Table 1. Performance comparison of models with different methods and datasets.

Method	Dataset	Sensitivity	Specificity	Accuracy	Balanced accuracy
ECMPride	D1	0.8925	0.9360	0.9340	0.9142
	D2	0.8783	0.8623	0.8638	0.8703
EcmPred	D1	0.8462	0.9158	0.9145	0.8810
	D2	0.6500	0.7700	0.8300	0.7100

Notes.

D1

Training dataset constructed in ECMPride’s model

D2

Training dataset constructed in EcmPred’s model