A Data-driven Approach for Biomarker Discovery based on U-centered Distance Correlation Network: Multi-omics Warning Signals for Non-small Cell Lung Cancer

Introduction/Objective: Lung cancer is the leading cause of cancer-related mortality worldwide, and non-small cell lung cancer (NSCLC) accounts for the majority of cases. Alterations in metabolic activities play important roles in NSCLC development, wherein related genes and metabolites interact with each other, involving multiple forms. Methods: To comprehensively understand the pathogenic mechanisms and improve the performance of clinical early, precise diagnosis, this study proposed a data-driven approach for biomarker discovery based on U-centered distance correlation network (DCN) to investigate NSCLC metabolism-related reactions. In DCN, changes in molecular relationships during NSCLC initiation and progression are measured using the t-statistics of U-centered distance correlation for network construction, in which prospective warning signals representing NSCLC onset can be identified without human intervention. Additionally, the network construction criterion in DCN can precisely and effectively capture both linear and nonlinear molecular relationships in simple and biologically relevant manners. Results: DCN was successfully employed to analyze NSCLC metabolism-related metabolomics and genomics datasets. Statistical analyses confirmed that compared with other algorithms, the gene and metabolite biomarker panels identified by DCN provided more reliable diagnostic capabilities for clinical NSCLC detection. Biological analyses revealed that disturbed energy metabolism and lipid metabolism occurred during tumor cell proliferation and growth in NSCLC patients. Discussion: The gene ASPA and metabolite aspartic acid were significantly decreased in NSCLC samples, suggesting that the corresponding amino acid metabolic activities were intricately linked to NSCLC progression. Conclusion: These findings demonstrated that DCN can further facilitate NSCLC studies to improve clinical outcomes in patients.