Table 1. Statistics of the final selected features after one-step and two-step feature selection for each class of experimental procedure.
| Feature type | CLF |
MF |
PF |
CF |
CRYs |
|||||
|---|---|---|---|---|---|---|---|---|---|---|
| OFS | TFS | OFS | TFS | OFS | TFS | OFS | TFS | OFS | TFS | |
| AAc | 1 | 0 | 3 | 3 | 1 | 1 | 1 | 1 | 1 | 1 |
| Tri-peptide | 2 | 2 | 6 | 4 | 10 | 9 | 3 | 1 | 9 | 6 |
| AAindex | 7 | 1 | 22 | 20 | 10 | 8 | 1 | 1 | 24 | 21 |
| KSAAP | 86 | 21 | 19 | 13 | 66 | 64 | 94 | 65 | 26 | 13 |
| GKSAAP | 4 | 0 | 50 | 38 | 13 | 13 | 1 | 0 | 40 | 24 |
| Total | 100 | 24 | 100 | 78 | 100 | 95 | 100 | 68 | 100 | 65 |
Feature selection was performed on the benchmark datasets.
OFS (One-method Feature Selection) denotes results after two-step feature selection (mRMR).
TFS (Two-method Feature Selection) denotes results after three-step feature selection (mRMR+FFS).
AAc: features generated based on 20 standard amino acids.