Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript. , Article number: (2026) Cite this article We are providing an unedited version of this manuscript to give early access to its findings. Before final publication, the manuscript will undergo further editing. Please note there may be errors present which affect the content, and all legal disclaimers apply. Accurately predicting the activity of a chemical in each bioactivity assay based on its already known properties is extremely useful in drug development. Unfortunately, we discovered that many assays in widely used assay-activity benchmark datasets directly relate to cell health and cytotoxicity. Many other assays intend to capture a more specific phenotype, but their active compounds impact cell count, while inactives do not. In both cases, counting cells achieves unexpectedly high performance in these benchmarks, making them less useful for discerning whether additional properties, such as phenotypic profiles (mRNA or Cell Painting), provide additional useful information on bioactivity. To accomplish this goal, we recommend filtering benchmarks to exclude such assays and including a cell-count baseline. Using a benchmark with 24 protein-target assays, we confirm that models leveraging Cell Painting image-based profiles outperformed the baseline cell count model. We propose several other practical recommendations for benchmarking machine learning models for predicting bioactivity and assessing the added value of mRNA, protein, or image-based profiles. The data used in this study have been deposited in the Zenodo database under accession code https://doi.org/10.5281/zenodo.17168185 [https://doi.org/10.5281/zenodo.14838603]. Supplementary Data 1 provides annotated assays from the Hofmarcher dataset, along with associated metadata, including assay type, target type, organism, and assay category. Supplementary Data 2 provides a comparison of the Protonet CP+ model (at a support set size of 64) with the baseline cell count model, as benchmarked on FSL-CP in Ha et al. Figure 2e-h can be reproduced with source data from Supplementary Data 3. All other Figures can be reproduced from data and notebooks deposited at Zenodo. Supplementary data are provided with this paper. The code used to develop the model, perform the analyses, and generate results in this study is publicly available and has been deposited in Zenodo at https://doi.org/10.5281/zenodo.1793112465 (and GitHub at https://github.com/srijitseal/The_Seal_Files), under MIT license. This includes a notebook with steps to filter the Cell Painting dataset for features highly correlated to cell count: https://github.com/srijitseal/The_Seal_Files/blob/main/02_Remove_Confounders_Cell_Painting_Dataset.ipynb. The specific version of the code associated with this publication is archived in Zenodo and is accessible via https://doi.org/10.5281/zenodo.1483860364. Bender, A. et al. Evaluation guidelines for machine learning tools in the chemical sciences. Seal, S. et al. Machine learning for toxicity prediction using chemical structures: pillars for success in the real world. Zhou, H. & Skolnick, J. Utility of the Morgan fingerprint in structure-based virtual ligand screening. Seal, S. et al. Merging bioactivity predictions from cell morphology and chemical fingerprint models using similarity to training data. Hanser, T., Barber, C., Marchaland, J. F. & Werner, S. Applicability domain: towards a more formal definition. SAR QSAR Environ. Liu, A., Seal, S., Yang, H. & Bender, A. Using chemical and biological data to predict drug toxicity. Serrano, E. et al. Progress and new challenges in image-based profiling. Preprint at ArXiv https://doi.org/10.48550/arXiv.2508.05800 (2025). Seal, S. et al. Cell Painting: a decade of discovery and innovation in cellular imaging. Chandrasekaran, S. N., Ceulemans, H., Boyd, J. D. & Carpenter, A. E. Image-based profiling for drug discovery: due for a machine-learning upgrade? Simm, J. et al. Repurposing high-throughput image assays enables biological activity prediction for drug discovery. Liu, G. et al. Learning molecular representation in a cell. International Conference on Learning Representations https://openreview.net/pdf?id=BbZy8nI1si (2025). Hofmarcher, M., Rumetshofer, E., Clevert, D.-A., Hochreiter, S. & Klambauer, G. Accurate prediction of biological assays with high-throughput microscopy images and convolutional networks. Moshkov, N. et al. Predicting compound activity from phenotypic profiles and chemical structures. Sanchez-Fernandez, A., Rumetshofer, E., Hochreiter, S. & Klambauer, G. CLOOME: contrastive learning unlocks bioimaging databases for queries with chemical structures. Ha, S. V., Leuschner, L. & Czodrowski, P. FSL-CP: a benchmark for small molecule activity few-shot prediction using cell microscopy images. Escher, B. I., Henneberger, L., König, M., Schlichting, R. & Fischer, F. C. Cytotoxicity burst? Differentiating specific from nonspecific effects in Tox21 in vitro reporter gene assays. Dahlin, J. L. & Walters, M. A. The essential roles of chemistry in high-throughput screening triage. Future Med. Ibrahim, M., Klindt, D. & Balestriero, R. Occam's razor for self supervised learning: what is sufficient to learn good representations? Preprint at arXiv https://doi.org/10.48550/arXiv.2406.10743 (2024). Lawrence, E. et al. Understanding biology in the age of artificial intelligence. Preprint at arXiv https://doi.org/10.48550/arXiv.2403.04106 (2024). & Glen, R. C. A discussion of measures of enrichment in virtual screening: comparing the information content of descriptors with increasing levels of sophistication. Walters, P. Comparing classification models - you're probably doing it wrong. How to lie with computational predictive models in drug discovery. Ahlmann-Eltze, C., Huber, W. & Anders, S. Deep-learning-based gene perturbation effect prediction does not yet outperform simple linear baselines. Csendes, G., Sanz, G., Szalay, K. Z. & Szalai, B. Benchmarking foundation cell models for post-perturbation RNA-seq prediction. Duan, B., Chen, S., Cheng, X. & Liu, Q. Multi-slice spatial transcriptome domain analysis with SpaDo. Genome Biol. AID 651635 - viability counterscreen for primary qHTS for inhibitors of ATXN expression - PubChem. AID 624202 - qHTS assay to identify small molecule activators of BRCA1 Expression - PubChem. Saito, T. & Rehmsmeier, M. The precision-recall plot is more informative than the ROC plot when evaluating binary classifiers on imbalanced datasets. PLoS ONE 10, e0118432 (2015). Varalda, M. et al. Psychotropic drugs show anticancer activity by disrupting mitochondrial and lysosomal function. Wang, C.-C., Chen, B.-K., Chen, P.-H. & Chen, L.-C. Hinokitiol induces cell death and inhibits epidermal growth factor-induced cell migration and signaling pathways in human cervical adenocarcinoma. Kim, J.-H., Chae, M., Choi, A.-R., Sik Kim, H. & Yoon, S. SP600125 overcomes antimitotic drug-resistance in cancer cells by increasing apoptosis with independence of P-gp inhibition. Cell Painting, a high-content image-based assay for morphological profiling using multiplexed fluorescent dyes. Luo, Y. et al. Characterization and analysis of biopharmaceutical proteins. in Handbook of Modern Pharmaceutical Analysis 283–359 (Elsevier, 2011). Schmitt, D. M. et al. The use of resazurin as a novel antimicrobial agent against Francisella tularensis. Liu, J. et al. Wnt/β-catenin signalling: function, biological mechanisms, and therapeutic opportunities. Chen, C. et al. Aberrant activation of Wnt/β-catenin signaling drives proliferation of bone sarcoma cells. & Wang, X. Targeting the Wnt/β-catenin signaling pathway in cancer. Ali, J., Camilleri, P., Brown, M. B., Hutt, A. J. & Kirton, S. B. In silico prediction of aqueous solubility using simple QSPR models: the importance of phenol and phenol-like moieties. Ishikawa, M. & Hashimoto, Y. Improvement in aqueous solubility in small molecule drug discovery programs by disruption of molecular planarity and symmetry. Peng, X. et al. Exploring a structural protein-drug interactome for new therapeutics in lung cancer. Cross-Zamirski, J. O. et al. Label-free prediction of cell painting from brightfield images. et al. A dataset of images and morphological profiles of 30,000 small-molecule treatments using the Cell Painting assay. Chandrasekaran, S. N. et al. JUMP cell painting dataset: morphological impact of 136,000 chemical and genetic perturbations. Preprint at bioRxiv https://doi.org/10.1101/2023.03.23.534023 (2023). Comolet, G. et al. A highly-efficient, scalable pipeline for fixed feature extraction from large-scale high-content imaging screens. Ewald, J. D. et al. Cell painting for cytotoxicity and mode-of-action analysis in primary human hepatocytes. Preprint at bioRxiv https://doi.org/10.1101/2025.01.22.634152 (2025). Hanley, J. & McNeil, B. J. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Seal, S. et al. From pixels to phenotypes: integrating image-based profiling with cell health data as BioMorph features improves interpretability. Pahl, A. et al. Morphological subprofile analysis for bioactivity annotation of small molecules. Seal, S. et al. Integrating cell morphology with gene expression and chemical structure to aid mitochondrial toxicity detection. Eve Bio: Pharmome data. Seal, S. et al. Insights into drug cardiotoxicity from biological and chemical data: The first public classifiers for FDA drug-induced cardiotoxicity rank. Seal, S., Yang, H., Vollmers, L. & Bender, A. Comparison of cellular morphological descriptors and molecular fingerprints for the prediction of cytotoxicity- and proliferation-related assays. Lambert, R. et al. Drug-induced cytotoxicity prediction in muscle cells, an application of the Cell Painting assay. PLoS ONE 20, e0320040 (2025). Rouquié, D. et al. The OASIS consortium: integrating multi-omics technologies to transform chemical safety assessment. Statistical functions (scipy.stats) — SciPy v1.11.3 Manual. & Kull, M. Precision-recall-gain curves: PR analysis done right. Neural Inf. Matlock, M. K., Hughes, T. B., Dahlin, J. L. & Swamidass, S. J. Modeling small-molecule reactivity identifies promiscuous bioactive compounds. Varela-Moreira, A. et al. Utilizing in vitro drug release assays to predict in vivo drug retention in micelles. Cerisier, N., Dafniet, B., Badel, A. Linking chemicals, genes and morphological perturbations to diseases. Subramanian, A. et al. A next generation connectivity map: L1000 platform and the first 1,000,000 profiles. Eve Bio Pharmome Data. Rücker, C., Rücker, G. & Meringer, M. y-Randomization