Table 3 Performance evaluation of the SVM models trained with various features based on independent testing dataset (3732 ubiquitylation sites and 10,664 non-ubiquitylation sites)
Training features | Sensitivity | Specificity | Accuracy | MCC |
|---|---|---|---|---|
20D binary coding (20D) | 62.59Â % | 65.85Â % | 65.00Â % | 0.253 |
Amino Acid Composition (AAC) | 66.37Â % | 64.63Â % | 65.08Â % | 0.274 |
Amino Acid Pair Composition (AAPC) | 69.05Â % | 69.05Â % | 69.05Â % | 0.340 |
Position Weight Matrix (PWM) | 73.90Â % | 67.29Â % | 69.01Â % | 0.364 |
Position-Specific Scoring Matrix (PSSM) | 73.20Â % | 68.45Â % | 69.68Â % | 0.369 |
Solvent-Accessible Surface Area (SASA) | 63.91Â % | 61.36Â % | 62.02Â % | 0.223 |
Secondary Structure (SS) | 55.60Â % | 51.34Â % | 52.45Â % | 0.061 |