Table 4 The Prediction performance using logistic regression with selected variables for each microbiota
Variables | AUC (95% CI) | Cut-off | SE (95% CI) | SP (95% CI) | PPV (95% CI) | NPV (95% CI) | Accuracy (95% CI) | |
|---|---|---|---|---|---|---|---|---|
Saliva | Logistic-total variables | 1.00 (1.00, 1.00) | 1.00 | 1.00 (0.85, 1.00) | 1.00 (0.78, 1.00) | 1.00 (0.85, 1.00) | 1.00 (0.79, 1.00) | 0.39 (0.24, 0.57) |
Logistic-selected variable | 0.91 (0.82, 1.00) | 0.60 | 0.87 (0.66, 0.97) | 0.80 (0.52, 0.96) | 0.87 (0.64, 0.97) | 0.80 (0.54, 0.96) | 0.84 (0.69, 0.94) | |
Support vector machine | 0.90 (0.77, 1.00) | 0.57 | 0.91 (0.72, 0.99) | 0.87 (0.60, 0.98) | 0.91 (0.70, 0.99) | 0.87 (0.61, 0.98) | 0.87 (0.72, 0.96) | |
Naïve Bayes | 0.93 (0.84, 1.00) | 0.10 | 0.91 (0.72, 0.99) | 0.87 (0.60, 0.98) | 0.91 (0.70, 0.99) | 0.87 (0.61, 0.98) | 0.89 (0.75, 0.97) | |
Neural network | 0.64 (0.44, 0.83) | 0.56 | 0.87 (0.66, 0.97) | 0.53 (0.27, 0.79) | 0.74 (0.48, 0.94) | 0.73 (0.44, 0.90) | 0.71 (0.54, 0.85) | |
Stool | Logistic-total variables | 1.00 (1.00, 1.00) | 1.00 | 1.00 (0.86, 1.00) | 1.00 (0.77, NA) | 1.00 (0.86, 1.00) | 1.00 (0.78, 1.00) | 0.36 (0.21, 0.53) |
Logistic-selected variables | 0.77 (0.59, 0.96) | 0.58 | 0.76 (0.55, 0.91) | 0.79 (0.49, 0.95) | 0.86 (0.63, 0.95) | 0.65 (0.41, 0.91) | 0.77 (0.61, 0.89) | |
Support vector machine | 0.97 (0.92, 1.00) | 0.67 | 0.92 (0.74, 0.99) | 0.93 (0.66, 1.00) | 0.96 (0.78, 1.00) | 0.87 (0.62, 1.00) | 0.90 (0.76, 0.97) | |
Naïve Bayes | 0.78 (0.61, 0.94) | 0.76 | 0.60 (0.39, 0.79) | 0.93 (0.66, 1.00) | 0.94 (0.69, 0.97) | 0.57 (0.35, 0.98) | 0.69 (0.52, 0.83) | |
Neural network | 0.45 (0.24, 0.66) | 0.61 | 1.00 (0.86, 1.00) | 0.21 (0.05, 0.51) | 0.69 (0.29, 1.00) | 1.00 (0.43, 1.00) | 0.69 (0.52, 0.83) |