Table 5 Performance comparison using sequence formulation techniques using F. vesca and R. chinensis
From: N6-methyladenine identification using deep learning and discriminative feature integration
Species | Methods | ACC (%) | SN (%) | SP (%) | MCC |
|---|---|---|---|---|---|
F. vesca | TAC | 90.52 | 93.32 | 83.21 | 0.811 |
NAC | 85.89 | 93.87 | 87.92 | 0.792 | |
Kmer | 85.90 | 88.45 | 80.13 | 0.786 | |
PseSNC | 80.13 | 83.34 | 79.93 | 0.774 | |
PseDNC | 90.95 | 92.65 | 87.83 | 0.788 | |
PseTNC | 89.32 | 91.34 | 85.54 | 0.808 | |
Hybrid feature (without feature selection) | 95.87 | 97.75 | 90.86 | 0.903 | |
Hybrid feature (with feature selection) | 97.70 | 98.01 | 97.30 | 0.951 | |
R. chinensis | TAC | 88.52 | 90.14 | 83.21 | 0.773 |
NAC | 85.89 | 75.32 | 87.32 | 0.736 | |
Kmer | 85.90 | 80.45 | 86.43 | 0.750 | |
PseSNC | 79.13 | 81.34 | 76.63 | 0.721 | |
PseDNC | 88.95 | 89.65 | 81.83 | 0.792 | |
PseTNC | 84.98 | 83.56 | 85.99 | 0.712 | |
Hybrid feature (without feature selection) | 91.75 | 93.09 | 90.33 | 0.891 | |
Hybrid feature (with feature selection) | 95.75 | 96.45 | 94.55 | 0.921 |