Cancer genomes are not fixed maps—they are flexible networks that constantly evolve. Recent studies reveal that up to 40% of human tumours contain extrachromosomal DNA (ecDNA)—circular, acentric DNA molecules carrying oncogenes such as MYC, EGFR, KRAS, and CDK4. Unlike chromosomal amplifications, these circles replicate independently and segregate unevenly during mitosis, creating variable oncogene copy numbers among daughter cells.
This unequal inheritance drives genetic diversity and fuels tumour adaptability under treatment pressure. In essence, ecDNA acts as the engine of cancer evolution—amplifying oncogenic expression and enabling rapid resistance. Understanding its origin and structure is therefore crucial for decoding the dynamics of tumour progression.
At the molecular level, ecDNA displays open chromatin architecture, functioning as transcriptional super-enhancers that boost oncogene activity far beyond chromosomal limits. Clinically, ecDNA-positive tumours are linked to poor prognosis and immune evasion. These unique properties have transformed ecDNA from an overlooked artifact into a defining feature of tumour plasticity.
In recent years, the discovery of extrachromosomal DNA (ecDNA) has revolutionized our understanding of tumor evolution. Once dismissed as artifacts, these circular, acentric DNA molecules—ranging from tens of kilobases to several megabases—have now been shown to carry potent oncogenes such as MYC, EGFR, KRAS, and CDK4. Unlike chromosomal amplifications, ecDNAs segregate unequally during mitosis, resulting in profound cell-to-cell variability in oncogene dosage. This randomness accelerates clonal diversification and endows tumors with extraordinary adaptability under therapeutic pressure.
Figure 1. Structural and Functional Features of ecDNA
From a molecular perspective, ecDNAs exhibit an open chromatin configuration, functioning as transcriptional super-enhancers that amplify oncogene expression far beyond chromosomal limits. Clinically, ecDNA-positive tumors are linked to aggressive phenotypes, immune evasion, and poor prognosis. Thus, decoding the structure and function of ecDNA is not merely a technical pursuit but a clinical necessity. However, their circularity, repetitive content, and complex rearrangements defy traditional sequencing frameworks optimized for linear genomes, necessitating novel approaches.
Sequencing ecDNA poses unique hurdles that extend beyond standard genomic analysis. Conventional next-generation sequencing (NGS) assumes linear templates and uniform coverage, while ecDNA molecules violate both assumptions. Their circular topology creates head-to-tail junctions that require specialized detection, and their formation through chromothripsis or breakage-fusion-bridge (BFB) cycles yields mosaics of rearranged chromosomal fragments.
Additional challenges include: (1) distinguishing ecDNA from intrachromosomal homogenously staining regions (HSRs); (2) maintaining DNA integrity during extraction, as circular molecules are prone to shearing; and (3) resolving large, repetitive structures that exceed typical read lengths. Overcoming these obstacles demands multimodal sequencing and analytical pipelines that integrate orthogonal evidence of circularity and amplification.
Conventional next-generation sequencing (NGS) assumes a linear genome. ecDNA defies that assumption. Circular topology and complex rearrangements demand methods that can detect, validate, and assemble complete ecDNA molecules.
An integrated sequencing framework now combines multiple technologies, each addressing a distinct aspect of the challenge:
By integrating these approaches, researchers can link structure to function—converting fragmented sequencing data into coherent biological insight.
Short-read WGS, predominantly via Illumina platforms, remains the foundation of ecDNA discovery in cancer genomics. By producing high-depth paired-end reads (~150 bp), WGS detects focal amplifications and structural variants (SVs) associated with ecDNA. Bioinformatics tools such as AmpliconArchitect (AA) and AmpliconClassifier analyze copy-number variation and breakpoint graphs to infer cyclic amplicons. These tools have identified ecDNA in up to 30–40% of human tumors across diverse cancer types.
Figure 2. Short-read WGS Workflow for ecDNA Analysis
However, WGS offers inference rather than direct observation. Short reads cannot span the full ecDNA circle, making it difficult to resolve sequence ordering, repeats, or multi-chromosomal mosaics. Ambiguities arise in regions with high homology or complex rearrangements. Despite these limitations, WGS provides a cost-effective, scalable first-pass tool for identifying ecDNA candidates in large cohorts, which can subsequently be validated by orthogonal methods.
