Atrial fibrillation (AF), the most common cardiac arrhythmia, affects millions worldwide and is increasing in prevalence due to aging populations and improved detection. While effective diagnostic and therapeutic technologies exist, life-threatening complications like stroke and heart failure remain significant problems. Besides established clinical risk factors (aging, obesity, hypertension, heart failure), a substantial genetic contribution to AF is recognized. Previous GWASs have identified over 100 AF-associated loci, implicating cardiac developmental, electrophysiological, contractile, and structural pathways. However, most GWASs have been conducted in European populations, leaving the genetic architecture of AF in other populations less understood. This study aimed to investigate AF genetics in a non-European population (Japanese) by performing a large-scale GWAS and a subsequent cross-ancestry meta-analysis, enhancing statistical power. The researchers also investigated the biological roles of identified loci using gene expression and epigenomic data, developing a polygenic risk score (PRS) to assess its impact on relevant phenotypes and long-term mortality. This work aims to provide new insights into AF genetics and explore the clinical utility of genetic risk prediction for AF.

The existing literature demonstrates a strong genetic component in atrial fibrillation (AF), with previous genome-wide association studies (GWAS) identifying over 100 susceptibility loci. These loci are linked to various pathways including cardiac development, electrophysiology, and structure. However, most studies have focused on populations of European ancestry, leaving the genetic architecture of AF in other populations less well-defined. The limited understanding of AF genetics in non-European populations hinders the development and application of polygenic risk scores (PRS) for risk stratification and prediction in these groups. There is also a need for further investigation into the biological mechanisms underlying AF susceptibility, which requires integration of GWAS findings with functional genomic data.

The study employed a two-stage design. First, a GWAS was conducted on a Japanese cohort from BioBank Japan (BBJ) comprising 9,826 AF cases and 140,446 controls. Genotyping was performed using Illumina arrays, followed by imputation using the 1000 Genomes Project Phase 3 reference panel. Association analysis was performed using logistic regression, adjusting for age, sex, and principal components to account for population stratification. Five novel AF risk loci were identified and subsequently replicated in an independent Japanese cohort. Second, a cross-ancestry meta-analysis was performed, combining the Japanese GWAS data with summary statistics from large-scale European GWASs, yielding a total of 77,690 cases and 1,167,040 controls. The MANTRA algorithm was used to account for ancestry heterogeneity. This meta-analysis identified 150 genome-wide significant loci, including 33 novel loci. Downstream analyses included transcriptome-wide association studies (TWAS) using GTEx data to identify candidate causal genes, enrichment analysis of ChIP-seq data to identify enriched transcription factors, and functional validation using human induced pluripotent stem cell-derived cardiomyocytes (iPSCMs). A polygenic risk score (PRS) was constructed from the cross-ancestry meta-analysis results, and its performance in predicting cardiovascular and stroke mortality was evaluated using long-term follow-up data from BBJ. Finally, Mendelian randomization (MR) analysis was performed to investigate the causal relationships between AF and other cardiovascular diseases and quantitative traits.

The Japanese GWAS identified 31 genome-wide significant loci associated with AF, including five novel loci. Two of these novel loci contained rare variants (rs202030113 and rs778479352) specific to East Asian populations, located within *SYNEL* and *FGF13*, genes implicated in nuclear envelope function and sodium channel activity respectively. The cross-ancestry meta-analysis identified a total of 150 genome-wide significant loci, with 33 being novel. TWAS analysis prioritized *IL6R* as a potential causal gene, highlighting the role of immune responses in AF pathogenesis. Integrative analysis using ChIP-seq data identified *ERRγ* as a key transcription factor regulating AF-associated genes. Functional studies using iPSCMs confirmed *ERRγ*'s role in modulating gene expression related to ion channels and sarcomere function. The PRS derived from the cross-ancestry meta-analysis showed robust predictive ability for cardiovascular and stroke mortality. It significantly predicted increased risk of cerebral infarction and cardioembolic stroke, even in individuals without a diagnosed AF, suggesting potential for identifying subclinical AF or related conditions. Individuals with high PRS had significantly younger age of AF onset and increased risks of cardiovascular and stroke related mortality.

This study significantly advances our understanding of the genetic architecture of atrial fibrillation (AF) by identifying numerous novel susceptibility loci across multiple ancestries. The discovery of East Asian-specific rare variants in *SYNEL* and *FGF13* highlights the importance of considering population-specific genetic variations when studying complex diseases. The identification of *IL6R* and *ERRγ* as key players in AF pathogenesis provides valuable insights into the underlying biological mechanisms, specifically highlighting the involvement of immune responses and transcriptional regulation of ion channels and sarcomere function. The successful development and validation of a highly predictive PRS underscores the potential for personalized risk stratification and early detection of AF, particularly cardioembolic stroke in individuals without prior diagnosis. This opens avenues for preventive strategies in high-risk individuals.

This large-scale, cross-ancestry GWAS and subsequent analyses have substantially increased our understanding of the genetic basis of atrial fibrillation. The identification of novel loci, key genes (*IL6R*), and transcription factors (*ERRγ*), combined with the development of a highly accurate polygenic risk score, provides crucial new insights into AF pathogenesis and risk prediction. Future research should focus on further elucidating the functional roles of the identified genes and pathways, and on refining the PRS for improved clinical utility in diverse populations.

While this study represents a significant advancement in AF genetics, some limitations should be acknowledged. The primarily Japanese and European ancestry of the samples may limit the generalizability of findings to other populations. The functional studies conducted using iPSCMs, while informative, may not fully replicate the complexity of in vivo processes. Finally, the long-term follow-up data used to assess PRS performance were collected from a single biobank, potentially influencing the generalizability of its predictive power.