Abstract
BACKGROUND: Lung cancer is one of the most commonly diagnosed cancer worldwide. As one of the liquid biopsy analytes, alternations in cell-free DNA (cfDNA) methylation could function as promising biomarkers for lung cancer detection. METHODS: In this study, differential methylation analysis was performed to identify candidate markers, and lasso regression with 10-fold cross-validation (CV) was used to establish the diagnostic marker panel. The performance of the binary classifier was evaluated using the receiver operating characteristic (ROC) curve and the precision-recall (PR) curve. RESULTS: We identified 4072 differentially methylated regions (DMRs) based on cfDNA methylation data, and then a 10-DMR marker panel was established. The panel achieved an area under the ROC curve (AUROC) of 0.922 and an area under the PR curve (AUPR) of 0.899 in a cfDNA cohort containing 29 lung cancer and 74 normal samples, showing outstanding performance. Besides, the cfDNA-derived markers also performed well in primary tissue datasets, which were more robust than the tissue-derived markers. CONCLUSION: Our study suggested that the 10-DMR marker panel attained high accuracy and robustness and may function as a novel and promising target for lung cancer detection.