Table 2. Classification Performance of the Consensus Approaches for Binding, Agonism, and Antagonism Endpointsaa.
| binding (34 models) |
antagonism (22 models) |
agonism (21 models) |
|||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| consensus approach | Sn (%) | Sp (%) | NER (%) | Cvg (%) | rank | Sn (%) | Sp (%) | NER (%) | Cvg (%) | rank | Sn (%) | Sp (%) | NER (%) | Cvg (%) | rank |
| MVL | 61.8 | 91.8 | 76.8 | 99.3 | 4 | 61.5 | 87.3 | 74.4 | 98.9 | 3 | 73.8 | 97.5 | 85.7 | 99.7 | 2 |
| MVI | 60.6 | 98.3 | 79.5 | 80.6 | 8 | 60.0 | 93.8 | 76.9 | 80.1 | 4 | 76.1 | 99.0 | 87.5 | 91.5 | 6 |
| MVS | 26.9 | 100 | 63.5 | 37.5 | 39 | 39.0 | 99.2 | 69.1 | 42.4 | 25 | 64.8 | 99.9 | 82.3 | 51.4 | 17 |
| B | 72.3 | 84.9 | 78.6 | 100 | 1 | 71.0 | 81.2 | 76.1 | 100 | 1 | 74.4 | 95.1 | 84.7 | 100 | 3 |
| Bp | 73.3 | 85.9 | 79.6 | 96.1 | 7 | 73.5 | 82.9 | 78.2 | 92.9 | 2 | 75.8 | 95.9 | 85.9 | 97.7 | 4 |
a
For each consensus approach, sensitivity (Sn), specificity (Sp), non-error rate (NER), coverage (Cvg), and total ranking are reported. B, Bayes; Bp, protective Bayes; MVL, majority voting loose; MVI, majority voting intermediate; MVS, majority voting strict.