Genome-wide maps of chromatin interactions have transformed our understanding of how the genetic blueprint is organized inside the cell nucleus, but a fundamental question has remained surprisingly difficult to answer: how often, in any given cell, is a particular loop actually formed? A new study published in Nature Structural & Molecular Biology now provides the first genome-wide answer to that question, and the result is striking. Chromatin loops, long visualized as stable architectural features of the genome, turn out to be fleeting events that exist only a small fraction of the time.

The research, led by a team working in mouse embryonic stem cells, tackles a blind spot that has been built into the most widely used three-dimensional genomics techniques since their inception. Methods such as Hi-C and its higher-resolution successor Micro-C work by cross-linking DNA, cutting it, and ligating fragments that sit close together in three-dimensional space. Sequencing the resulting molecules yields a contact map: a matrix showing which regions of the genome pair with which, and how frequently relative to one another. These maps have revealed loops anchored by the architectural protein CTCF, enhancer-promoter contacts, and the large-scale compartmentalization of chromosomes into active and inactive domains. What they have never delivered, however, is an absolute number. A contact map can say that two loci interact more often than two others, but it cannot say whether the “stronger” pair is together ten percent of the time or one percent of the time. The reason lies in how the assay works. Cross-linking captures proximity imperfectly, digestion and ligation introduce their own biases, and the normalization procedures used to correct for these effects deliberately rescale the data, discarding absolute information in the process. The result is a genome described in relative terms only.

The new study overcomes this by borrowing a measurement from an entirely different experimental tradition: live-cell imaging. In recent years, fluorescence-based approaches have allowed researchers to watch two specific genomic loci in real time, measuring directly the fraction of time they spend in physical contact. These imaging experiments are precise but painfully slow in genomic terms; a single pair of loci can be tracked per experiment, making it impossible to survey tens of thousands of loops this way. The key insight of the new work is that this limitation can be circumvented. Rather than imaging every loop individually, the team used live-imaging data for a subset of loci to calibrate the Micro-C contact maps, effectively anchoring the relative probabilities from sequencing to true absolute probabilities measured in living cells. Once calibrated in mouse embryonic stem cells, the approach could be extended across the entire dataset, converting thousands of relative contact frequencies into genuine probabilities.

The scale of the resulting catalog is considerable. The authors quantified absolute looping probabilities for 65,929 chromatin loops identified by Micro-C in mouse embryonic stem cells. Each loop now carries a number with a straightforward physical interpretation: the estimated probability that, at any given moment in a typical cell of that population, the two anchors of the loop are in contact. Averaged across all quantified loops, that number is just 1.2 percent. In other words, at a randomly chosen instant, fewer than one in eighty of the annotated loops is actually formed. Even the strongest loop in the entire dataset reaches only about 25 percent, meaning that even the most persistent architectural contact in the genome is absent three-quarters of the time.

The finding reframes how chromatin loops should be conceptualized. In textbook diagrams and in the popular imagination, loops appear as stable structures, drawn as definite arcs connecting enhancers to promoters or insulating one domain from another. The new data suggest that such diagrams are better understood as population averages or tendencies rather than as structures that exist in individual cells most of the time. A loop detected robustly in a contact map, and even assigned high confidence by computational loop-calling algorithms, may nonetheless be a rare event at the level of a single cell and a single moment. This probabilistic view generalizes to the whole genome what a handful of live-imaging studies had previously suggested for individual loci: that genome architecture is dynamic, heterogeneous, and dominated by transient encounters rather than durable connections.

The dataset also reveals a hierarchy among different classes of loops. Loops anchored by convergent CTCF binding sites, the canonical motif of the loop-extrusion model, turn out to be substantially more stable than the rest: their average looping probability is 2.2 percent. Loops classified as cis-regulatory, meaning contacts between regulatory elements such as enhancers and promoters, are weaker still, with average probabilities below one percent. The distinction is meaningful for models of gene regulation. Enhancer-promoter communication, which is central to the activation of developmental genes and to the misexpression that drives many cancers, is often assumed to require physical proximity. If such contacts occur less than one percent of the time, then either gene regulation tolerates extremely intermittent enhancer contact, or the functional interaction happens within a subset of cells, or the relevant contact geometry differs from what loop-detection algorithms annotate. Each possibility carries different implications for how transcriptional control should be modeled.

