Table 3 Classifier Performance
| Â | Demographics only | Demographics plus genes |
|---|---|---|
Number of subjects (N) | 115 (NC:57, CA:58) | 115 (NC:57, CA:58) |
Demographic variable used | 3: gender, age and pack-years | 3: gender, age and pack-years |
Picked Gene Features | 0 | E2F1 ERCC5 XRCC1 GPX1 TP63.2 GSTP1 CDKN1A TP53.2 ERCC4.2 RB1 ERCC5.2 KEAP1.2 ERCC1.2 |
AUC based on 50 MI* | 68.3% (95%R CI: 63.3%–73.1%) | 97.5% (95%R CI: 96.0%–98.9%) |
Classification Accuracy based on 50 MI* | 65.5% (95% CI: 56.8%–74.1%) | 93.0% (95% CI: 88.4%–97.7%) |
PPV* | 64.8% (95% CI: 52.9%–76.6%) | 93.1% (95% CI: 86.6%–99.6%) |
NPV* | 66.3% (95% CI: 53.6%–79.0%) | 93.0% (95% CI: 86.4%–99.6%) |
Sensitivity* | 69.2% (95% CI: 57.4%–81.1%) | 93.1% (95% CI: 86.6%–99.6%) |
Specificity* | 61.6% (95% CI: 49.0%–74.2%) | 92.9% (95% CI: 86.3%–99.5%) |