Table 5.

Performances of deep neural network/protein–protein interaction (DNN-PPI) framework on the benchmark dataset.

Test Set	Accuracy	Recall	Precision	F-Score	MCC
1	0.9853	0.9845	0.9849	0.9847	0.9706
2	0.9877	0.9876	0.9865	0.9871	0.9754
3	0.9916	0.9909	0.9913	0.9911	0.9831
4	0.9892	0.9911	0.9867	0.9889	0.9784
5	0.9941	0.9963	0.9915	0.9939	0.9883
Hold-out	0.9878	0.9891	0.9861	0.9876	0.9757