Table 2 10-fold cross-validation performances of iProDNA-CapsNet on the training dataset (PDNA-543) under various decision thresholds
From: iProDNA-CapsNet: identifying protein-DNA binding residues using capsule neural networks
Setting | ACC (%) | SN (%) | SP (%) | PR (%) | MCC | AUC |
|---|---|---|---|---|---|---|
Threshold = 0.5 | 74.73 | 77.38 | 74.55 | 17.32 | 0.282 | 0.832 |
FPR ≈ 5% | 91.21 | 36.31 | 95.00 | 33.34 | 0.301 | 0.832 |
FPR ≈ 15% | 83.66 | 64.21 | 85.00 | 22.78 | 0.313 | 0.832 |
SP ≈ SN | 76.02 | 76.02 | 76.02 | 17.93 | 0.287 | 0.832 |