Fig. 4 | Scientific Reports

Fig. 4

From: Comprehensive multi-omics and machine learning framework for glioma subtyping and precision therapeutics

Fig. 4

Evaluation of Prognostic Model Performance and Gene Significance Across TCGA, CCGC, and GEO Cohorts. (A) C-index distribution for various models trained on TCGA and validated on CCGC and GEO datasets. (B) Meta-analysis of univariate Cox regression results for survival prediction across TCGA, CCGC, and GEO datasets. (C) C-index distribution for the "Lasso + SuperPC" model across TCGA, CCGC, and GEO datasets. (D) Coefficient estimates from multivariate Cox regression analysis for selected genes identified by the Lasso + SuperPC model. (E) Forest plot of hazard ratios (HR) and 95% confidence intervals (CI) from univariate Cox regression analysis for selected genes across TCGA, CCGC, and GEO datasets. (F). Kaplan–Meier survival curves for the TCGA, CCGC, and GEO datasets based on the "Lasso + SuperPC" model.

Back to article page