and its variants in QSPR/QSAR. Pedregosa, F. et al. Scikit-learn: machine learning in Python. Preprint at arXiv https://doi.org/10.48550/arXiv.1201.0490 2825–2830 (2012). Seal, S. et al. Counting cells can accurately predict small-molecule bioactivity benchmarks. Preprint at Zenodo. Seal, S. et al. Counting cells can accurately predict small-molecule bioactivity benchmarks Srijitseal/the_seal_files: counting cells can accurately predict small-molecule bioactivity benchmarks. Preprint at Zenodo. Data Release #4. https://data.evebio.org/. S. Seal acknowledges Reid Olsen (Recursion), Anna Lobley (Independent), Barak Gilboa (Novo Nordisk), and Hassan A Ali (University of Miami) for their comments on a pre-print and to the independent reviewers for their suggested analyses that greatly enhanced this publication. S. Seal acknowledges funding from the Cambridge Centre for Data-Driven Discovery (C2D3) Accelerate Programme for Scientific Discovery. S. Seal, S. Singh, and A.E.C. acknowledge funding from the National Institutes of Health (NIH MIRA R35 GM122547 to A.E.C. ), the Massachusetts Life Sciences Center Bits to Bytes Capital Call program for funding the data production (to S. Singh), as well as their Data Science Internship program (to S. Singh), and the OASIS Consortium organized by HESI. acknowledges funding from the Swedish Research Council (grants 2020-03731, 2020-01865, 2024-03566, 2024-04576), FORMAS (grant 2022-00940), Swedish Cancer Foundation (22 2412 Pj 03 H), and Horizon Europe grant agreement #101057014 (PARC) and #101057442 (REMEDI4ALL). W. Dee and G. Slabaugh acknowledge the UKRI/BBSRC Collaborative Training Partnership in AI for Drug Discovery, led by Exscientia Plc. in partnership with Queen Mary University of London. The Collaborative Training Partnership was funded by the Biotechnology and Biological Sciences Research Council, grant reference BB/X511791/1. G. Slabaugh also acknowledges EPSRC grant EP/Y009800/1, through Keystone project funding from Responsible AI UK (KP0016) and also acknowledges the support of the National Institute for Health and Care Research Barts Biomedical Research Centre (NIHR203330); a delivery partnership of Barts Health NHS Trust, Queen Mary University of London, St George's University Hospitals NHS Foundation Trust and St George's University of London. Open access funding provided by Uppsala University. These authors contributed equally: Srijit Seal, William Dee. Department of Chemistry, University of Cambridge, Cambridge, UK Srijit Seal & Andreas Bender Broad Institute of MIT and Harvard, Cambridge, MA, USA Srijit Seal, Adit Shah, Esteban Miglietta, Shantanu Singh & Anne E. Carpenter Digital Environment Research Institute (DERI), Queen Mary University of London, London, UK William Dee & Gregory Slabaugh Université Paris Cité, INSERM U1133, CNRS UMR 8251, Paris, France Natacha Cerisier & Olivier Taboureau Health Sciences and Technology, Harvard-MIT, Cambridge, MA, USA Andrew Zhang Axiom Bio, San Francisco, CA, USA Katherine Titterton, Ángel Alexander Cabrera, Daniil Boiko & Alex Beatson Department of Pharmaceutical Biosciences and Science for Life Laboratory, Uppsala University, Uppsala, Sweden Jordi Carreras Puigvert & Ola Spjuth Pixl Bio AB, Uppsala, Sweden Jordi Carreras Puigvert & Ola Spjuth Center for Biotechnology, College of Medicine and Health Sciences, Khalifa University of Science and Technology, Abu Dhabi, UAE STAR-UBB Institute, Babeş-Bolyai University, Cluj-Napoca, Romania Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar S. Seal conceived the study, designed all analyses, performed the modeling, carried out benchmarking and data interpretation, curated datasets, generated figures, and wrote the manuscript. ran several model comparisons and co-wrote the manuscript with S. Seal. A.S. supported code development and preprocessing of Cell Painting data. N.C. contributed to the gene expression analyses. helped with code and data handling. E.M. assisted in the analysis of selected Cell Painting images. K.T., Á.A.C., D.B., and A. Beatson contributed on behalf of Axiom Bio to describe the experiments involving cell count versus concentration data. provided comments on the manuscript and guidance on model evaluation. assisted in writing the manuscript. S. Singh provided expertise on Cell Painting data processing and interpretation. co-supervised the work together with S. Seal, advising throughout study design, model interpretation, and manuscript writing and revision. All authors reviewed, contributed to, and approved the final manuscript. Correspondence to Srijit Seal, Ola Spjuth, Andreas Bender or Anne E. Carpenter. serve as scientific advisors for companies that use image-based profiling and Cell Painting (A.E.C. : Recursion, SyzOnc, Quiver Bioscience; S. Singh: Waypoint Bio, Dewpoint Therapeutics, DeepCell) and receive honoraria for occasional talks at pharmaceutical and biotechnology companies. declare ownership in Phenaros Pharmaceuticals. serves as a scientific advisor to BioAIHealth and has a collaborative project with AstraZeneca involving image-based profiling and Cell Painting. The remaining authors declare no competing interests. Nature Communications thanks Bin Duan, Diego Galeano, Song He and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available. Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Reprints and permissions Seal, S., Dee, W., Shah, A. et al. Counting cells can accurately predict small-molecule bioactivity benchmarks. Received: 09 February 2025 Anyone you share the following link with will be able to read this content: Sorry, a shareable link is not currently available for this article. Provided by the Springer Nature SharedIt content-sharing initiative © 2026 Springer Nature Limited Sign up for the Nature Briefing: Translational Research newsletter — top stories in biotechnology, drug discovery and pharma.

