) t-SNE of 12 clusters. (G) t-SNE of CD4 T cell and CD8 T cell. (H) Dot plot showing cell markers of CD4 and CD8 T cell.
Fig 2.
Comprehensive analysis of T cells.
(A) GO enrichment analysis of T cells marker genes. (B) KEGG enrichment analysis of T cells marker genes. (C) GSEA enrichment analysis of T cells marker genes. (D) Bubble diagram showing ligand-receptor pair-mediated interactions between T cells and other cells. (E) Differential expressed gene along the pseudotime clustered into four subclusters and the top annotated GO terms in each cluster were provided. (F) SCENIC analysis predicts important TFs in T cell according to RSS.
Table 1.
Clinical and pathological characteristics of HCC patients in the TCGA-LIHC and ICGC-LIRI-JP cohorts.
Fig 3.
The construction of risk model.
(A) Lasso regression identify 4 genes. (B) Cross-validation for the parameters in TCGA dataset. (C) Kaplan-Meier curves of survival analysis in TCGA dataset based on risk score. (D) The distribution of risk score and survival status in TCGA dataset. (E) Time ROC analysis at 1, 2, 3 years in TCGA dataset. (F) Heatmap of identified gene expression in TCGA dataset. (G) Univariate Cox regression analysis revealed the relationship between patients’ overall survival and clinicopathological parameters in TCGA dataset. (H) Multivariate Cox regression analysis in TCGA dataset.
Fig 4.
The validation of risk model and construction of nomogram.
(A) Kaplan-Meier curves of survival analysis in LIRI-JP dataset based on risk score. (B) The distribution of risk score and survival status in LIRI-JP dataset. (C) Time ROC analysis at 1, 2, 3 years in LIRI-JP dataset. (D) Heatmap of identified gene expression in LIRI-JP dataset. (E) Univariate Cox regression analysis revealed the relationship between patients’ overall survival and clinicopathological parameters in LIRI-JP dataset. (F) Multivariate Cox regression analysis in LIRI-JP dataset. (G) Nomogram depicting the predicted survival rates of patients at 1, 3, and 5 years. (H) The calibration curves for predicting 1, 3, and 5 years.
Fig 5.
(A) Box plot of TIDE score between high-risk and low-risk groups.
(B) Box plot of 16 immune checkpoint level between high-risk and low-risk groups. (C) Box plot of 22 immune cells’ infiltration level between high-risk low-risk groups.
Fig 6.
Immunohistochemistry showing the protein expressions of PTTG1 and BATF.
(A) The expression of PTTG1 in HCC and Paired non-tumorous tissue. (B) Box plot of PTTG1 expression values. (C) The expression of BATF in HCC and Paired non-tumorous tissue. (D) Box plot of BATF expression values.