Take the Next Step: Explore Related Services
Learn More
To directly detect circular DNA, enrichment-based methods such as Circle-Seq, CIDER-Seq, and exonuclease-based protocols have been developed. Circle-Seq digests linear DNA using exonucleases and selectively amplifies circular molecules via rolling-circle amplification (RCA), followed by sequencing. This approach enhances the signal-to-noise ratio and reveals rare ecDNA molecules otherwise masked by chromosomal DNA. Similarly, CIDER-Seq integrates circular enrichment with long-read sequencing to capture entire ecDNA molecules without PCR-induced artifacts.
Figure 3. Enrichment-Based and Circular DNA-Selective Sequencing Methods for
ecDNA Detection
While these techniques provide biochemical validation of circularity, they are biased toward smaller circles (<100 kb) due to size-dependent recovery and amplification efficiency. Moreover, RCA can generate chimeric concatemer artifacts that complicate downstream analysis. Combining Circle-Seq enrichment with short- or long-read sequencing mitigates some of these biases and enables more comprehensive ecDNA discovery.
Long-read sequencing technologies have transformed ecDNA research by enabling direct reconstruction of complex genomic architectures. Pacific Biosciences (PacBio) HiFi sequencing produces highly accurate reads (99.9%) up to 25 kb, while Oxford Nanopore Technologies (ONT) offers ultra-long reads exceeding 100 kb, occasionally spanning entire ecDNA molecules in a single read. These platforms overcome short-read limitations, allowing accurate assembly of repetitive, rearranged, or multi-chromosomal amplicons.
Figure 4. Schematic Representation of short-read WGS for the Analysis of ecDNA
For example, long-read analyses of glioblastoma and neuroblastoma have revealed ecDNA structures resulting from chromothripsis—catastrophic shattering of chromosomes followed by random re-ligation. Such insights demonstrate that ecDNA acts as a genomic "escape hatch," reorganizing oncogenic segments into self-replicating circles that drive tumor aggressiveness. Despite higher costs and input DNA requirements, long-read sequencing remains indispensable for defining ecDNA topology, breakpoint resolution, and epigenetic modifications, including methylation and nucleosome positioning.
No single platform fully captures ecDNA complexity. Hybrid strategies that integrate WGS, long-read sequencing, optical mapping, and chromatin assays yield the most complete reconstructions. AmpliconReconstructor (AR), for instance, combines NGS-derived breakpoint graphs with optical maps to scaffold ecDNA structures across hundreds of kilobases. Optical mapping visualizes DNA molecules stretched in nanochannels, offering megabase-scale confirmation of ecDNA circularity.
Figure 5. Hybrid and Multi-omic approaches for Comprehensive ecDNA Analysis
Functional integration via ATAC-seq (Assay for Transposase-Accessible Chromatin) and ChIP-seq (Chromatin Immunoprecipitation Sequencing) reveals that ecDNAs exhibit hyper-accessible chromatin and high transcriptional activity. ATAC-seq identifies "ecDNA hubs16" enriched in enhancer elements that regulate oncogene overexpression. Moreover, emerging single-cell ATAC-seq and Hi-C variants allow simultaneous profiling of ecDNA accessibility and nuclear topology, linking structure to transcriptional regulation within individual cells.
Bioinformatic tools have become central to ecDNA detection and characterization. AmpliconArchitect reconstructs cyclic amplicons from short-read data, while AmpliconClassifier distinguishes ecDNA from linear amplifications. ECCsplorer automates the analysis of Circle-Seq data, identifying candidate circular reads. For long-read datasets, CIDER-Seq and CoRAL assemble and validate complete circular sequences, whereas AmpliconReconstructor integrates optical and sequencing data for higher fidelity.
Circle-Map and Circle-finder reanalyze alignment files to identify split reads supporting circular junctions, offering computational validation of circularity. Visualization platforms such as Circos and Ribbon enable interpretation of complex rearrangement networks. Increasingly, AI-driven graph assembly algorithms and deep-learning models are being incorporated to improve structural prediction and reduce false positives in ecDNA detection.
