Atrial fibrillation, or AF for short, is a common persistent arrhythmia, and more than 33 million people across the world suffer from this disease; advancing age can greatly raise people’s risk of developing AF. AF patients often have typical symptoms such as palpitations, chest tightness and reduced exercise tolerance, and these patients also face higher risks of ischemic stroke, heart failure and cardiovascular-related death [1]. As research on atrial fibrillation continues to progress, catheter ablation, anticoagulation therapies and anti-arrhythmic drugs have achieved great improvements; still, pathological mechanisms of AF are complicated and vary greatly among individuals, and many patients obtain poor treatment effects or experience AF recurrence after clinical intervention. Researchers have not fully clarified the complete pathogenesis of atrial fibrillation, and it is relatively difficult to carry out studies about how genetic variants regulate gene expression. It is of great practical significance to fill this research gap; if relevant studies make major breakthroughs, clinicians can obtain new biomarkers and discover new therapeutic targets for diseases. Traditional research models can hardly explain the complex genetic features behind atrial fibrillation. The onset and continuous progression of AF are affected by many different factors; these factors include atrial electrophysiological remodeling driven by ion channel dysfunction and action potential changes, as well as atrial structural remodeling, tissue fibrosis, massive extracellular-matrix deposition, activated inflammatory responses and disturbed body metabolism. Genome-Wide Association Studies (GWAS) have identified many AF risk loci, including PITX2, TBX5, ZFHX3 and PRRX1; these genes participate in atrial development, myocardial signal conduction and transcriptional regulation. Yet most known risk loci are located in non-coding regions, and these sites may affect gene expression by acting on enhancers and promoters [2]. Epigenetic mechanisms such as DNA methylation, histone modification, chromatin remodeling and non-coding RNAs can regulate gene expression without changing DNA sequences, and they further get involved in the occurrence of atrial fibrillation. High-throughput sequencing technologies drive the development of cardiovascular epigenetics; ATAC-seq uses Tn5 transposase to draw genome-wide open-chromatin maps, and this technology can locate promoters, enhancers and transcription-factor-binding sites [3]. RNA-seq is able to detect transcriptome-wide expression changes, and it provides experimental evidence for researchers to analyze disease-related signaling pathways [4]. When researchers use these two sequencing technologies together, they can correlate the open status of regulatory regions with gene expression levels, and they can better reconstruct transcription-factor networks and sort out disease-relevant signaling pathways.

Such multi-omics approaches are being adopted with increasing frequency in complex diseases, and for AF research they supply a regulatory framework. For example, integrated analyses have demonstrated that AF risk variants are enriched in cardiac-specific open chromatin; deleting a single regulatory region at Tbx5, for instance, can shift gene dosage and modify arrhythmia susceptibility [5][6].

This review covers its application in AF research: disease-related regulatory regions, candidate genes, genetic risk loci, cell type-specific mechanisms, and therapeutic targets. It also discusses challenges (sample heterogeneity, data integration, causality) and prospects (single-cell multi-omics, spatial omics, AI) for AF precision medicine. For our literature search, we queried PubMed, Web of Science, and Google Scholar for studies published between January 2015 and September 2026, using the terms: (“atrial fibrillation” OR “AF”) AND (“chromatin accessibility” OR “ATAC-seq” OR “scATAC-seq”) AND (“transcriptome” OR “RNA-seq” OR “scRNA-seq”) AND (“multi-omics” OR “integrative analysis”). Studies were included if they 1) used human or animal atrial tissue, 2) combined chromatin accessibility with transcriptome data, and 3) reported AF-related regulatory mechanisms or targets. Studies using only single-omics or non-cardiac tissues were excluded. We categorized the included studies into paired multi-omics (same nuclei/samples) and independent dataset integration.

