The technical hurdle the team faced is one that has haunted genome biologists for years: the three-dimensional arrangement of enhancers and promoters at the scale of individual nucleosomes—the fundamental spools around which DNA wraps—has remained essentially invisible. Popular chromosome conformation capture methods such as Hi-C average over millions of cells and blur out anything smaller than tens of kilobases. Super-resolution imaging can reach the nanoscale, but only for a handful of selected loci at a time. To bridge the gap, the researchers developed a simulation framework that leverages region-capture Micro-C contact maps, an assay that zooms in on megabase-scale windows with nucleosome-level resolution, to infer full conformational ensembles of genomic regions.

A central innovation of the work is a data-processing step the authors call neighbor balancing. Standard Micro-C analysis assumes that contact density is uniform across the genome, an assumption that quietly erases one of the most biologically meaningful signals: the fact that some stretches of DNA are simply more densely packed, and therefore more frequently in contact, than others. The new balancing strategy identifies variation in contact density instead of normalizing it away. When the team applied the method, sharp dips in contact density appeared precisely at active promoters and enhancers—places where the chromatin fiber opens up and nucleosomes thin out. These dips had been smoothed over in coarser, genome-wide datasets, but region-capture data preserved them, and neighbor balancing made them interpretable.

With the corrected contact maps in hand, the researchers turned to a maximum-entropy inversion approach, a technique rooted in statistical mechanics that finds the least-biased set of polymer structures consistent with the experimental constraints. The resulting simulated ensembles were then put through a battery of validation tests. The simulated structures reproduced pairwise distance distributions measured independently by chromatin tracing, a super-resolution imaging method that follows the positions of dozens of genomic loci in single cells. The simulations also contained packing domains and nucleosome clutches—discrete clusters of neighboring nucleosomes—matching structures that electron microscopy and super-resolution imaging studies had previously observed directly in cells. In other words, the model did not merely fit the data it was trained on; it recovered structural features seen by entirely different experimental techniques.

The most striking finding emerged when the team examined what those structures actually look like. Far from being uniformly open, euchromatin generally forms condensed domains with packing densities comparable to those of heterochromatin. The difference lies in scale: euchromatic domains are smaller than their heterochromatic counterparts, but the chromatin fiber inside them is nearly as tightly packed. Local nucleosome concentrations inside domain interiors frequently exceeded 400 micromolar, an extraordinarily crowded environment for DNA. This result aligns with a growing body of imaging evidence suggesting that so-called open chromatin is condensed but liquid-like in living cells, and it forces a rethinking of what the light and dark bands of classical cytology actually represent.

If euchromatin is mostly dense, how do genes get read? The answer, according to the simulations, lies in the geometry of the domain surfaces. Kilobase-scale regions at promoters and enhancers often protrude from the condensed domains, extending outward into the surrounding nuclear space where they become highly accessible. The team found that protrusion probability, the positions of clutch boundaries, and local contact density all correlate closely with ATAC-seq coverage, a genome-wide measure of chromatin accessibility. Where the fiber pokes out of a domain, transcription factors can find their binding sites; where it is buried inside, it cannot. Accessibility, in this picture, is not a bulk property of open chromatin but a surface phenomenon.

This arrangement effectively compartmentalizes regulatory elements from the surrounding chromatin, and the authors argue that the geometry serves a functional purpose. By lifting enhancers and promoters onto domain surfaces, the cell facilitates protein binding and enhancer–promoter communication, the long-range conversations between distant regulatory elements that switch genes on. The condensed cores may act as scaffolds that keep related regulatory elements in physical proximity, while the exposed protrusions provide the docking sites. The study also found that both packing domains and clutches tend to be smaller around ATAC-seq peaks, suggesting that local accessibility reshapes chromatin organization at multiple scales simultaneously.

The hierarchical picture that emerges runs from nucleosomes to clutches to domains, and it holds across species and cell types. The team validated their framework in mouse embryonic stem cells and in several human cell lines, finding condensed domains throughout. Even when the researchers examined data from cells depleted of cohesin, the ring-shaped protein complex that extrudes DNA loops, the local structure was largely maintained: chromatin expanded at large scales, but clutch-scale organization and the correlation between protrusions and accessibility persisted. This indicates that the condensed-domain architecture of euchromatin is not simply a byproduct of loop extrusion but reflects more fundamental physicochemical interactions within the chromatin fiber itself.

