Table 6.
Top 10 logistic models consisting of 8 lung cancer antigen markers in terms of AUC and their sensitivity at 80/90% specificity and AUC under 5-fold cross-validation.
| Serial No. | Marker combination | Sensibility at 80% specificity | Sensibility at 90% specificity | AUC |
|---|---|---|---|---|
| Model 1 | ZNF573 + BRAF + TM4SF1 + SOX2 + MAGE.A4 + BMI1 + FXR1 + HuC | 0.7311 | 0.5507 | 0.8420 |
| Model 2 | BRAF + TM4SF1 + SOX2 + MAGE.A4 + BMI1 + FXR1 + HuC + ESO1 | 0.7251 | 0.5418 | 0.8400 |
| Model 3 | ZNF573 + BRAF + TM4SF1 + SOX2 + MAGE.A4 + BMI1 + FXR1 + ESO1 | 0.7217 | 0.5382 | 0.8398 |
| Model 4 | ZNF573 + BRAF + SOX2 + MAGE.A4 + BMI1 + FXR1 + HuC + ESO1 | 0.7187 | 0.5599 | 0.8394 |
| Model 5 | ZNF573 + BRAF + TM4SF1 + SOX2 + MAGE.A4 + BMI1 + HuC + ESO1 | 0.7340 | 0.5810 | 0.8392 |
| Model 6 | ZNF573 + BRAF + TM4SF1 + SOX2 + MAGE.A4 + BMI1 + FXR1 + CRYAA | 0.7034 | 0.5444 | 0.8368 |
| Model 7 | ZNF573 + BRAF + TM4SF1 + SOX2 + BMI1 + FXR1 + HuC + ESO1 | 0.7341 | 0.5476 | 0.8366 |
| Model 8 | ZNF573 + BRAF + TM4SF1 + SOX2 + MAGE.A4 + BMI1 + HuC + CRYAA | 0.7249 | 0.5598 | 0.8354 |
| Model 9 | ZNF573 + BRAF + TM4SF1 + SOX2 + MAGE.A4 + BMI1 + GAD2 + FXR1 | 0.7217 | 0.5139 | 0.8354 |
| Model 10 | ZNF573 + BRAF + TM4SF1 + MAGE.A4 + BMI1 + FXR1 + HuC + ESO1 | 0.7340 | 0.5416 | 0.8352 |