Every cell in the human body harbors over six feet of DNA compressed into the microscopic confines of its nucleus. This compaction is achieved through the wrapping of DNA strands around nucleosome cores, which consist of histone proteins assembled into eight-part complexes. Traditionally, it was assumed that these nucleosomes rendered DNA segments fully occluded or completely exposed, dictating gene activity in a simple on/off scheme. The current study challenges this paradigm by demonstrating that most nucleosomes exhibit partial accessibility, with DNA regions only loosely wrapped and therefore partially exposed to cellular machinery, revealing an unexpected layer of genomic control.

The research team utilized a pioneering method named IDLI—Iteratively Defined Lengths of Inaccessibility—building on a previously developed sequencing technology known as SAMOSA. While SAMOSA pioneered the precise mapping of nucleosome locations along individual DNA molecules, IDLI takes a more intricate approach. Through advanced AI modeling trained to interpret subtle structural variations within sequence data, IDLI probes the internal architecture of each nucleosome, revealing distinct conformational states rather than mere positioning. This technique analyzes both the longitudinal axis of DNA and the radial dimensions inside each nucleosome, enabling unprecedented resolution of chromatin structure.

Insights gleaned from applying IDLI to mouse embryonic stem cells uncovered a startling revelation: more than 85% of nucleosomes were found to adopt distorted shapes, characterized by missing or weakened histone components. These structural perturbations result in partial unwrapping of DNA from the nucleosome surface, effectively increasing the genome’s accessibility to transcription factors and other regulatory proteins. Far from random molecular noise, these distortions appear to be meticulously programmed by cells, operating as fine-tuned switches that modulate gene expression levels rather than flipping a simple on/off switch.

Further analyses identified fourteen distinct nucleosome conformations, each correlating with varying intensities of gene activity. Remarkably, these conformational states were conserved across diverse biological contexts, including human stem cells differentiating into liver-like cells and adult mouse liver tissue. This consistency highlights an evolutionarily conserved “chromatin grammar,” where nucleosome shape and composition encode regulatory information, functioning much like the syntax and punctuation in language to modulate genetic expression patterns.

The study also interrogated the interplay between transcription factors—specialized proteins that regulate gene expression—and nucleosome architecture. The researchers demonstrated that transcription factors exert direct structural influence over nucleosomes, guiding them into specific states of distortion that either facilitate or hinder DNA accessibility. Experimental removal of key transcription factors resulted in predictable shifts in nucleosome conformations, underscoring their regulatory role as molecular architects that dynamically sculpt chromatin landscapes to orchestrate complex gene expression programs.

These findings fundamentally shift how the scientific community conceives chromatin biology. Instead of viewing nucleosomes as static gatekeepers enforcing a binary code, this research positions them as dynamic, programmable entities with a rich repertoire of structural states. Such versatility enables cells to finely tune gene activity through a gradient of accessible chromatin configurations, thereby greatly expanding the regulatory toolkit available for controlling cellular identity, function, and response to environmental cues.

The implications for human health and disease are profound. Many complex diseases, including cancer and neurodegenerative disorders, lack clear genetic mutations that explain their onset. This may be because pathogenic states stem from subtle, coordinated shifts in gene regulation across many loci, rather than overt ‘on’ or ‘off’ gene statuses. The newly identified nucleosome conformations provide a highly sensitive readout for such gradations in gene accessibility, potentially offering novel biomarkers or therapeutic targets for diseases rooted in dysregulated chromatin states.

The research team envisions that the IDLI method could also significantly impact aging studies. Chromatin architecture undergoes known alterations during cellular aging, some of which may be reversible. By mapping nucleosome states across different tissues and age groups, scientists could gain vital insights into the molecular underpinnings of aging and identify interventions to restore or maintain youthful chromatin configurations, potentially delaying or mitigating age-associated pathologies.

Ultimately, this work opens new horizons in epigenetics and genome biology, revealing a complex language of chromatin organization that transcends simplistic models. “We’re reading the language of nucleosomes,” says Hani Goodarzi, PhD, co-leader of the study, “but now we want to learn how to speak and modify it.” This vision of rationally manipulating chromatin structure heralds exciting prospects for future therapeutics aimed at precisely controlling gene expression to treat a broad spectrum of diseases.

The study “Pervasive and programmed nucleosome distortion on single chromatin fibers” was published on April 29, 2026, in Nature. The interdisciplinary collaboration brought together experts in molecular biology, computational modeling, and genomics from Gladstone Institutes, Arc Institute, University of California San Francisco, the Netherlands Cancer Institute, and Georgia Institute of Technology. Supported by prominent funding bodies including the National Institutes of Health and the California Institute for Regenerative Medicine, this research exemplifies the power of integrating AI and biotechnology to unlock previously hidden dimensions of genome regulation.

As this new paradigm takes hold, it promises to transform both fundamental biology and translational medicine by providing a refined framework to understand and manipulate how our genetic blueprint is accessed and expressed within living cells. The dynamic and graded chromatin states uncovered by this work challenge decades-old dogmas and offer a road map towards nuanced insights into cellular regulation, disease mechanisms, and the aging process.