OpenClaw is an open-source artificial-intelligence agent designed to assist users with everyday tasks, such as sending e-mails and managing their calendars.Credit: Thomas Fuller/SOPA Images/LightRocket via Getty OpenClaw is an AI agent capable of performing tasks on personal devices, such as scheduling calendar events, reading e-mails, sending messages through apps and using the Internet to make purchases. Most of the popular AI tools such as OpenAI's ChatGPT chatbot works by interacting directly with user prompts, whereas agentic AI models such as OpenClaw can carry out actions autonomously in response to instructions. Improvements in the capabilities of large language models have made it possible to create more versatile AI tools, researchers say. “OpenClaw promises something especially appealing: a capable assistant embedded in the everyday apps people already rely on,” says Barbara Barbosa Neves, a sociologist who focuses on technology at the University of Sydney in Australia. “It's a kind of chaotic, dynamic system that we're not very good at modelling yet,” he adds. Although agents can act autonomously on the platform, Cohney says that many posts are shaped in some way by humans. Users can choose the underlying large language model that will run their agent and give it a personality. For example, they could ask it to behave like a “friendly helper”, he says. But agents do not possess intentions or goals and draw their abilities from large swathes of human communication. “It is still worth studying because it tells us something important about how people imagine AI, what they want agents to do and how human intentions are translated, or distorted, through technical systems,” she adds. Joel Pearson, a neuroscientist at the University of New South Wales in Sydney, Australia, says that when people see AI agents chatting between themselves, they are likely to anthropomorphize the AI models' behaviour — that is, see personality and intention where none exists. The risk of that, he says, is that it makes people more likely to form bonds with AI models, becoming dependent on their attention or divulging private information as if the AI agent were a trusted friend or family member. “As the AI models get bigger and more complicated, we'll probably start to see companies leaning into achieving that sort of autonomy.” ArXiv says submissions must be in English: are AI translators up for the job? ‘It means I can sleep at night': how sensors are helping to solve scientists' problems ‘It means I can sleep at night': how sensors are helping to solve scientists' problems Aligning with world-class university standards, CityUHK (Dongguan) has now launched a global recruitment campaign for faculty members The University invites individuals from diverse backgrounds to apply for faculty positions in this field The University invites individuals from diverse backgrounds to apply for faculty positions in this field An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday. Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

You are using a browser version with limited support for CSS. “I will never forget that Saturday evening when I first saw Russian tanks on the streets of my city,” says Viktoriya Voropayeva, a systems engineer and vice-rector at the Donetsk National Technology University (DonNTU). In 2014, after Russian-backed forces took over Donetsk, the unofficial capital of Ukraine's Donbas region, Voropayeva and many of her colleagues chose to leave, setting up their university in exile. The university's first new home was in Pokrovsk, a small city about 60 kilometres away from Donetsk, still in the Donbas region, where it already had a sister institution. In April 2022, two months after Russia's full-scale invasion of Ukraine, the university moved again. The city council offered several buildings to turn into offices and classrooms. How the invasion of Ukraine is affecting Russian expat researchersThe European nation is not the only place where war or political unrest has forced universities and their staff members into exile. Others around the world, including in Sudan and Myanmar, have also had to relocate. How the invasion of Ukraine is affecting Russian expat researchers More than 50% of DonNTU first-year students are from the Lviv region. Illya Khadzhynov, an economist and vice-rector for scientific work, was last in Donetsk in July 2014, when his institution, Donetsk National University (DonNU) was taken over by the pro-Russian separatist government. It authorized the university's roughly 700-km move west to a former jewellery factory in Vinnytsia, in the west-central region of Ukraine, a building with no lecture theatres or laboratories. But with