Table 1 Performance measures and dimensions for the different features
From: A deep learning method to more accurately recall known lysine acetylation sites
Feature | Dimension | Accuracy | Specificity | Sensitivity | AUC | MCC |
|---|---|---|---|---|---|---|
One-hot | 651 | 76.25% | 74.00% | 78.50% | 0.7506 | 0.5256 |
BLOSUM62 | 651 | 76.23% | 71.68% | 80.77% | 0.7880 | 0.5267 |
CKSAAP | 2205 | 73.61% | 70.79% | 76.44% | 0.7290 | 0.4731 |
IG | 1 | 53.22% | 64.02% | 42.43% | 0.5430 | 0.0660 |
AAindex | 434 | 63.65% | 53.92% | 73.38% | 0.6904 | 0.2783 |
PSSM | 30 | 49.50% | 60.46% | 38.53% | 0.4941 | −0.0103 |
Word2vec | 31 | 52.78% | 56.89% | 48.57% | 0.4382 | 0.1814 |