Skip to main content
Fig. 1 | Proteome Science

Fig. 1

From: A risk score model based on TGF-β pathway-related genes predicts survival, tumor microenvironment and immunotherapy for liver hepatocellular carcinoma

Fig. 1

Development of the prognostic model with 8 TGF-β pathway-related gene in TCGA-LIHC dataset. A LASSO Cox regression analysis on 22 TGF-β pathway-related genes. The coefficients of genes closed to zero with the increasing lambda value. Dashed red line indicates the optimal lambda value = 0.0372 for constructing an optimal model with the least number of genes. B 95% CI of partial likelihood deviance under different lambda values evaluated by cross-validation. Red dot indicates the lambda value corresponding with the optimal model. The left (green dots) and right (blue dots) of the red dot indicate the lambda values with the decreasing and increasing number of genes respectively. C The distribution of samples with different survival status, and expression of eight prognostic genes for each sample ranking by risk score from low to high. Censored and uncensored indicate dead and alive status respectively. Horizontal axis indicates samples. Z-score of expression from green to red indicates low to high expression. D Kaplan-Meier survival curve of low- and high-risk groups classified by the prognostic model. Log-rank test was conducted. E ROC and AUC of the prognostic model in predicting 1-year, 3-year and 5-year OS. CI, confidence interval. HR, hazard ratio. ROC, receiver operation characteristic. AUC, area under ROC curve

Back to article page