Table 2 Comparison of the predictive performance of our BEC-Pred model with other machine learning methods at level 3 of the EC number
From: A general model for predicting enzyme functions based on enzymatic reactions
Data | ACC | MCC | F1 score |
|---|---|---|---|
KNN-morgan2 | \(0.843\pm 0.005\) | \(0.827\pm 0.005\) | \(0.843\pm 0.005\) |
RF-morgan2 | \(0.768\pm 0.003\) | \(0.744\pm 0.005\) | \(0.739\pm 0.005\) |
DNN—morgan2 | \(0.861\pm 0.005\) | \(0.847\pm 0.005\) | \(0.848\pm 0.006\) |
KNN-drfp | \(0.837\pm 0.006\) | \(0.820\pm 0.005\) | \(0.837\pm 0.006\) |
DNN-drfp | \(0.852\pm 0.004\) | \(0.834\pm 0.005\) | \(0.837\pm 0.004\) |
GNN | \(0.861\pm 0.002\) | \(0.847\pm 0.002\) | \(0.854\pm 0.003\) |
BEC-Pred (ours) | \(0.916\pm 0.003\) | \(0.907\pm 0.003\) | \(0.913\pm 0.004\) |