Table 2.
Overall classifier performance after excluding poorly performing GO term classifiers.
| Threshold F-measure | F-Measure | Percentage of Proteins Annotated | ||||
|---|---|---|---|---|---|---|
| AAPFC | MultiPfam2GO | PoGO | AAPFC | MultiPfam2GO | PoGO | |
| 0.7 | 0.8955 | 0.9903 | 0.9814 | 1.8 | 2.8 | 25.4 |
| 0.6 | 0.8317 | 0.9705 | 0.9336 | 3.8 | 5.2 | 31.0 |
| 0.5 | 0.6835 | 0.9169 | 0.7518 | 9.8 | 10.7 | 54.3 |
| 0.4 | 0.6619 | 0.8861 | 0.6751 | 17.6 | 13.4 | 65.6 |
| 0.3 | 0.5353 | 0.8520 | 0.5214 | 34.9 | 15.6 | 86.3 |
| 0.2 | 0.4396 | 0.5835 | 0.4456 | 58.1 | 28.2 | 96.2 |
Properties were computed after removing individual GO term classifiers with an F-measure below the indicated threshold.