Fig. 3.

ROC curves for lung cancer detection. a ROC curves comparing the four genes with the largest areas under the curve in plasma. b ROC of the combined methylation status of CDO1, TAC1, and SOX17 from plasma with the largest area under the curve. c ROC of the combined methylation status of CDO1, SOX17, and HOXA7 from plasma with the largest area under the curve. d, e and f ROC curves assessing the accuracy of the predictions for lung cancer using gene methylation panel with clinical risk factors(age, pack-year, COPD status, nodule size, and pulmonary function values). d Plot is obtained using clinical predictors alone. e Plot is obtained using clinical predictors plus the combination of CDO1, TAC1, and SOX17. f Plot is obtained using clinical predictors plus the combination of CDO1, SOX17, and HOXA7