Meta-analysis of Transcriptomic Data Reveals Pathophysiological Modules Involved with Atrial Fibrillation

Abstract

Atrial fibrillation (AF) is a complex disease and affects millions of people around the world. The biological mechanisms that are involved with AF are complex and still need to be fully elucidated. Therefore, we performed a meta-analysis of transcriptome data related to AF to explore these mechanisms aiming at more sensitive and reliable results. Public transcriptomic datasets were downloaded, analyzed for quality control, and individually pre-processed. Differential expression analysis was carried out for each individual dataset, and the results were meta-analytically aggregated using the r-th ordered p-value method. We analyzed the final list of differentially expressed genes through network analysis, namely topological and modularity analysis, and functional enrichment analysis. The meta-analysis of transcriptomes resulted in 589 differentially expressed genes, whose protein-protein interaction network presented 11 hubs-bottlenecks and four main identified functional modules. These modules were enriched for, respectively, 23, 54, 33, and 53 biological pathways involved with the pathophysiology of AF, especially with the diseases structural and electrical remodeling processes. Stress of the endoplasmic reticulum, protein catabolism, oxidative stress, and inflammation are some of the enriched processes. Among hubs-bottlenecks genes, which are highly connected and probably have a key role in regulating these processes, we found HSPA5, ANK2, CTNNB1, and VWF. Further experimental investigation of our findings may shed light on the pathophysiology of the disease and contribute to the identification of new therapeutic targets and treatments.