Table 10 Results of the experimental comparison between the proposed MODAS method and competing structural class prediction methods on the D498 dataset.
Classifier used (name of the method, if any) | Feature vector | Reference | Accuracy | ||||
|---|---|---|---|---|---|---|---|
|  |  |  | α | β | α/β | α+β | Avg |
Component-coupling | AA composition | 70 | 93.5 | 88.9 | 90.4 | 84.5 | 89.2 |
Neural network | AA composition | 80 | 86.0 | 96.0 | 88.2 | 86.0 | 89.2 |
Rough sets | AA composition and physicochemical properties | 49 | 87.9 | 91.3 | 97.1 | 86.0 | 90.8 |
SVM with RBF kernel (SCPRED) | custom | 79 | 94.9 | 91.7 | 94.2 | 86.1 | 91.5 |
SVM | AA composition | 82 | 88.8 | 95.2 | 96.3 | 91.5 | 93.2 |
Fuzzy k-nearest neighbor algorithm | protein sequence | 68 | 95.3 | 93.7 | 97.8 | 88.3 | 93.8 |
Nearest Neighbor (NN-CDM) | protein sequence | 69 | 96.3 | 93.7 | 95.6 | 89.9 | 93.8 |
LogitBoost | AA composition | 71 | 92.5 | 96.0 | 97.1 | 93.0 | 94.8 |
SVM with RBF kernel (SCEC) | PSI-BLAST based p-collocated AA pairs | 75 | 98.0 | 93.3 | 95.6 | 93.4 | 94.9 |
IB1 | PSI-BLAST based p-collocated AA pairs | 75 | 95.0 | 95.8 | 97.8 | 94.2 | 95.7 |
SVM with polynomial or RBF kernels (MODAS) | custom | this paper | 96.7 | 97.5 | 95.6 | 97.1 | 96.8 |