Inferring Chromatin Structure from Cell-Free DNA Fragmentation Patterns

This direction is merged into another direction: fragment length analysis. Its last version stays readable and citable here.

Version 1 of 122 papers46 conceptsfrozen 2026-10-04

When cells die, they release fragments of DNA into the bloodstream and other bodily fluids. For years, researchers treated this cell-free DNA (cfDNA) primarily as a source of genetic sequences—a way to find mutations without a tissue biopsy. However, a growing body of research suggests that the physical properties of these fragments—their lengths, where they start and end, and their jagged edges—contain a rich "fragmentomic" record of the source cell’s epigenetic state.

By analyzing these patterns, researchers can now reconstruct the 3-dimensional architecture of chromatin and the methylation status of the original tissue, effectively using blood and urine as a high-resolution window into the body’s internal regulatory landscape.

The Nucleosome Footprint

The core intuition behind fragmentomics is that DNA fragmentation is not a random process. In a living cell, DNA is wrapped around proteins called histones to form nucleosomes. This packaging acts as a physical shield. When a cell undergoes apoptosis (programmed cell death), enzymes like DNASE1 chop the DNA into pieces.

As Pavel Stejskal and colleagues explain, these enzymes preferentially cut the "linker" DNA between nucleosomes while the DNA wrapped around the histone core remains protected. This creates a characteristic size distribution: cfDNA fragments often peak at roughly 167 base pairs, the length of DNA required to wrap around a single nucleosome plus a small linker. Even extrachromosomal circular DNA (eccDNA) follows this rule; Du Wu found that small circular DNAs in adipose stem cells often cluster around 200–240 base pairs, reflecting the physical constraints of nucleosomal packaging.

Methylation as a Driver of Fragmentation

The relationship between a fragment's length and its chemical state is a powerful diagnostic tool. Taylor J. Jensen and Cosmin Deciu discovered that unmethylated DNA fragments tend to be significantly shorter than methylated ones.

This happens because unmethylated regions usually correspond to "open" chromatin—areas where the DNA is less tightly packed and more active. Because these regions are more accessible, they are more susceptible to enzymatic cleavage. This discovery allowed researchers to identify "Placenta Hypomethylated Domains" (PHDs), which distinguish fetal DNA from maternal DNA in pregnancy, and to differentiate tumor-derived DNA from healthy background DNA.

In cancer, this signal becomes even more pronounced. A. Rose Brannon and the team behind the MSK-ACCESS assay demonstrated that tumor-derived fragments are consistently shorter than those from healthy cells. This physical signature provides an "orthogonal" layer of evidence: if a mutation is found on a particularly short fragment, it is much more likely to be a true somatic tumor mutation rather than a sequencing error or a benign age-related mutation.

Reconstructing the 3D Genome from Liquid Biopsies

If fragmentation patterns reflect nucleosome positions, then the density of fragment "ends" can be used to map the entire landscape of chromatin accessibility. Yaping Liu and colleagues developed a method called FREE-C that treats cfDNA as a "liquid biopsy" of 3D chromatin organization.

By mapping where DNA fragments break, they can identify which parts of the genome were active (open) or inactive (closed) in the source tissue. This allows for "tissue-of-origin" mapping: because a liver cell and a lung cell have different open chromatin regions, their cfDNA "breakage maps" look different. Pengfei Yu showed that analyzing these patterns at transcription start sites can even serve as a proxy for gene expression, allowing for the detection of gastric cancer with high sensitivity without needing to sequence the entire genome at high depth.

Decoding the "Jagged" Ends

The information isn't just in the length of the fragment, but in the fine structure of its ends. When enzymes cut DNA, they don't always leave clean, blunt ends. Instead, they often leave "jagged ends"—single-stranded overhangs.

Tingting Xie and Suk Hang Cheng discovered that urinary cfDNA has a unique 10-nucleotide periodicity in these jagged ends. This periodicity is a physical echo of the helical pitch of DNA as it wraps around a histone. In patients with renal cell carcinoma, these structural signatures change, providing a highly accurate diagnostic signal that goes beyond simple sequence mutations. Similarly, Meihui Chen showed that the presence of specific "end motifs" (the specific sequence of nucleotides where the cut occurred) can identify which specific nucleases, such as DNASE1, were active, further narrowing down the biological source of the DNA.

Predicting Epigenetics Without Chemical Treatment

Traditionally, measuring DNA methylation requires harsh chemical treatments like bisulfite conversion, which can destroy the very DNA fragments researchers are trying to study.

The latest frontier in this field, exemplified by the FinaleMe model developed by Yaping Liu and Sarah C. Reed, uses machine learning to predict methylation directly from fragmentation patterns. By analyzing the spatial positioning of CpG sites relative to the center of a fragment, the model can infer whether a site was methylated or not. This effectively turns standard, non-destructive whole-genome sequencing into an epigenetic assay, unlocking hidden regulatory information from millions of existing datasets.

Papers in this version

History & citationVersion 1 · 2026-10-04

Merged into another direction · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_inferring_chromatin_structure_from_cell__16be813c,
  title = {Inferring Chromatin Structure from Cell-Free DNA Fragmentation Patterns},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/inferring-chromatin-structure-from-cell-free-dna-fragmentation-patterns/rdr_16be813cd05a4e8a8067091ec39b8fd2},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}

This exact version (version 1, frozen text):

@misc{lacuna_inferring_chromatin_structure_from_cell__16be813c_v1,
  title = {Inferring Chromatin Structure from Cell-Free DNA Fragmentation Patterns},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/inferring-chromatin-structure-from-cell-free-dna-fragmentation-patterns/rdr_16be813cd05a4e8a8067091ec39b8fd2/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}