Table 5.
Performances of prediction algorithms related to validated interactions in Table 4
| Algorithm | ACC % | PPV % | TPR % | TNR % | FNR % | FPR % | F1 % | MCC |
|---|---|---|---|---|---|---|---|---|
| PITA | 5.3 | 99.3 | 5.0 | 88.6 | 94.9 | 11.4 | 9.6 | –0.01617 |
| MiRanda | 25.5 | 99.5 | 25.4 | 59.3 | 74.5 | 40.6 | 40.5 | –0.01946 |
| TargetScan | 11.4 | 99.3 | 11.2 | 78.0 | 88.8 | 21.9 | 20.0 | –0.01912 |
| Pictar | 7.1 | 99.1 | 6.9 | 81.3 | 93.1 | 18.7 | 12.9 | –0.02581 |
| DIANA-microT | 20.1 | 99.5 | 19.9 | 74.0 | 80.0 | 26.0 | 33.2 | –0.00847 |
| miRNATIP | 30.6 | 99.7 | 30.5 | 69.9 | 69.4 | 30.1 | 46.7 | 0.00055 |
Statistical measures reported in this table are accuracy (ACC), precision (PPV), sensitivity (TPR), specificity (TNR), miss-rate (FPR), F1-measure (F1) and Matthews correlation (MCC), respectively