Skip to main content
Fig. 3 | Journal of Translational Medicine

Fig. 3

From: Integrative single-cell transcriptomic analyses reveal the cellular ontological and functional heterogeneities of primary and metastatic liver tumors

Fig. 3

Characterization of the heterogeneous T/NK cell populations in the primary and metastatic liver tumors. A UMAP plot of scRNA-seq profile from T/NK cells that are separated into 13 cell clusters. Cells are colored according to different clusters. B UMAP plot showing 13 cell clusters in different types of tissues. C Dotplot showing GO biological processes enriched by the highly expressed genes of CD4+ T cells in NTs, PTs and MTs, respectively. The dot size indicates the ratios of the enriched genes in the GO term versus all highly expressed genes; and the dot color indicates the P values determined by Hypergeometric distribution test. D Dotplot showing GO biological processes enriched by the highly expressed genes of CD8+ T cells in NTs, PTs and MTs. Dot size indicates the ratios of the enriched genes in the GO term versus all highly expressed genes; and dot color indicates the P values determined by Hypergeometric distribution test. E Heatmap showing the normalized expression (z-score) of T cell function-associated gene sets in each T cell cluster. F Feature plots showing the typical markers of CD4+ T cells, CD8+ T cells and NK cells. G Tissue prevalence of T cell clusters estimated by Ro/e scores. H Violin plots showing the regulatory and exhausted scores of FOXP3+ Treg across different types of tissues. Wilcox test was used to assess the difference between groups. “****” represents “P < 0.0001”. I Dotplot showing the percentage of expressed cells and average expression levels of regulatory and exhaustion-related genes of Treg in three types of tissues. J Forest plot showing the prognostic values of each CD4+ T cell cluster infiltration in the primary HCC cohort (TCGA-LIHC, n = 346) and the metastatic liver tumors cohort (MT2020, n = 198). The CD4+ T cell cluster infiltration was estimated based on the top 50 gene expression signatures of each cell cluster by GSVA. The median GSVA score was used to categorize patients into “high” and “low” groups. The hazard ratios (HRs) with 95% confidence intervals (CIs), and P values were determined by univariate Cox proportional hazards regression analyses. K Violin plots showing the cytotoxicity and exhaustion scores of TCF7+ progenitor-like memory CD8+ T cells (Tpm) across different types of tissues. Wilcox test was used to assess the difference between groups. “****”, “***” and “**” represent “P < 0.0001”, “P < 0.001” and “P < 0.01”, respectively. L Forest plot showing the prognostic values of each CD8+ T cell cluster infiltration in the primary HCC cohort (TCGA-LIHC; n = 346) and the metastatic liver tumors cohort (MT2020; n = 198). The CD8+ T cell cluster infiltration was estimated based on the top 50 gene expression signatures of each cell cluster by GSVA. The median GSVA score was used to categorize patients into “high” and “low” groups

Back to article page