Table 5 Comparison of REGAL vs. Random Forests. REGAL also outperforms a second technique from [7] using random forest trees for the identification of C → U editing sites in mitochondrial genomes.
From: Genetic algorithm learning as a robust approach to RNA editing site prediction
Random Forests | REGAL | |||||
|---|---|---|---|---|---|---|
Sensitivity | Specificity | Accuracy | Sensitivity | Specificity | Accuracy | |
A. thaliana | 0.70 | 0.81 | 0.74 | 0.81 | 0.91 | 0.86 |
B. napus | 0.73 | 0.81 | 0.77 | 0.81 | 0.91 | 0.86 |
O. sativa | 0.72 | 0.81 | 0.72 | 0.85 | 0.92 | 0.88 |
Overall | 0.71 | 0.81 | 0.84 | 0.82 | 0.91 | 0.87 |