Table 4 Independent testing performance of our reproduction of state-of-the-art models using the PDB186 dataset. The column P represents results collected from the corresponding publications
From: StackDPP: a stacking ensemble based DNA-binding protein prediction model
Models | ACC(%) | SN(%) | SP(%) | MCC | ||||
|---|---|---|---|---|---|---|---|---|
| Â | P | R | P | R | P | R | P | R |
DPP-PseAAC [17] | 77.42 | 75.81 | 83.87 | 83.87 | 70.97 | 67.74 | 0.79 | 0.52 |
LocalDPP [11] | 79.00 | 72.50 | 92.50 | 86.02 | 65.60 | 59.14 | 0.63 | 0.47 |
Adilina et al. [20](Group) | 82.26 | 79.04 | 95.0 | 86.02 | 69.90 | 61.30 | 0.67 | 0.62 |
Adilina et al. [20](RFE) | 76.88 | 77.95 | 77.00 | 93.55 | 76.9 | 62.37 | 0.55 | 0.58 |
iDNAProt-ES [14] | 80.64 | 72.04 | 81.31 | 83.87 | 80.00 | 60.21 | 0.84 | 0.46 |