Table 3.

Cross-validation results

Descriptor	Methoda	AUC-ROC	BACC	Sensitivity	Specificity
MOE	BayesNet—DU	0.81	0.76	0.72	0.79
	RF—MU	0.86	0.78	0.79	0.76
	SVM—DU	0.85	0.77	0.79	0.76
RDKit	BayesNet	0.79	0.73	0.61	0.85
	RF—MU	0.86	0.78	0.78	0.78
	SVM—DU	0.83	0.77	0.83	0.72
QNA	BayesNet—MU	0.75	0.69	0.62	0.75
	RF—MU	0.83	0.75	0.72	0.78
	SVM—DU	0.85	0.80	0.80	0.80

^aDU and MU stand for diversity under-sampling and multiple under-sampling, respectively