support from international donors, including US$350,000 from the International Renaissance Foundation, a Ukrainian charity founded by Hungarian-born US philanthropist George Soros, DonNU refurbished the factory building and re-established a campus there, including labs for research and teaching. Khadzhynov estimates that about half of the university's 12,000 students moved to Vinnytsia. Serhii Radio, a chemistry researcher at DonNU, says that back in 2014, not everyone felt able to leave everything behind. Those who did took only “the most necessary things that you can carry in two hands”, he says. He was unable to take any lab equipment that might stand out at the checkpoints they had to pass, controlled by armed pro-Russian separatist groups. Their current 650-strong staff includes 180 who have been displaced from other cities. Now I work on ways to help the country's soil heal The war itself has brought the added problem of electricity cuts, which make it difficult for Radio to perform experiments for his studies using single-crystal X-ray diffraction. “An experiment needs more than eight hours, and this is a very long time for electricity,” he says. At DonNTU, Voropayeva says: “Laboratory equipment has been in storage since 2022 or was destroyed during shelling.” The university is gradually restoring its labs, and a pooled centre for collective use of scientific equipment enables Ukrainian academic institutions to share resources, sometimes on a fee-for-service basis. Salaries have become catastrophically low — Voropayeva says that a full-time associate professor receives the equivalent of €300–350 (US$350–410) a month, which is comparable to pre-war wages but now buys considerably less because of rising prices. Another challenge for both institutions has been the move from the heavily industrial Donbas, where many institutions excelled in applied science, to a region with a different industrial history. The war in Ukraine has also forced some Russian academics into exile. Smolny was the first liberal-arts college in Russia, awarding degrees from 2003 and following a broad multidisciplinary curriculum. But in June 2021, Bard College was declared an “undesirable” organization by the Russian prosecutor's general office and all ties were cut. Last semester, they ran 23 courses and now offer a two-year associate-degree programme accredited by Bard College. Fedchin says that about 50% of its students taking non-degree courses are in Russia. He and his family relocated from Russia to France in March 2022, where, as well as teaching maths and statistics for Smolny Beyond Borders, he has a temporary teaching position at the University of Lorraine in Nancy. It is currently under way in Kenya, Jordan, Bangladesh and parts of East Africa. “Some of the universities are still not working today,” says Ibrahim. Others, including the University of Khartoum, have adopted online teaching and have set up centres in regional universities to host exams and crucial practical training in subjects such as medicine. Engineer Gihad Ibrahim had to leave his home in the Sudanese city of Khartoum North when civil war broke out.Credit: Mashreq University This includes the private Mashreq University, which specializes in applied science, engineering and medicine. In Cairo, Mashreq University rented an entire unused college building. Student numbers have also decreased by 45% owing to drop-outs and temporary suspensions of study. “Many students were not able to pay their fees because we had 300–400% inflation after the war and lots of people have lost their jobs,” says Ibrahim. The college has tried to provide scholarships where it can. After two-and-a-half years working this way, Ibrahim says with pride and determination that they are sustaining teaching “regardless of the difficult situation”. Most of the universities in conflict zones have also been destroyed and looted. Subscribe to Nature Briefing: Careers, an unmissable free weekly round-up of help and advice for working scientists. ‘Armed groups entered the lab': Sudan's researchers flee violent military conflict How the invasion of Ukraine is affecting Russian expat researchers War shattered Ukrainian science — its rebirth is now taking shape AI could transform research assessment — and some academics are worried The number of China's elite scientists who have been trained abroad is falling Epstein files reveal deeper ties to scientists than previously known AI could transform research assessment — and some academics are worried Artificial intelligence tools expand scientists' impact but contract science's focus When two years of academic work vanished with a single click Aligning with world-class university standards, CityUHK (Dongguan) has now launched a global recruitment campaign for faculty members The University invites individuals from diverse backgrounds to apply for faculty positions in this field The University invites individuals from diverse backgrounds to apply for faculty positions in this field ‘Armed groups entered the lab': Sudan's researchers flee violent military conflict How the invasion of Ukraine is affecting Russian expat researchers War shattered Ukrainian science — its rebirth is now taking shape An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday. Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). You have full access to this article via your institution. Stigma is a massive barrier to seeking help for alcohol dependence in academia. Two researchers share their experiences with Adam Levy.Your browser does not support the audio element.Download MP3See transcript Wendy Dossett tells Adam Levy why the stigma of having an alcohol dependence in academia can be a huge barrier to seeking help. “We're not supposed to be struggling with cognitive issues, mental health problems, damaging ourselves in the way that somebody with an alcohol addiction is doing.” Dossett, now an emeritus professor of religious studies at the University of Chester, UK, says that as an early career researcher she saw alcohol as the fuel to her academic life, driving her creativity and making the social elements of academic life easier to navigate. When, in her 30s, a colleague suggested she might need help, Dorsett says she felt a “mixture of horror and absolute gratitude that somebody had the courage and care for me”. She went on to research the spiritual elements of recovery from addiction, which she says is less talked about in academia than, say, depression and anxiety. Victoria Burns, a social work scholar at the University of Calgary in Alberta, Canada, founded Recovery on Campus Alberta after telling her dean that she had an alcohol dependence. He told her she was the first academic to disclose in his 26-year career, prompting her to research other deans' experiences of faculty members disclosing addiction and recovery. This is the fifth episode of Off Limits, a podcast series exploring topics that are often perceived as taboo in the workplace, including religion, bereavement, activism and sizeism. Subscribe to the Nature Careers Working Scientist podcast on Apple Podcasts, Spotify, YouTube or your favourite podcast app. Subscribe to Nature Briefing: Careers, an unmissable free weekly round-up of help and advice for working scientists. Hello, I'm Adam Levy, and this is Off Limits: Academia's Taboos, a podcast from Nature Careers. Alcohol and academia have a complex relationship. On the one hand, in many countries and contexts, consuming alcohol is incredibly normalized. In fact, as we heard in last week's episode, people who don't drink for religious reasons can struggle to find social spaces in academia that don't centre around drinking. And so it's perhaps unsurprising that many academics can struggle with alcohol dependency. But while alcohol consumption may be the norm, coming to terms with and openly discussing harmful relationships with the substance remain incredibly taboo. Over her career, she's investigated a number of questions related to faith. In the last decade or so of my career, I turned my religious studies methodologies onto the phenomenon of recovery from addiction. This is a topic close to Wendy's heart. And so we spoke about her journey with alcohol. But looking back on my student days and my early career, alcohol was a terrible problem for me. Much like lots of people who suffer with substance issues, I had low self esteem. I felt that I didn't fit in, and substance use enabled me to function for many years. I experienced sexual abuse at the age of 21 and I reacted to that experience by drinking more, and my problems with alcohol escalated. The way you describe it is, of course, with hindsight. Did you have this kind of understanding at the time that alcohol was something you were using to cope with these other things in academia and in your personal life? So alcohol felt like it was a fuel to my academic life. A lot of it is about networking, and, you know, making friends with the right people, and, you know, being able to tell people your ideas in a confident manner. I probably had some kind of inkling that it was also causing me trouble, but that kind of remained at an unconscious level, really, all through my 20s. Were there any particular moments which, at the time, or perhaps in hindsight, really highlighted, ”Okay, this is becoming a problem.” And actually getting in the way of the things you thought it would be helping for? There were some spectacular ones and some just the daily grind of waking up every morning or coming round, I should say, every morning and trying to get myself ready to go to work. Those were the kind of day in, day out realizations that there was there was something wrong. For example, I chaired an international symposium in blackout. You know, it doesn't mean I was unconscious. At the end of it, a very good friend of mine, a colleague, said to me that they thought I needed to get some help, and that they were worried about my drinking. Were you surprised by this, I suppose almost intervention from a colleague? It was an intervention, and this was well over 20 years ago now, and I I look back on that moment with a mixture of horror and absolute gratitude that somebody had the courage and care for me, really, to actually say that they thought I had a problem. You know, who would want to admit that they were out of control? You know, especially in academia, perhaps there is a great stigma around mental health problems. We're supposed to be making contributions moving the frontiers of knowledge forward. We're not supposed to be struggling with cognitive