Table 3.
Prediction performance for the test dataset for each of the tasks in two-stage AMP prediction
| Stage | Prediction task | SEN (%) | SPEC (%) | MeanACC (%) |
|---|---|---|---|---|
| First | AMP | 93.34 | 91.37 | 92.36 |
| Second | Anti-Gram-positive | 91.14 | 88.42 | 89.78 |
| Anti-Gram-negative | 89.58 | 88.79 | 89.19 | |
| Antivirus | 88.87 | 81.05 | 84.96 | |
| Antifungal | 93.86 | 57.92 | 75.89 | |
| Anticancer | 84.33 | 81.24 | 82.79 | |
| Mammalian inhibition | 78.49 | 79.08 | 78.79 |
TP, TN, FP and FN denote true positive, true negative, false positive and false negative, respectively. The sensitivity (SEN), specificity (SPEC) and mean accuracy (MeanACC) are defined as follows: SEN = TP/(TP + FN); SPEC = TN/(TN + FP); and MeanACC = 0.5(SEN + SPEC). The value 0.5 is a default cutoff of confidence values used to determine the positive or negative predictions.