Abstract
BACKGROUND: Research has demonstrated that the homeostasis of mitochondria and programmed cell death (PCD) are intimately linked to chronic obstructive pulmonary disease (COPD). Consequently, identifying biomarkers of COPD from mitochondria-related genes (MRGs) and programmed cell death-related genes (PCD-RGs) is of paramount importance. METHODS: Differentially expressed genes (DEGs) from the GSE42057 dataset and COPD-related genes (COPD-RGs) via weighted gene co-expression network analysis (WGCNA) were intersected with MRGs and PCD-RGs to select candidates. Machine learning identified biomarkers, validated across GSE42057 and GSE94916 datasets. Pathway enrichment, immune infiltration, and drug prediction analyses were performed. RESULTS: Eight candidate genes were derived from intersecting DEGs, COPD-RGs, MRGs, and PCD-RGs. Five biomarkers (BCL2, CCR7, FAM162A, FOXO1, RPS3) were identified, showing consistent dysregulation in COPD. These biomarkers activated the "ribosome" pathway. CCR7 and FOXO1 correlated positively with naïve B cells, while BCL2 negatively correlated with M0 macrophages. BCL2 exhibited strong binding to dolastatin 10, beauvericin, and micellar paclitaxel. RT-qPCR confirmed biomarker expression. CONCLUSION: BCL2, CCR7, FAM162A, FOXO1, and RPS3 are biomarkers for COPD, providing a new breakthrough point for the treatment of this disease.