Subject of Research: Chromatin Structure, Nucleosome Dynamics, Gene Regulation

Article Title: Pervasive and programmed nucleosome distortion on single chromatin fibres

News Publication Date: April 29, 2026

Web References: https://www.nature.com/articles/s41586-026-10418-6

Image Credits: Gladstone Institutes

Keywords: Nucleosomes, Chromatin, DNA, Genetic structure, Artificial intelligence, Cell nuclei, Gene regulation, Epigenetics, Stem cells, Transcription factors, Genome dynamics, Aging

Phylogenetic trees, or species trees, are fundamental frameworks that allow scientists to decode the evolutionary relationships among species, offering insights not only into the history of life but also into practical fields such as drug discovery, zoonotic disease control, and conservation biology. Constructing these trees typically requires experts to select genetic markers, annotate genomes, and establish orthologous relationships among genes—a process that is both time-consuming and requires considerable domain expertise. ROADIES sidesteps these obstacles by implementing a fully automated pipeline that operates directly on raw genome assemblies, democratizing access to accurate phylogenetic inference for a broad range of researchers.

At the core of ROADIES lies a clever strategy that relies on the random sampling of genomic loci rather than predetermined protein-coding genes or functional markers. This choice defies conventional wisdom, which holds that only carefully selected, conserved genomic regions can produce reliable phylogenetic signals. Yet, UC San Diego’s research, led by Yatish Turakhia and published in the prestigious journal Proceedings of the National Academy of Sciences, demonstrates that random loci sampling not only simplifies the data processing but also maintains, or even enhances, adherence to evolutionary models, resulting in species trees that match those derived from more laborious methods.

This random sampling approach, coupled with novel computational algorithms, allows ROADIES to forgo genome annotation entirely. Genome annotation—the process of identifying and labeling functional elements within a DNA sequence—is a major bottleneck that usually requires extensive manual input and computational power. By bypassing this requirement, ROADIES drastically reduces the time and resources needed to move from raw sequencing data to evolutionary insights, a leap forward that could catalyze a new wave of comparative genomic studies.

Another significant hurdle conquered by ROADIES is the issue of orthology inference. Orthology involves distinguishing between genes in different species that originated from a common ancestral gene, a process complicated by gene duplication events that produce multiple gene copies across genomes. Many extant phylogenetic tools struggle with paralogs—these duplicated genes—leading to inaccuracies if misclassified. ROADIES incorporates sophisticated algorithms developed in the lab of Siavash Mirarab that accept multi-copy genes without relying on explicit orthology assignments. This discordance-aware methodology ensures robust phylogenetic inference even when faced with complex gene family histories.

The implications of removing these two major steps—annotation and orthology inference—are profound. ROADIES can process extensive datasets containing hundreds of genomes, inferring species trees that are concordant with expert-generated, large-scale phylogenies but require only a fraction of the computational investment. The scalability of ROADIES opens doors for its application to the massive genomic datasets expected in upcoming biodiversity projects, such as the Earth BioGenome Project, which aims to sequence nearly every eukaryotic life form on the planet.

The study showcased ROADIES’s impressive performance across a diverse array of taxa, including placental mammals, pomace flies, birds, and budding yeasts. The tool’s versatility highlights its applicability across the tree of life, underscoring its potential as a game-changer in evolutionary research. By facilitating rapid and automated species tree inference, ROADIES not only accelerates phylogenomic studies but also broadens participation in this research area beyond specialized bioinformatics groups.

Looking ahead, the team behind ROADIES plans to enhance the tool’s capabilities further. One exciting avenue is the implementation of algorithms for the placement of new taxa on preexisting species trees, making incremental updates more feasible. Additionally, leveraging GPU computing resources could exponentially increase throughput, enabling the phylogenetic analysis of tens of thousands—or even hundreds of thousands—of genomes, aligning with the scale of current and future genomic sequencing endeavors.

The potential applications of ROADIES extend beyond academic research. By enabling faster identification of functional genomic regions and evolutionary patterns, this technology could expedite the development of new pharmaceuticals, provide early warnings for zoonotic disease outbreaks, and inform targeted conservation strategies for vulnerable species. The tool’s capacity to integrate complex genetic data at scale represents a substantial leap forward in translating genomic information into actionable knowledge.

With genome assembly technologies continuously improving and sequencing becoming more accessible, the bottleneck in extracting meaningful evolutionary insights has shifted towards computational analysis. ROADIES epitomizes the next generation of bioinformatics tools, characterized by automation, accuracy, and scalability. The research community eagerly anticipates the widespread adoption of ROADIES, which promises to accelerate discoveries in evolutionary biology and related fields.

In conclusion, the advent of ROADIES marks a pivotal moment in phylogenetics. By reimagining how species trees can be inferred from raw genomic data, this tool paves the way for a deeper, more comprehensive understanding of the tree of life. The work of Turakhia, Mirarab, and colleagues exemplifies the synergy of engineering and biology, setting a new standard for innovation in the life sciences. As large-scale sequencing initiatives progress, tools like ROADIES will be indispensable in unlocking the secrets held within the genomes of Earth’s astonishing diversity.

Subject of Research: Not applicable

Article Title: Accurate, scalable, and fully automated inference of species trees from raw genome assemblies using ROADIES

News Publication Date: 2-May-2025

Web References:
https://www.pnas.org/doi/10.1073/pnas.2500553122

References:
Turakhia, Y., Mirarab, S., et al. (2025). Accurate, scalable, and fully automated inference of species trees from raw genome assemblies using ROADIES. Proceedings of the National Academy of Sciences. https://doi.org/10.1073/pnas.2500553122

Image Credits: Artwork by Alice Grishchenko

Keywords: Phylogenetics, Genome mapping