Extended Data Table 1 Comparison of performance between the classification models and other currently available prediction tools based on the benchmarking set or the test set
Model | Type | TP | FP | TN | FN | Precision (%) | F1 Score (%) | MCC (%) | Specificity (%) | Sensitivity (%) |
|---|---|---|---|---|---|---|---|---|---|---|
AMP Prediction | ||||||||||
 AMPSorter | Transformer | 632 | 65 | 1,006 | 93 | 90.67 | 88.89 | 81.66 | 93.93 | 87.17 |
 AMPlifyimbal | Neural Network | 618 | 147 | 924 | 107 | 80.78 | 82.95 | 70.96 | 86.27 | 85.24 |
 Macrel | Random Forest | 379 | 16 | 1,055 | 346 | 95.95 | 67.68 | 60.15 | 98.51 | 52.28 |
 AMPlifybal | Neural Network | 523 | 155 | 916 | 202 | 77.14 | 74.55 | 58.36 | 85.53 | 72.14 |
 iAMP Pred | Support Vector Machines | 524 | 149 | 922 | 201 | 77.86 | 74.96 | 59.16 | 86.09 | 72.28 |
 AMP Scanner v2 | Neural Network | 538 | 212 | 859 | 187 | 71.73 | 72.95 | 54.13 | 80.21 | 74.21 |
 Bert-Protein | Transformer | 708 | 1,010 | 61 | 17 | 41.21 | 57.96 | 8.07 | 5.70 | 97.66 |
 AMPir | Support Vector Machines | 68 | 110 | 961 | 657 | 38.20 | 15.06 | -1.46 | 89.73 | 9.38 |
 AmPEP | Random Forest | 357 | 738 | 333 | 368 | 32.60 | 39.23 | -19.78 | 31.09 | 49.24 |
Toxin Prediction | ||||||||||
 BioToxiPept | Transformer | 329 | 59 | 336 | 49 | 84.79 | 85.90 | 72.08 | 85.06 | 87.04 |
 ToxinBTL | Neural Network | 365 | 24 | 371 | 13 | 93.83 | 95.18 | 90.46 | 93.92 | 96.56 |
 ToxinpredRF | Random Forest | 369 | 247 | 148 | 9 | 59.90 | 74.25 | 43.60 | 37.47 | 97.62 |
