Identification of Cuproptosis-Related Diagnostic Biomarkers in COPD by Intergrating Machine Learning and Bioinformatics

慢性阻塞性肺疾病(chronic obstructive pulmonary disease，COPD)作为全球主要致死疾病之一，与铜死亡密切相关。本研究旨在从铜死亡相关基因中筛选COPD诊断生物标志物。我们分析了来自GEO数据库的三个数据集(GSE106986、GSE103174、GSE71220)。在标准化及归一化后，基于文献中确定的铜死亡相关基因，我们在GSE106986中进行差异表达分析，识别出7个铜死亡相关差异表达基因及2个患者亚群(C1/C2)。通过GO和KEGG富集分析、相关性分析及单样本基因集富集分析，提示差异基因与炎症、氧化应激和免疫应答相关通路有关。进一步采用WGCNA、LASSO和SVM算法筛选出5个核心基因，其中RALB和SLC31A1为关键诊断标志物。在GSE103174和GSE71220中验证其诊断性能，ROC曲线显示RALB和SLC31A1联合AUC为0.77。Chronic obstructive pulmonary disease (COPD), a leading global cause of mortality, is closely linked to cuprotosis. This study aimed to identify diagnostic biomarkers for COPD from cuprotosis-related genes. Three GEO datasets (GSE106986, GSE103174, GSE71220) were analyzed. After normalization, differential expression analysis in GSE106986 identified 7 cuprotosis-related DEGs and 2 patient subgroups (C1/C2) using literature-curated cuprotosis genes. GO/KEGG enrichment, correlation, and single-sample GSEA revealed associations with inflammation, oxidative stress, and immune pathways. WGCNA, LASSO, and SVM identified 5 hub genes, with RALB and SLC31A1 as key diagnostic markers. Validation in GSE103174 and GSE71220 showed combined AUC of 0.77 for RALB and SLC31A1.