Table 1 Average sensitivity (SEN), specificity (SPC), two-state accuracy (Q2) and Matthews correlation coefficient (MCC) for 10 different data sets in each experiment with different databases.
From: Predicting mostly disordered proteins by using structure-unknown protein data
| Â | SEN | SPC | Q2 | MCC |
|---|---|---|---|---|
30000 unlabeled sequences | ||||
Swiss-Prot | 0.684 | 0.978 | 0.938 | 0.718 |
Uniprot | 0.581 | 0.976 | 0.922 | 0.636 |
TrEMBLE | 0.567 | 0.980 | 0.925 | 0.643 |
70000 unlabeled sequences | ||||
Swiss-Prot | 0.701 | 0.973 | 0.936 | 0.715 |
Uniprot | 0.570 | 0.979 | 0.923 | 0.638 |
TrEMBLE | 0.504 | 0.978 | 0.914 | 0.585 |