Skip to main content

Table 3 Independent test performance (data set S238) of three currently available methods, our previous work, the single-model method, and the three methods with multiple models.

From: Feature-based multiple models improve classification of mutation-induced stability changes

Method AUC MCC Q2 Se Sp PPV NPV
MUpro 0.65 0.24 79.41 28.89 91.19 43.33 84.62
MuStab 0.67 0.26 77.31 40.00 86.01 40.00 86.01
I-Mutant2.0 0.70 0.27 65.97 68.89 65.28 31.63 90.00
EASE-AA 0.83 0.45 82.35 60.00 87.56 52.94 90.37
EASE-AA2 0.72 0.36 82.77 35.56 93.78 57.14 86.19
EASE-ASA 0.81 0.43 83.19 51.11 90.67 56.10 88.83
EASE-SS 0.82 0.48 83.19 62.22 88.08 54.90 90.91
EASE-MM 0.85 0.47 81.09 68.89 83.94 50.00 92.05
  1. EASE-MM is a consensus method of EASE-ASA and EASE-SS