Table 1.

Summary of QSAR models considered in the current study for identifying the TLR3/7/8/9 antagonist.

Machine Learning Method	Descriptors	Prediction CCR	Accuracy	Sensitivity	Specificity
k-Nearest Neighbor	MOE 2D	0.711 ± 0.031	0.748	0.875	0.547
k-Nearest Neighbor	DragonX-H	0.733 ± 0.083	0.766	0.881	0.585
Random Forest	MOE 2D	0.737 ± 0.057	0.777	0.632	0.869
Random Forest	DragonX-H	0.717 ± 0.029	0.732	0.635	0.809
Support Vector Machine	MOE 2D	0.773 ± 0.031	0.762	0.721	0.794
Support Vector Machine	DragonX-H	0.705 ± 0.043	0.745	0.594	0.839

QSAR, quantitative structure–activity relationship; CCR, correct classification rate; TLR, Toll-like receptor; MOE, molecular operating environment.