Fig. 5
From: Integrative analysis of thiamethoxam induced hepatocellular carcinoma toxicity mechanisms
Machine learning for screening the key genes. (A) ROC curves for SVM, RF, XGB, GLM; (B) Box plots of residuals for SVM, RF, XGB, GLM; (C) Reverse cumulative distribution of residuals for SVM, RF, XGB, GLM machine learning models; (D) Bar plot showing the feature importance of SVM, RF, XGB, GLM machine learning models; (E) Nomogram was constructed based on five feature genes; (F) Calibration curve of the feature gene nomogram for THX-induced HCC; (G) Decision curve for the feature gene nomogram for THX-induced HCC; (H) ROC curve for the test GEO dataset; (I) ROC curves for external validation using the GEO dataset.
