Fig 2: Two different tumor microenvironment of CD8+T cells infiltrated in IRGPI-high and IRGPI-low: (A, B) Representative multiplex immunofluorescence images to show the distribution of CD8 + T cells in IRGPI-low and IRGPI-high, respectively: MAPT (green), GHR (red), CD5L (yellow), CCL14 (orange), CD8 (white) and DAPI (blue). Scale bar, 50 µm. (C) Statistical graphs to show the proportion of CD8 + T cells between IRGPI-low and IRGPI-high groups. p = 0.0328; Wilcoxon test. (D) K-M curves for IRGPI-high and low in the hospital cohort. p = 0.025

Fig 3: Construction of IRGPI: (A) Weighted Gene Co-expression Network Analysis (WGCNA) immune-related genes and their four modules based on their degree of co-expression. The number in the module indicates the correlation between the module genes, HCC, and the p-values. (B) Univariate Cox proportional hazard regression analysis screens out the genes that have the greatest impact on the survival time of HCC and hazard ratio. (C) K-M curves for high and low IRGPI groups in the TCGA cohort. Log rank test, p < 0.001. (D) K-M curves for high and low MAPT expression groups in the TCGA cohort. Log rank test, p < 0.001. (E) K-M curves for high and low CCL14 expression groups in the TCGA cohort. Log rank test, p < 0.001. (F) K-M curves for high and low GHR expression groups in the TCGA cohort. Log rank test, p < 0.001. (G) K-M curves for high and low CD5L expression groups in the TCGA cohort. Log rank test, p < 0.001. (H, I) Univariate and multivariate independent prognostic analyses of IRGPI.