AUTHOR=Liang Qianqian , Wang Yide , Li Zheng 

TITLE=Comprehensive bioinformatics analysis identifies metabolic and immune-related diagnostic biomarkers shared between diabetes and COPD using multi-omics and machine learning

JOURNAL=Frontiers in Endocrinology

VOLUME=Volume 15 - 2024

YEAR=2025

URL=https://www.frontiersin.org/journals/endocrinology/articles/10.3389/fendo.2024.1475958

DOI=10.3389/fendo.2024.1475958

ISSN=1664-2392

ABSTRACT=BackgroundDiabetes and chronic obstructive pulmonary disease (COPD) are prominent global health challenges, each imposing significant burdens on affected individuals, healthcare systems, and society. However, the specific molecular mechanisms supporting their interrelationship have not been fully defined.MethodsWe identified the differentially expressed genes (DEGs) of COPD and diabetes from multi-center patient cohorts, respectively. Through cross-analysis, we identified the shared DEGs of COPD and diabetes, and investigated alterations of signaling pathways using Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and gene set enrichment analysis (GSEA). By using weighted gene correlation network analysis (WGCNA), key gene modules for COPD and diabetes were identified, and various machine learning algorithms were employed to identify shared biomarkers. Using xCell, we investigated the relationship between shared biomarkers and immune infiltration in diabetes and COPD. Single-cell sequencing, clinical samples, and animal models were used to confirm the robustness of shared biomarkers.ResultsCross-analysis identified 186 shared DEGs between diabetes and COPD patients. Functional enrichment results demonstrate that metabolic and immune-related pathways are common features altered in both diabetes and COPD patients. WGCNA identified 526 genes from key gene modules in COPD and diabetes. Multiple machine learning algorithms identified 4 shared biomarkers for COPD and diabetes, including CADPS, EDNRB, THBS4 and TMEM27. Finally, the 4 shared biomarkers were validated in single-cell sequencing data, clinical samples, and animal models, and their expression changes were consistent with the results of bioinformatic analysis.ConclusionsThrough comprehensive bioinformatics analysis, we revealed the potential connection between diabetes and COPD, providing a theoretical basis for exploring the common regulatory genes.