Fig. 7

Construction and validation of clinical predictive models. A LASSO regression coefficient graph. B Partial likelihood deviation of the LASSO coefficient distribution. The two vertical dashed lines represent lambda.min and lambda.1 se. C Intersecting gene sets screened under 5 different results, retain KPNA2, YBX1, and MED8. Unicox represent Univariate cox regression. D Kaplan‒Meier survival analysis of the HCC patient (TCGA-LIHC) with risk scoring. E ROC analysis for predicting the 1/3/5-year survival rate of HCC (TCGA-LIHC). F ROC curve analysis for risk scores and other clinicopathological indicators. G Kaplan‒Meier survival analysis of HCC patients (ICGC-LIRI-JP) with risk scoring. H ROC analysis for predicting the 1/3/5-year survival rate of HCC (ICGC-LIRI-JP). I ROC curve of the NASH cohort. J, K TCGA patients were categorized into high-risk and low-risk groups, based on the median cutoff of the risk score (J), and the distribution of risk scores and patient survival between low and high risk (K). L Heatmap displaying the expression of prognostic genes of the SPCG model in the high- and low-risk groups. M Nomogram of risk groupings and clinical characteristics for predicting survival at 1, 3, and 5 years. N Calibration curves for testing the agreement between actual and predicted outcomes at 1, 3, and 5 years