Fig. 4 | Scientific Reports

Fig. 4

From: QSAR machine learning-guided discovery of renieramycin right-half derivatives with cytotoxic activity against NSCLC cells

Fig. 4

The experimental vs. predicted pIC50 of training (●) and test (×) set were obtained from the selected QSAR models. The figure shows the top 5 models among all, i.e., (A) GA-QC-RF, (B) GA-QC-SVR, (C) GA-MS-SVR, (D) GA-MS-XGB, and (E) GA-MS-MLR. RF: Random Forest, SVR: Support Vector Regression, XGB: eXtreme Gradient Boosting/XGBoost, GA: genetic algorithm feature selection, VI: variable importance feature selection, QC: quantum chemical descriptors, MS: non-quantum chemical descriptor.

Back to article page