Fig. 5 | Scientific Reports

Fig. 5

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

Fig. 5

Comprehensive Evaluation of the "Lasso + SuperPC" Model Across Multiple Databases and Metrics. (A and B). Comparison of the C-index and HR for our model versus existing models in the MIME package across TCGA, CCGC, and GEO databases. (C) Results from multivariate Cox regression analysis, identifying significant predictors of prognosis. (D) A nomogram that estimates 1-year, 3-year, and 5-year survival probabilities. (E). Calibration plot comparing observed versus nomogram-predicted overall survival (OS) at multiple time points, confirming the model’s accuracy. (F) Decision curve analysis (DCA) that quantifies the clinical usefulness of different predictive factors used in the model. (G) Trends in the C-index over a 10-year period, comparing two prognostic models for assessing long-term outcomes in glioma patients.

Back to article page