issues, mental health problems. We're not supposed to be damaging ourselves in the way that somebody with an alcohol addiction is doing. So, yes, the stigma is a massive barrier to help-seeking. I think we've made such great strides with, you know, especially depression and anxiety. Is that something you kept separate from your colleagues? I didn't mention it to any colleagues for many years. Because academia is such a drinking culture, you've got a few options if you don't drink. But over time, I became more comfortable, more secure in my own recovery status, and I think that gave me a little bit of confidence to raise it with peers and to say, ”Yeah, I don't drink. But it was a whole other thing to bring that up with senior members of the university. And I was anxious about many different things. I was anxious about causing myself issues with promotion or respect from senior colleagues, or, you know, trust from senior colleagues. And can I ask, I mean, you mentioned that at academic events with alcohol, at first, you found it very difficult to be open about how you were navigating alcohol. Were these events also challenging, just in terms of yourself and in terms of being in the context where drinking alcohol is not only the norm, but maybe also expected? I don't know many people in recovery who are fully comfortable in an environment where there's lots of alcohol, or lots of drunkenness. I find it quite difficult sometimes being around drunk colleagues. But certainly in the early days of my recovery I found it incredibly difficult. Not necessarily because I might be tempted to drink, but that is always a possibility, but because I just found it so stressful to be in that environment. Academia has historically at least functioned around drinking. Also other drugs as well, especially the kinds of drugs that enhance cognitive performance in the short term. In order to be competitive in an academic environment, there is a sense that people feel they need to have some kind of enhanced performance in order to deliver. I felt as though this research really was me finding my academic voice, actually, and finding the contribution that I could make to the recovery world, and to the academic study of recovery. You know, recovery from substance misuse is different to recovery from other kinds of mental health issues. And of course, the recovery scene is not homogenous. There are many, different ways of recovering from alcohol or other drug use. Those all derive from a spiritual notion that the person suffering with the addiction, their condition is a state of powerlessness, and their willpower is not sufficient to overcome the problem. So therefore they must access a higher power. So I was, I was trying to show how members of 12 step programs are spiritual innovators in the way they they think about higher power. I mean, you're speaking to me now. You also spoke to Nature for a feature published back in 2025. It's been a long and difficult journey to become more open. What should academics listening do to better support their their colleagues who are going through challenges with dependency? I'd love colleagues to become much more informed about addiction, and in particular, much more informed about recovery, and to know that recovery is possible. There can often be a tendency to judge people who struggle with addiction. I'd love it if more academics knew about the recovery-friendly university movement, and that means they're actively seeking ways to support both staff and students who have substance issues. They're saying that we are proud to have people in recovery within our university community. I think communities are the places in which recovery can happen in a in a community where there are people in recovery and they're known and it's seen that they're welcomed and they're celebrated, then that creates its own contagion. And it means that other people can find their way to recovery as well. So universities are actually really places where we can seed recovery for the wider community. We'll hear from her again in a later episode, as we discuss the struggles and stigma surrounding fertility. Next up, though, I spoke with another researcher, Victoria Burns, associate professor of social work at the University of Calgary in Canada. She is the founder and director of Recovery on Campus Alberta. We'll get to her work with this organization in just a moment. But first we discussed how Victoria began to reflect on her relationship with alcohol. And then I didn't drink my first year of university. But then my second year, I did an exchange program to France. And that's where I really started to drink in earnest. I always held down jobs and was an A student. But I also started to drink more, and it was actually at 19 when a boyfriend gave me an ultimatum to stop drinking because I would black out and end up in risky situations. No, so I did a lot more research. I said bye to the boyfriend, and thought that he was just being too controlling. And during that period, there were many, I guess, increasingly dangerous situations that I was in related to my drinking. So I like to say every time I drank, something bad didn't happen, but anytime something bad happened, alcohol was involved. You just kind of mentioned it in passing there. But did it