Table 2.
Comparison of kernels for l-mer content with their AA-property enhanced counterparts.
| Method | auROC50 | #Wins |
|---|---|---|
| Spectrum (l = 5) | 15.2% | 7/54 |
| Spectrum-RBF (l = 5, σ = 1) | 42.1% | 45/54 |
| Mismatch (l = 5, m = 1) | 42.3% | 13/54 |
| Mismatch-RBF (l = 5,m = 1, σ = 1) | 43.6% | 36/54 |
| Profile (l = 5, τ = 7.5) | 82.1% | 3/54 |
| Profile-RBF (l = 5,τ = 7.5, σ = 100) | 82.2% | 10/54 |
Comparison of the three kernels proposed in [11,21,22], with their AA-property enhanced counterparts for remote homology detection of 54 protein families. auROC50 is the average auROC50 score and #Wins the number of families for which each method outperforms its counterpart (Spectrum vs. Spectrum-RBF, Mismatch vs. Mismatch-RBF, Profile vs. Profile-RBF). The kernels taking advantage of AA properties lead to a higher average accuracy in all three cases (p-Values: 6.92 10−8 for spectrum, 0.0045 for mismatch, and 1.0 for profile kernels). For l and τ we use the published parameter settings. For σ we chose the best result among σ = {0.1,1,10,100,1000}.