Table 2 Performance comparison of DeEPsnap, DNN, DeepHE, SVM, RF, Adaboost, and GAT.
From: A deep ensemble framework for human essential gene prediction by integrating multi-omics data
Method | AUROC | AUPRC | MCC | F1 | Accuracy |
|---|---|---|---|---|---|
DeEPsnap | 0.9616 \({\pm }\) 0.0059 | 0.9383 \({\pm }\) 0.0083 | 0.7592 \({\pm }\) 0.0186 | 0.8062 \({\pm }\) 0.0149 | 0.9236 \({\pm }\) 0.0059 |
DNN | 0.9511 \({\pm }\) 0.0062 | 0.9232 \({\pm }\) 0.0088 | 0.7329 \({\pm }\) 0.0190 | 0.7850 \({\pm }\) 0.0157 | 0.9152 \({\pm }\) 0.0062 |
DeepHE | 0.9355 \({\pm }\) 0.0072 | 0.8953 \({\pm }\) 0.0148 | 0.6860 \({\pm }\) 0.0295 | 0.7467 \({\pm }\) 0.0246 | 0.9016 \({\pm }\) 0.0105 |
SVM | 0.9562 \({\pm }\) 0.0097 | 0.8527 \({\pm }\) 0.0249 | 0.7371 \({\pm }\) 0.0358 | 0.7887 \({\pm }\) 0.0285 | 0.9123 \({\pm }\) 0.012 |
RF | 0.9502 \({\pm }\) 0.0092 | 0.8532 \({\pm }\) 0.0206 | 0.6899 \({\pm }\) 0.0272 | 0.7268 \({\pm }\) 0.0239 | 0.9103 \({\pm }\) 0.0072 |
Adaboost | 0.9460 \({\pm }\) 0.0078 | 0.8428 \({\pm }\) 0.0219 | 0.7040 \({\pm }\) 0.0243 | 0.7582 \({\pm }\) 0.0196 | 0.9102 \({\pm }\) 0.0076 |
GAT | 0.9171 \({\pm }\) 0.0216 | 0.8554 \({\pm }\) 0.0360 | 0.6681 \({\pm }\) 0.1099 | 0.6568 \({\pm }\) 0.1255 | 0.8710 \({\pm }\) 0.0217 |