The implications reach well beyond structural biology. Misregulated enhancer–promoter communication underlies many developmental disorders and cancers, and drugs that target chromatin regulators, such as bromodomain inhibitors, are already in clinical use. A model in which accessibility depends on whether a regulatory element sits on a domain surface offers a concrete structural hypothesis for how such drugs work and why their effects are so context-dependent. It also reframes a long-standing question in the field: rather than asking how euchromatin stays open, biologists may now ask how specific regulatory elements are actively extruded or maintained on domain surfaces, and what molecular machinery performs that positioning.

The study’s computational framework, along with its simulation trajectories and analysis code, has been made publicly available, allowing other groups to apply the approach to their own region-capture datasets. As region-capture Micro-C spreads through the genomics community, the MIT team’s method could become a standard tool for converting contact maps into physically realistic, nucleosome-resolution structures. What began as a technical exercise in matrix balancing has delivered a conceptual shift: the active genome is not a loose tangle waiting to be read, but a dense, well-organized material whose most important working parts hang, deliberately and accessibly, on the outside.

Subject of Research: Three-dimensional organization of euchromatin and regulatory elements at nucleosome resolution

Article Title: Euchromatin forms condensed domains with short active regions on the surface

Article References: Paggi, J. M., Long, L. Y., & Zhang, B. (2026). Euchromatin forms condensed domains with short active regions on the surface. Nature Genetics. https://doi.org/10.1038/s41588-026-02775-9

Image Credits: AI Generated

DOI: 10.1038/s41588-026-02775-9

Keywords: chromatin, euchromatin, heterochromatin, nucleosomes, Micro-C, enhancers, promoters, gene regulation, 3D genome organization, polymer simulations, chromatin accessibility, super-resolution imaging

The central insight behind single-molecule genomics, often abbreviated SMG, is that conventional genomic assays average their measurements across millions of DNA fragments, obscuring the very heterogeneity that drives genome function. A bulk chromatin accessibility profile, for example, might show that half of the molecules covering a promoter are open, but it cannot reveal whether the same half of molecules also carry a particular histone modification or bound transcription factor. SMG techniques change this picture fundamentally. By profiling epigenetic modifications, transcription factor binding and chromatin organization on single DNA molecules, they can quantify molecular heterogeneity and the co-occurrence of regulatory events genome-wide. This makes it possible to ask which combinations of regulatory marks travel together on the same molecule, a question that bulk methods are mathematically incapable of answering. The authors argue that this co-occurrence information reveals dynamics of chromatin interactions that cannot be measured by conventional genomics assays, and that SMG therefore offers a unique platform to study how regulatory events combine to control genome activity.

The technical foundations of the field draw on two complementary sequencing regimes. Short-read sequencers, which generate reads typically between 150 and 600 base pairs, provide accurate and high-throughput analysis of genomic DNA but fragment the molecule-level context. Long-read sequencing technologies, including nanopore platforms, generate reads spanning several to tens of kilobases, enabling the analysis of long-range chromatin features, repetitive regions and structural variation while simultaneously detecting endogenous or experimentally introduced DNA base modifications on native, unamplified DNA. The choice between these regimes shapes what a SMG experiment can deliver: short reads give statistical power and base-level resolution within a limited window, while long reads preserve multi-kilobase architectural context, allowing researchers to connect distal enhancers, promoters and silencers into coherent regulatory states on individual chromatin fibers.

Among the methods the guide surveys, single-molecule footprinting holds a foundational place. First introduced genome-wide in 2017 by the Krebs laboratory, the technique uses DNA methyltransferases to tag accessible DNA in permeabilized cells; regions protected by bound transcription factors or nucleosomes remain unmethylated, and methylation patterns read out at base resolution reveal which proteins occupied which positions on each individual molecule. Subsequent work showed that molecular co-occupancy can identify transcription factor binding cooperativity in vivo, and that genome-wide footprinting quantitatively captures high RNA polymerase II turnover at paused promoters. More recent developments have pushed the concept further. Cytosine deaminases discovered in recent years enable methylome mapping with a single enzyme, and deaminase-based footprinting has been combined with CRISPR scanning to create high-throughput functional analysis of cis-regulatory elements, linking perturbation directly to molecule-level chromatin readouts.

