Table 3.
The performance of different subcellular localization methods on blind/independent dataset used in BaCelLo.
| Blind1 Dataset (Animal Proteins) | Blind2 Dataset (Fungal Proteins) | |||||
| Method | Cov* | nAcc | GAv | Cov | nAcc | GAv |
| BaCelLo | 66.1 | 56.4 | 61.1 | 66.4 | 71.3 | 68.8 |
| Loctree | 62.2 | 49.5 | 55.5 | 66.4 | 66.9 | 66.6 |
| Psort II | 70.2 | 43.0 | 54.9 | 71.1 | 44.2 | 56.1 |
| SubLoc | 67.8 | 37.2 | 50.2 | 70.5 | 38.4 | 52.0 |
| ESLpred | 79.1 | 35.8 | 53.2 | 84.4 | 37.5 | 56.3 |
| LOCSVMpsi | 80.2 | 38.7 | 55.7 | 88.5 | 51.0 | 67.2 |
| pTARGET | 73.3 | 64.2 | 68.6 | 62.3 | 63.5 | 62.9 |
| NpPred | 87.3 | 74.3 | 80.5 | 93.4 | 72.7 | 82.4 |
Cov is coverage/sensitivity; nAcc is normalized accuracy; GAv = geometric average between coverage and normalized accuracy (*see Additional file 1 for detail description of these parameters).