In eukaryotic cells, DNA is not naked but packaged into chromatin, winding around histone octamers to form nucleosomes. Two principal states are recognized: open euchromatin and closed heterochromatin. The open state tends to be enriched in promoters, enhancers, insulators, and other regulatory elements, which allows transcription factors to bind and activate genes [7]. Shifts between these states lie at the heart of gene expression regulation. In AF, atrial tissue responds to electrical, mechanical, and inflammatory stress by remodeling its chromatin, a process that alters transcriptional networks and implies that accessibility changes help govern disease-related genes. ATAC-seq takes advantage of a hyperactive Tn5 transposase, which fragments DNA and inserts sequencing adapters into open chromatin at the same time. Because stable nucleosome structures do not exist inside open chromatin areas, the Tn5 enzyme tends to insert itself into these accessible genome segments; once researchers finish the sequencing work, they carry out peak-calling steps to mark out these open genome segments, and scientists can later assign annotations to these marked areas as promoters, enhancers or binding locations for transcription factors [8]. ATAC-seq shows higher detection sensitivity, needs smaller amounts of starting sample material, and follows a more straightforward operating pipeline, if people make comparisons against DNase-seq and FAIRE-seq; many research projects have adopted this technique to find regulatory segments unique to cardiomyocytes, and researchers can also use genetic datasets to build links between non-coding sequence changes and their actual biological roles [9]. Even so, results generated from ATAC-seq only reflect possible regulatory capacity, and this method cannot give direct descriptions for real-world gene expression status, so research teams have to bring in transcription-related datasets, to judge whether one given open genome segment can really change the activity of its downstream target genes. On the other hand, RNA-seq lets investigators track expression shifts over the full set of transcript sequences; researchers can pick out differentially expressed genes (DEGs) by comparing transcript profiles collected under varied experimental settings, and enrichment tests alongside pathway analysis help workers figure out the related biological processes. For atrial-fibrillation-related research, RNA-seq is capable of spotting abnormal signals linked with inflammatory responses, extracellular matrix reconstruction, energy turnover and the contraction performance of heart muscle cells [10][11]. Higher gene expression values may come from activated enhancer regions, strengthened attachment of transcription factor molecules or other epigenetic control patterns; researchers cannot get direct explanations for upstream controlling factors, if they only make use of RNA-seq outputs, thus transcriptome data on its own is not enough for people to work out complete regulatory networks for complicated illnesses. When researchers use ATAC-seq alongside RNA-seq, they can merge information about how open regulatory segments are together with gene expression records, and this practice builds up connections among regulatory DNA fragments, transcription factor molecules and their corresponding target genes. A genome segment is likely to serve as an active regulatory fragment, when this region gains higher accessibility meanwhile its neighboring genes show elevated expression levels; investigators can further locate disease-promoting molecular networks, once they add motif forecasting, enhancer labeling and risk-site examination into their analytical workflow. This combined analytical plan using ATAC-seq and RNA-seq can therefore not only discover differentially expressed genes (DEGs), it can also show researchers which regulatory DNA regions have gone through alterations, which transcription factors exert control over these regions, and how these factors alter the expression of illness-relevant genes, allowing people to clarify epigenetic working mechanisms behind atrial fibrillation.

Researchers can uncover the disease-causing mechanisms of AF, when they combine chromatin accessibility data together with transcriptome analysis, yet plenty of practical obstacles still exist for present-day research work.

The first problem lies in the heterogeneity of samples, which is the main obstacle. Most studies use human atrial tissues obtained during cardiac surgery, and the sample size is usually small. Meanwhile, patient age, co-existing illnesses, particular forms of atrial fibrillation, and the clinical interventions given to AF patients, can all shape chromatin conditions as well as gene expression profiles, and this situation lowers result repeatability across separate study groups [6][16]. Besides that, a range of technical variables and biological conditions can create extra disturbance, when researchers carry out comparisons between atrial samples taken from AF subjects and non-AF subjects. Investigators generally collect atrial tissue samples from several common positions, and these positions cover the left atrial wall, right atrial wall and atrial free wall; sample sites will bring about differences in cell make-up, fibrosis severity and chromatin profiles, and such differences will add position-related bias to experimental outputs. For the control subjects included in our research, many of them suffer from other cardiovascular disorders, for instance valvular heart injury and heart failure; these extra health issues are able to alter chromatin states and gene expression on their own. What is more, atrial fibrillation will bring shifts in cellular components as the disease goes on, it will cut down the number of cardiomyocytes, while fibroblasts and immune cells will grow in quantity, and these changes may produce misleading transcription and epigenetic signals for the whole tissue sample. Single-cell and single-nucleus testing tools can ease this trouble to some degree, since these tools allow researchers to separate signals belonging to distinct cell types, yet these approaches still have drawbacks; problems like biased cell separation and unsatisfactory nucleus capture rate will affect final data. The second major difficulty comes from inadequate tests for causal relationships; ATAC-seq and RNA-seq are capable of showing correlations between accessible chromatin segments and gene expression levels, but these observed correlations cannot confirm real causal links. Researchers need to run practical functional tests, such as CRISPR gene editing, chromatin interaction assays including Hi-C and Capture Hi-C, plus in-cell and animal model experiments, so they can confirm whether one certain open chromatin segment really controls its target gene [17]. The third main limitation is the absence of uniform processing pipelines; researchers adopt different peak searching settings, data normalization strategies and data combination algorithms; these differences reduce the comparability among experimental outputs and give rise to conflicting findings.

When research workers make use of single-cell multi-omics tools that contain scATAC-seq and scRNA-seq, they can examine cell-specific regulatory networks with higher accuracy, and uncover how cardiomyocytes, fibroblasts, endothelial cells together with immune cells work jointly during AF onset and disease progression [6]. Spatial transcriptome and epigenome methods can supply positional data for individual cells, and they help illustrate how cell locations interact with the diseased microenvironment inside atrial tissue. Artificial intelligence alongside machine learning methods gets more frequent use for merging different sets of multi-omics data; deep learning models can combine datasets with large internal differences, and raise the reliability for AF risk evaluation as well as potential target identification [18]. If multi-omics datasets are combined with clinical symptom records, imaging results and electrophysiology measurements, people can build multi-layer prediction models, and such models will boost accuracy for AF diagnosis and patient treatment. To draw a brief conclusion, joint measurement of chromatin accessibility and transcript information has turned into one key way to study molecular mechanisms behind atrial fibrillation; as multi-omics kits, computing algorithms and functional verification tools become more reliable, this analytical method can describe AF-related regulatory networks in a more complete manner, and lay more solid groundwork for early-stage disease detection, biomarker screening and targeted clinical therapy.

*Corresponding author.