| Method | Principle | Strengths | Limitations | Ideal Use |
|---|---|---|---|---|
| Short-Read WGS | Detects focal CNAs and breakpoints | Scalable, cost-effective, suitable for cohort studies | Indirect inference, limited structural resolution | Population-level ecDNA screening |
| Circle-Seq / CIDER-Seq | Enriches circular DNA molecules for sequencing | High sensitivity, biochemical validation of circularity | Bias toward small circles, amplification artifacts | Discovery of rare or low-abundance ecDNAs |
| Long-Read Sequencing | Directly sequences long fragments to resolve full structures | Resolves repeats and chromothripsis events; detects methylation | High cost, requires high-quality DNA | Structural validation and complete ecDNA reconstruction |
| Optical Mapping | Nanochannel-based imaging of DNA molecules | Provides megabase-scale structural confirmation | Specialized equipment, limited resolution for small circles | Validation of large and complex ecDNA architectures |
| ATAC/ChIP-seq | Profiles chromatin accessibility and histone modifications | Links structure to transcriptional activity | Does not reveal complete DNA sequence | Functional characterization of ecDNA-driven regulation |
| Hybrid / Multi-Omic Approaches | Combines WGS, LRS, and functional assays | Comprehensive view of structure and function | Complex data integration and cost | Multi-layered ecDNA profiling and mechanism discovery |
The next frontier in ecDNA research lies at the intersection of single-cell genomics, artificial intelligence, and clinical application. Single-cell DNA and ATAC sequencing now enable quantification of ecDNA copy number variability across individual tumor cells, revealing the stochastic inheritance patterns that underpin resistance evolution. Coupled with transcriptomic profiling, these approaches expose how ecDNA dynamically reshapes gene expression networks during therapy.
AI-assisted assembly algorithms promise to revolutionize ecDNA reconstruction by integrating multi-omic signals into unified structural maps. Deep-learning-based models trained on simulated ecDNA data can identify circular amplicons with high accuracy, even from noisy sequencing reads. Clinically, circulating ecDNA fragments detected via liquid biopsy offer non-invasive biomarkers for tumor burden and treatment response monitoring. As ecDNA-targeted therapies—such as replication inhibitors and synthetic lethality strategies—advance toward clinical testing, precise sequencing will become a cornerstone of personalized oncology.
The study of ecDNA epitomizes the convergence of genomics innovation and cancer biology. Short-read sequencing provides the foundation for detection, long-read and optical platforms resolve structure, and functional assays illuminate transcriptional impact. Computational frameworks synthesize these data into interpretable models, bridging molecular architecture with oncogenic function. As technologies mature, ecDNA sequencing will transition from discovery to clinical implementation, enabling precision diagnostics and therapeutic interventions that target the very engines of tumor evolution.
Extrachromosomal DNA, or ecDNA, refers to circular, acentric DNA molecules that exist outside the chromosomes. These elements often carry amplified oncogenes such as MYC, EGFR, and CDK4. They replicate independently of chromosomes and contribute to genetic diversity and therapy resistance in many tumours.
Large-scale genomic studies (e.g., Nature, 2019) have revealed that ecDNA occurs in approximately 30–40% of human tumours, although the prevalence varies by cancer type. Tumours such as glioblastoma and neuroblastoma often show especially high levels of ecDNA amplification.
ecDNA breaks the assumptions of conventional sequencing, which expects linear DNA templates and uniform coverage. Circular topology, repetitive regions, and complex rearrangements make ecDNA difficult to assemble. These features require combined use of short-read, long-read, and enrichment-based methods.
Researchers commonly combine several complementary technologies:
Together, these approaches provide a complete picture of ecDNA structure and activity.
Several bioinformatic pipelines support ecDNA reconstruction and classification, including AmpliconArchitect, AmpliconClassifier, ECCsplorer, CIDER-Seq, and Circle-Map. These tools identify circular junctions, visualise rearrangements, and distinguish ecDNA from linear amplifications.
ecDNA can drive oncogene overexpression, rapid tumour evolution, and resistance to targeted therapies. Detecting ecDNA helps researchers understand tumour heterogeneity and may provide new biomarkers for monitoring disease progression through liquid biopsy or other non-invasive methods.
Single-cell DNA and ATAC sequencing allow scientists to measure ecDNA copy number variation between individual tumour cells. This reveals how circular DNA segregates randomly during cell division, fuelling genetic heterogeneity and resistance evolution within a tumour population.
The field is moving toward multi-omic integration and AI-assisted reconstruction. Deep learning algorithms are improving circle assembly from complex datasets, while liquid biopsy applications are extending ecDNA detection beyond tissue samples. These advances promise to connect ecDNA dynamics with clinical outcomes in precision oncology.
References
CD Genomics is transforming biomedical potential into precision insights through seamless sequencing and advanced bioinformatics.