Integrated bioinformatics and network search identify biomarkers and comorbid mechanisms of atrial fibrillation and ischemic stroke

This study combines bioinformatics, network search, and immune infiltration analysis to explore the mechanisms and biomarkers underlying the comorbidity of atrial fibrillation (AF) and ischemic stroke (IS). The Gene Expression Omnibus database acquired 2 microarray datasets each for AF and IS. After removing batch effects and merging the data, potential comorbid targets were identified through differential expression analysis and weighted gene co-expression network analysis. These targets were then explored using a protein–protein interaction network to identify comorbid targets with similar functions, followed by enrichment analysis. Subsequently, hub genes were determined via Lasso regression, and predictive models based on these genes were constructed for both IS and AF to assess the models’ predictive capabilities. Finally, immune infiltration analysis was conducted to uncover the comorbid mechanisms of IS and AF from an immunological perspective. From the AF dataset, 2063 differentially expressed genes were identified, along with 4776 genes from the most relevant turquoise module. From the IS gene set, 947 differentially expressed genes were identified, along with 1900 genes from the most pertinent blue module, yielding 31 intersecting genes as potential comorbid targets. Based on the protein–protein interaction network, a restart random walk algorithm focused on these 31 potential targets identified 50 related comorbid targets, primarily enriched in processes such as fatty acid metabolism, peroxisome proliferator-activated receptor signaling pathway, and ErbB signaling pathway. Hub genes cortactin , FOS , JUN , and cathepsin B were identified through Lasso regression and exhibited excellent predictive performance in both AF and IS datasets. Immune infiltration analysis revealed a higher infiltration abundance of monocytes, CD8 + T cells, and neutrophils in AF and IS samples. This study clarifies the comorbidity mechanisms of AF and IS, linking them to lipid metabolism and activated peroxisome proliferator-activated receptor/ErbB signaling pathways. Hub comorbid genes FOS , JUN , cortactin , and cathepsin B were identified, and the AF/IS early prediction models built on these genes show excellent performance, meriting clinical focus. Additionally, neutrophils, monocytes, and CD8 + T cells may participate in the shared pathogenesis of AF and IS, though whether hub genes regulate immune cell infiltration to affect AF/IS requires further experimental validation.