Table 4.

Performance of six tools trained on the merged unbiased training dataset and tested on BTD, HBTD, and EBTD separately

Tool	Sensitivity			Specificity			MCC
	BTD	HBTD	EBTD	BTD	HBTD	EBTD	BTD	HBTD	EBTD
Local-DPP	0.5168	0.5103	0.9701	0.6642	0.6512	0.9277	0.168	0.1503	0.9007
DNABP	0.5422	0.537	0.9492	0.6125	0.6044	0.8492	0.1412	0.1297	0.8068
StackDPPred	0.4273	0.4187	0.9653	0.6152	0.5972	0.9358	0.0393	0.0148	0.9029
PDBP-Fusion	0.4989	0.4932	0.9445	0.6409	0.6316	0.8293	0.1291	0.1158	0.7837
LSTM-CNN_Fusion	0.5023	0.4962	0.9575	0.6491	0.6365	0.893	0.14	0.1231	0.8553
PB_DBP	0.4092	0.4016	0.9562	0.7093	0.6931	0.9032	0.1089	0.0914	0.8631