Table 2.

Models’ scores after testing with an independent SAPPHIRE dataset.^a

Model	Accuracy	Sensitivity	Specificity	MCC	ROC AUC
TemStaPro-t80	0.923	0.849	0.997	0.856	0.990
TemStaPro-t75	0.930	0.868	0.992	0.867	0.989
TemStaPro-t70	0.960	0.951	0.968	0.919	0.993
TemStaPro-t65	0.957	0.957	0.957	0.914	0.993
TemStaPro-t60	0.951	0.987	0.916	0.905	0.995
TemStaPro-t55	0.934	0.989	0.879	0.873	0.996
TemStaPro-t50	0.908	0.989	0.827	0.828	0.993
TemStaPro-t45	0.907	0.989	0.825	0.825	0.992
TemStaPro-t40	0.896	0.995	0.798	0.808	0.993
SAPPHIRE	0.942	0.951	0.933	0.884	0.980
SCMTPP	0.865	0.849	0.881	0.731	—
ThermoPred	0.86	0.938	0.782	0.729	—

^aEach column’s highest values are emphasized in bold.