Methodologically, the work addresses a long-standing tension in the field between resolution and interpretability. Micro-C, which uses micrococcal nuclease rather than restriction enzymes to fragment chromatin, achieves near-nucleosome resolution and detects far more loops than conventional Hi-C. But higher resolution has, until now, only sharpened the relative picture. By supplying an absolute scale, the calibration approach converts Micro-C from a comparative tool into a quantitative one, allowing statements of the form “this loop exists with probability X” rather than “this loop is stronger than that loop.” The authors also show that the framework is not confined to the mouse embryonic stem cells in which it was calibrated. Under certain assumptions, the approach can be extended to human cells for which Micro-C data are already available, opening the possibility of absolute loop catalogs across cell types and organisms using existing public datasets.

The implications extend to how looping is detected and validated in future studies. If true looping probabilities are uniformly low, then the signal-to-noise considerations that underlie loop-calling algorithms deserve renewed scrutiny: a contact that appears at a frequency many times above the genomic background may still correspond to an absolute probability of one or two percent. Conversely, the low absolute values provide a benchmark that computational models of chromatin folding, including polymer simulations and machine-learning predictors of structure, can now be tested against. A model that reproduces the qualitative pattern of a contact map may nonetheless fail badly at reproducing the actual fraction of time loops are formed, and absolute data make such failures visible.

There are also consequences for interpreting perturbation experiments. Deleting a CTCF site or degrading cohesin, the motor complex credited with extruding DNA loops, typically changes contact frequencies in maps. With an absolute scale in hand, such changes can now be read as changes in the fraction of time a loop exists, which is closer to the quantity that matters mechanistically. A perturbation that halves a relative contact frequency may, in absolute terms, move a loop from two percent probability to one percent, or from twenty percent to ten percent, and the biological meaning of those two scenarios is very different.

For a field that has spent two decades mapping the genome in three dimensions, the study marks a shift from topology to kinetics and probability. The genome, seen through this calibrated lens, is not a wire-frame of stable loops but a fluctuating ensemble in which defined contacts flicker in and out of existence, dominated by absence rather than presence. The average loop spends nearly 99 percent of its time unformed. Understanding how transcription, replication, and genome maintenance proceed in the brief windows when specific contacts do occur is likely to become a central question for the next phase of research into genome organization.

Subject of Research: Genome-wide absolute quantification of chromatin looping probabilities by calibrating Micro-C contact maps with live-cell imaging data in mouse embryonic stem cells

Subject of Research: Biology

Article Title: Genome-wide absolute quantification of chromatin looping

Article References: Jusuf, J. M., Yang, J. H., Toppen, J., Grosse-Holz, S., Gabriele, M., Mach, P., Flyamer, I. M., Zechner, C., Giorgetti, L., Mirny, L. A., & Hansen, A. S. (2026). Genome-wide absolute quantification of chromatin looping. Nature Structural & Molecular Biology, 33(7), 1105-1114. https://doi.org/10.1038/s41594-026-01819-2

Image Credits: AI Generated

DOI: 10.1038/s41594-026-01819-2

Keywords: chromatin looping, Micro-C, Hi-C, absolute looping probability, CTCF, cis-regulatory loops, mouse embryonic stem cells, live-cell imaging, 3D genomics, genome organization, loop extrusion, gene regulation

Cite Scienmag News
APA MLA Chicago

Juliet Wilcox. (September 5, 2026). New method counts chromatin loops across the entire genome. Scienmag. https://scienmag.com/new-method-counts-chromatin-loops-across-the-entire-genome/

Copy citation Download RIS

Tags: advances in 3D genomics methodschromatin loop dynamicschromatin looping frequencychromosome compartmentalizationchromosome contact frequency analysisCTCF protein role in chromatin architectureCTCF-bound chromatin loopsdynamic nature of chromatin structural featuresenhancer-promoter contact dynamicsenhancer-promoter contact frequencyfleeting chromatin loopsgenome folding and compartmentalizationgenome folding and loopinggenome-wide chromatin interaction mappingHi-C and Micro-C sequencing techniquesHi-C and Micro-C techniqueslive-cell chromatin architecturenuclear architecture in embryonic stem cellsquantitative analysis of chromatin interactionssingle-cell chromatin interaction analysisthree-dimensional genome organizationtransient chromatin loops