Identification of Key Genes Associated with Atrial Fibrillation and Hypoxia Using WGCNA and Machine Learning Technology

Chao Wang¹Mardan Muradil²Jianbin Huang¹Jie Cai¹Fangbao Ding¹Li Zhang¹Mengda Li³Ju Mei^1*Zhaolei Jiang^1*

• ¹Department of Cardiothoracic Surgery, Xinhua Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China
• ²Spine Center, Xinhua Hospital Affiliated to Shanghai Jiaotong University School of Medicine, Shanghai, China
• ³Henan Provincial People's Hospital, Zhengzhou, Henan Province, China

Background: Atrial fibrillation (AF) is among the most prevalent cardiac arrhythmias worldwide, and its incidence is steadily rising due to global aging. Hypoxia, a well-recognized trigger of AF, plays a pivotal role in the onset and progression of AF. However, the molecular mechanisms underlying the interplay between AF and hypoxia remain unclear, and specific biomarkers for this condition are lacking. This study aimed to identify key hypoxia-related genes associated with AF through an integrated bioinformatics approach that combines weighted gene co-expression network analysis (WGCNA) with machine learning (ML) algorithms, and to assess their potential diagnostic significance. Methods: This study employed an integrative approach combining weighted gene co-expression network analysis (WGCNA) and machine learning (ML) to identify key genes associated with AF under hypoxic conditions. AF-related gene expression data were sourced from the Gene Expression Omnibus (GEO) database, and hypoxia-related gene sets from the Molecular Signatures Database (MSigDB) database. WGCNA was employed to identify gene modules associated with AF, which were then intersected with hypoxia-related genes. Candidate hub genes were identified using random forest and least absolute shrinkage and selection operator regression. Their diagnostic performance was evaluated using receiver operating characteristic (ROC) curve analysis. A predictive nomogram was developed, and immune infiltration analysis and gene set enrichment analysis (GSEA) were performed to explore associated biological pathways and alterations in the immune landscape. Results: WGCNA identified 34 gene modules, with the most AF-relevant module comprising 624 genes. Intersection analysis and ML algorithms identified SLC6A6, BGN, and PFKP as key genes. ROC analysis demonstrated strong diagnostic potential. Immune cell profiling showed increased infiltration of M2 macrophages and dendritic cells in AF samples, with significant correlations to the expression of these hub genes. Conclusion: This study identified SLC6A6, BGN, and PFKP as key genes associated with AF under hypoxic conditions and successfully developed a diagnostic model with promising clinical applicability. These genes likely play important roles in hypoxia-mediated AF pathogenesis and are closely associated with immune cell infiltration,providing potential biomarkers for early diagnosis and precision treatment of AF. This study provides novel insights into the molecular mechanisms underlying the interplay between hypoxia and AF.