Table 3.
Performance of logistic regression models with various biomarkers for discriminating healthy controls and lung cancer patients.
Top 5 significant metabolites | Top 10 significant metabolites | Top 20 significant metabolites | ||||
---|---|---|---|---|---|---|
Discovery | Validation | Discovery | Validation | Discovery | Validation | |
AUC | 0.918 (± 0.103) | 0.869 | 0.973 (± 0.065) | 0.947 | 0.947 (± 0.125) | 0.964 |
Accuracy | 0.836 (± 0.155) | 0.829 | 0.902 (± 0.162) | 0.857 | 0.893 (± 0.159) | 0.900 |
Precision | 0.850 (± 0.174) | 0.829 | 0.933 (± 0.214) | 0.866 | 0.903 (± 0.204) | 0.905 |
Recall | 0.855 (± 0.208) | 0.829 | 0.890 (± 0.161) | 0.857 | 0.908 (± 0.117) | 0.900 |