Fig. 3: Predictive assessment of risk factors for the progression of HCC.

A Workflow of the relationship between the altered proteome of the PME and HCC progression. B Distribution of RI and survival time for each HCC patient. C Correlation between the RI and survival time. D Kaplan–Meier curve analysis based on the RI value grouped by the median. Differential proportion of distribution for AFP (E) and single or multiple tumors (F) according to different RI levels (f, RI^high, RI > 484.18; g, RI^high, RI > 437.32). G Proportion of risk factors for the progression of HCC in each patient. Detailed workflow of calculations for RIs described as Fig. S3A. H Comparison of the RI among different risk factors related with HCC progression. Data are presented as the median with range. ^*P < 0.05, ^***P < 0.001 compared with inflammation; ^&&P < 0.01, ^&&&P < 0.001 compared with angiogenesis; ^#P < 0.05, ^###P < 0.001 compared with metabolism; ^$$$P < 0.001 compared with immunity; ^%%P < 0.01 compared with proliferation and invasion. RI, risk index. Overall survival curves (I, J) and ROC curves (K, L) of the HCC progression prediction model with three markers combined (CMPK2, TYMP, NADSYN1) according to the median PI in our dataset (I, K) and in the validation dataset (J, L). PI, prognostic index. Detailed calculations for PI described in METHODS section. M Comparison of the proteins for occurrence and progression in predicting HCC occurrence. Number of proteins was counted according to AUC of ROC curve analysis. P value of four-fold table was calculated by Chi-square test and Fisher exact tests. N Comparison of the proteins for occurrence and progression in predicting HCC progression. Number of proteins was counted according to Kaplan-Meier survival curve analysis and log-rank test. P value of four-fold table was calculated by Chi-square test and Fisher exact tests.