Table 2.

The performance of classification models for the manually curated T_MC test set

Model	AUC	Acc	SN	SP	Threshold
Default threshold
SYBA	0.903	0.844	0.913	0.775	0.0
SAScore	0.865	0.617	0.934	0.300	6.0
RF	0.875	0.819	0.863	0.775	0.5
Optimized threshold
SYBA	0.903	0.871	0.902	0.840	19.1
SAScore	0.865	0.859	0.799	0.919	3.9
SCScore	0.528	0.601	0.707	0.496	3.7
RF	0.875	0.842	0.855	0.828	0.6

Quality measures AUC, Acc, SN and SP, as well as thresholds, are reported as their average values over 30 T_MC instances