Cell-free DNA (cfDNA) fragmentomics refers to the quantitative and qualitative analysis of DNA fragments circulating in blood, particularly the patterns of fragment sizes, ends, and genomic coverage that reflect underlying tissue-specific biology. Unlike conventional liquid biopsy approaches that focus mainly on variant allele frequencies (e.g., single-nucleotide variants), fragmentomics extracts “how the DNA is cut and released” information. This matters because cfDNA is not merely a passive biomarker; it carries structural signatures shaped by apoptosis, necrosis, and active secretion, as well as by the chromatin landscape of the tissue of origin.
Mechanistic basis begins with the observation that nucleosome organization and DNA damage/repair processes influence cfDNA generation. In typical apoptotic pathways, DNA is cleaved in a nucleosome-aligned manner, producing characteristic fragment size distributions (often enriched around ~160–180 bp mononucleosome lengths and their multiples). Tumor cells can alter these fragmentation patterns through dysregulated cell death pathways, increased replication stress, and altered chromatin remodeling. The resulting cfDNA fragments therefore show measurable differences in fragment size peaks, end motifs, and genomic coverage biases. Fragmentomics may also leverage fragment end positions and preferred cleavage sites, yielding inference about nuclease activity and nucleosome positioning.
In practice, fragmentomics is implemented via high-depth sequencing and specialized bioinformatic pipelines. Key outputs include: (1) fragment length histograms across the genome; (2) nucleosome footprint-like signatures using end position distributions; (3) “coverage” patterns that may highlight tissue-of-origin chromatin states; and (4) integration with other modalities such as copy-number variation or methylation. Because fragmentomics is sensitive to pre-analytical variables, rigorous sample handling is essential. Factors such as time-to-processing, tube type, centrifugation protocol, and freeze-thaw cycles can change the cfDNA size profile by increasing background DNA contamination from lysed blood cells. Therefore, clinical adoption requires standardized collection and QC metrics.
A major clinical objective is improving early cancer detection and monitoring. Early malignancies often shed low quantities of cfDNA, and variant-based assays may have limited sensitivity when tumor fraction is extremely low. Fragmentomic features can sometimes detect subtle shifts in fragmentation and tissue-of-origin signal even when canonical mutations are scarce. The approach can be used alongside error-suppressed sequencing and targeted panels to improve effective signal-to-noise. Additionally, fragmentomics can potentially support dynamic monitoring: treatment-induced tumor cell death may change cfDNA fragmentation distributions, enabling detection of response or recurrence earlier than imaging in some settings.
Another domain is identifying tissue of origin and clarifying biological context. Since chromatin architecture differs between tissue types, fragment end patterns and size distributions can be used as a proxy for where cfDNA originated. This may help distinguish cancers from other sources of cfDNA shedding such as inflammation, autoimmune activity, trauma, or recent procedures. These non-malignant sources can confound liquid biopsy results; fragmentomics may reduce false positives by recognizing patterns inconsistent with tumor-associated fragmentation.
Despite promise, fragmentomics faces scientific and translational challenges. First, cfDNA is heterogeneous: it includes fragments from multiple cell types and processes concurrently. Second, blood-based confounders—age, BMI, comorbidities, medications, and inflammatory states—may affect cfDNA kinetics and fragmentation. Third, sequencing depth and computational complexity can limit throughput and raise costs. Fourth, model generalizability across cohorts and platforms is crucial; fragmentomic signatures may vary with library preparation chemistry and bioinformatic choices. Robust external validation and prospective clinical trials are required before fragmentomics can be considered a routine diagnostic tool.
Future directions include multimodal integration. By combining fragmentomic features with methylation, copy-number profiles, fragment end motifs, and unique molecular identifier strategies, models may better estimate tumor fraction and predict clinical outcomes. Machine learning approaches can fuse these features to derive risk scores for screening, prognosis, and treatment selection. Additionally, increasing focus on longitudinal sampling will help interpret fragmentomic trajectories rather than relying on single time points.
Overall, cell-free DNA fragmentomics represents an evolution of liquid biopsy from purely genetic to structural and process-informed genomic readouts. By capturing nucleosome-associated cleavage patterns and end signatures shaped by tumor biology, it offers a route to detect malignancy earlier, monitor response more sensitively, and improve biological interpretation of circulating DNA. Source: [Creator/Source]
Source: Kevin Chua (@drkevin_chua_lm) via ASCOBT26 day 3 keynote mention.
Kevin Chua: #ASCOBT26 day 3 kicks off with a keynote address by Professor Dennis Lo on the emerging field of cell-free DNA fragmentomics. @ASCO @DrMLChua. #breaking
— @drkevin_chua_lm May 1, 2026
SHOP AMAZON BEST SELLERS, CLICK TO BUY FROM AMAZON.
SHOP AMAZON BEST SELLERS, CLICK TO BUY FROM AMAZON.
