Table 5 Comparing the performance of DEEPLYESSENTIAL on down-sampled dataset against methods that solely use sequence features; numbers in boldface indicate the best performance
From: DeeplyEssential: a deep neural network for predicting essential genes in microbes
Method | # features | AUC | Sensitivity | PPV |
|---|---|---|---|---|
Liu et al. 2017 | 40 | 0.794 | 0.715 | 0.243 |
Azhagesan et al. 2018 | 267 | 0.838 | 0.754 | 0.321 |
ZUPLS | 274 | 0.705 | 0.663 | 0.255 |
DeeplyEssential | 89 | 0.842 | 0.801 | 0.749 |