Table 1.
Performance of the models based on amino acid composition of the peptides on training datasets.
| Classifier | Sensitivity | Specificity | Accuracy | MCC |
|---|---|---|---|---|
| AAC-RF with CDHit(1) | 72.73 | 100.00 | 89.26 | 0.79 |
| AAC-RF without CDHit(2) | 71.43 | 87.50 | 80.00 | 0.60 |
| AAC-SVM with CDHit(3) | 63.64 | 100.00 | 85.71 | 0.72 |
| AAC-SVM without CDHit(4) | 78.57 | 93.75 | 86.67 | 0.74 |
AAC: amino acid composition; RF: Random Forest algorithm; SVM: support vector machine algorithm; with CDHit: CDHit-screened datasets; without CDHit: CDHit-unscreened datasets.