Long-range fiber-based approaches extend these principles to chromosome-scale architecture. Chromatin fiber sequencing, or Fiber-seq, described in 2020, maps chromatin accessibility and regulatory architecture on individual DNA molecules at high resolution across multi-kilobase distances. Its single-cell descendant, DAF-seq, uses cytidine deaminases to profile diploid chromatin architecture in individual cells with haplotype phasing, reconstructing chromosome-specific regulatory states. Related technologies use antibody-directed methylation to map protein-DNA interactions on long molecules: DiMeLo-seq directs a methyltransferase to specific proteins of interest, producing long-read maps of where a given factor or histone modification sits along each DNA molecule. Combined approaches now integrate histone modifications, protein-DNA interactions and DNA methylation on multi-kilobase molecules, and BrdU labeling has been coupled with footprinting to profile the accessibility of newly replicated chromatin fibers, opening a window onto how epigenetic information is maintained through DNA replication.

The applications documented in the guide span the breadth of modern genome biology. SMG has been used to examine epigenetic memory in pluripotent and somatic cells, to detect epigenetic heterogeneity in neural stem cells and glioblastoma, and to characterize the variable nucleosome content of active promoters. It has illuminated the organization of centromeres, revealing conserved dichromatin organization, the influence of DNA methylation on centromere positioning, and the heterogeneous packaging of mitochondrial DNA. It has resolved the chromatin impact of mosaic genetic variants through targeted fiber sequencing, mapped allele-specific and tissue-specific regulatory elements over long distances, and quantified how nucleosome density shapes the regulatory output of mammalian chromatin remodelers. Quantitative frameworks have even connected in vitro transcription factor binding affinities to single-molecule chromatin states in living cells, and parallel testing of synthetic DNA sequences has identified how cumulative transcription factor binding and p300-mediated histone acetylation drive enhancer activation frequency.

For laboratories considering adoption, the guide emphasizes careful experimental design. The authors recommend that researchers begin by defining the biological question at the level of individual molecules: whether the goal is measuring co-occurring transcription factor occupancy, resolving haplotypes, mapping a specific histone mark over long ranges, or foot-printing RNA polymerase dynamics. That choice dictates the assay, the sequencing chemistry, and the depth required. Sequencing coverage, defined as the number of usable reads spanning a locus after de-duplication and quality filtering, becomes a critical budget item because molecule-level conclusions demand sufficient observations per molecule rather than per position. Targeted enrichment strategies can raise coverage at predefined loci while reducing cost, and the authors stress that pilot experiments calibrating enzyme accessibility, crosslinking conditions and read length are essential before committing to genome-scale runs.

Quality control receives equally detailed treatment. The guide outlines recommended controls including spike-in standards, assessment of antibody specificity for directed methods, and validation of footprint calls against known factor positions. Computational analysis is presented as an integral component rather than an afterthought: modification calling, the computational identification of introduced or endogenous DNA base modifications from sequencing data, and footprint calling, the unbiased detection of protected regions on individual molecules, each require carefully chosen algorithms and calibrated error models. The authors highlight open-source tools for methylation calling, alignment and multimodal integration of long-read epigenetic assays, and they note emerging deep-learning models that parse base-resolution chromatin accessibility data to extract transcription factor footprints and predict regulatory variant effects. Reproducibility practices, including versioned workflows and deposition of raw and processed data in public archives, are framed as non-negotiable pillars of credible SMG research.

The outlook sketched by the authors suggests that single-molecule genomics is poised to become a standard lens on genome function. As complete telomere-to-telomere human genome assemblies replace older references, SMG readouts anchored to those references can resolve centromeric chromatin, telomeric structure and satellite-derived transcription factor binding platforms that were previously invisible. Integration with single-cell methods is already delivering regulatory maps within individual cells, and synchronized profiling of genome, methylome, epigenome and transcriptome from the same sample promises a truly unified view of how sequence, chromatin state and gene expression interlock. For the field’s newest practitioners, the message of the guide is both practical and ambitious: the genome is not a single static script but a vast ensemble of molecules in different regulatory states at any moment, and single-molecule genomics provides the tools to read that ensemble directly, one DNA molecule at a time.

Subject of Research: Practical guidance for using single-molecule genomics to profile regulatory events on individual DNA molecules genome-wide

Article Title: A practical guide to studying genome function using single-molecule genomics

Article References: Krebs, A. R., Altemose, N., Churchman, L. S., Greenleaf, W. J., Ramani, V., Stadler, M. B., & Stergachis, A. B. (2026). A practical guide to studying genome function using single-molecule genomics. Nature Reviews Molecular Cell Biology. https://doi.org/10.1038/s41580-026-01017-4

Image Credits: AI Generated

DOI: 10.1038/s41580-026-01017-4

Keywords: single-molecule genomics, chromatin, epigenetics, transcription factors, long-read sequencing, DNA methylation, nucleosomes, footprinting, 3D genome organization, enhancers, gene regulation, nanopore sequencing