Fig. 3

Identification of MAG-related biomarkers using multiple machine learning algorithms and development of a nomogram diagnostic model. (A) AUC values of 113 algorithm combinations based on 10 different machine learning algorithms, used to determine the optimal predictive model. (B–E) Expression levels of MAG-related biomarkers. (F) ROC curve analysis of the identified MAG-related biomarkers. (G) Construction of a nomogram diagnostic model based on the expression profiles of MAG-related biomarkers. (H) Decision curve analysis (DCA) of the nomogram model. (I) Calibration curve analysis evaluating the predictive accuracy of the nomogram. (J) ROC curve analysis of the nomogram-based scoring model.