Fig. 3.

The prognostic signature of the risk score analyses for the nine matrix stiffness-related ECM genes. (A) The optimal turning parameters (log λ) were selected using tenfold cross-validation. (B) Construction of a Lasso regression model. (C-D) Kaplan‒Meier analysis of the OS rate of HCC patients in the high- and low-risk groups in the training and testing cohorts. (E-F) The ROC curves of the training and testing cohorts for overall survival at 1 year, 2 years, and 3 years. (G-H) Distribution of risk scores of the matrix stiffness-related ECM gene model in the training and testing cohorts. (I-J) Overall survival state distributions for the training and testing cohorts. (K-L) Principal component analysis (PCA) between the high- and low-risk groups in the training and testing cohorts. (M-N) T-distributed stochastic neighbour embedding (t-SNE) between the high- and low-risk groups in the training and testing cohorts