shake you up quite a bit to end up having to seek, you know, medical support in this situation? I did seek medical support earlier on at my university. And after some particularly bad binges I went to the student wellness and no one really gave me supports related to my drinking. They just told me I needed to drink less and were quite judgmental. And when I ended up in the hospital, that last time, I ended up seeing three therapists, actually. It was the abstinence-based one who had recommended I try AA. And then I found a woman's meeting. And I had tried, actually, 14 months before my final bottom to go to rehab, but they wouldn't accept me because I was quote, unquote, too organized as a full time student. So on paper, things looked okay, but I was really dying inside. And now, after you did successfully go to AA and transition away from alcohol, how did you approach discussing, disclosing this with academic colleagues and with supervisors, for example? So I didn't disclose to anybody in my academic circles, except for a couple of senior academics. I did tell not to tell anyone noise in recovery, because it could negatively affect my job prospects. I didn't want to do anything that would negatively affect my career. So I really led a double life in recovery for the first five years. And I am a social work scholar. And this became increasingly unmanageable as well. What do we know about how researchers who struggle with substance dependence, disclose or don't disclose? I was researching harm reduction and housing first at the time. So I ended up disclosing to my Dean. And he gave me a green light to conduct research in this area, because he said, out of his 26-year career, he never had anyone disclosed to him about being in recovery. However, there were many incidents where faculty members were showing up to the classroom intoxicated or concerns about people's drinking or drug use, but Deans and service providers not really knowing what that process was to help them. And several of the participants actually said, ”Well, as long as they're bringing in grants, no one really cares.” And it's also problematic, because like me, I have never met anyone in recovery in academia, and I was a student for 14 years. So one of the recommendations of this work was to shift more into this vulnerable, authentic leadership and role model. You know, recovery is not something to be ashamed of. It's something that's actually a source of pride, often committed to service and helping others. I find it almost paradoxical, because, on the one hand, alcohol is so visible in academia. You know, there's so many contexts in academia, at least in certain countries where alcohol and also alcohol abuse can be quite visible. But then there seems to be a silence around both the misuse of alcohol and the recovery from that misuse. Yeah, there definitely is a paradox around that. I had been sober for about two years at the time, and I successfully passed my defence and my two supervisors, the tradition was to have some champagne, and they had known me as a drinker as well. And then the third time I said no, thank you. He said, ”Come on, it's not like you're an alcoholic.” And I took the glass and I pretended to drink it and then I threw it away, but it was such a common kind of innocuous tradition, seemingly. But these are the kind of things we need to, I realize, educate people on how to be a recovery ally. Because if I had not been as secure in my recovery, I think, you know, that could have been a really a shakier moment than it was. And I'm just really grateful that I was able to kind of push through that and not pick up. But these things happen all the time. Yeah, so another recommendation from that research was to start a peer support group for faculty. I was really committed to creating a recovery-friendly workplace. So if someone did put their hand up and say, ”I'm struggling with substance use,” that they would get the help and be met with compassion and non judgment. There wasn't any recovery programs on campuses until 2019. And it was actually a student at University of British Columbia who started the first campus recovery program in Canada. And it was through that work, I was able to to hire a part time coordinator, and now we have 10 going on, 11 peer support meetings a week. We've grown from that $8,000 to over $5 million. With this in mind, what's your hope for the academia of the future, and academia where, instead of maybe exacerbating people's issues around substances, we provide a more supportive environment? I don't want a student or a staff member to feel the shame and stigma of being in recovery or of seeking help. The visibility of people in recovery, I think, is very important, because there's still such a narrow perception of what it means to, well, A, have an addiction, and B, what it means to be in recovery. So allowing folks who go to treatment, they have that community on campus where they have a safe place to go, that we're normalizing the idea that you could have a really enriching, fun, joyful university experience without the need to be intoxicated. There are many people who can use substances and safety, but there are also many people who can't. And up until recently, they didn't really have that space where they could fully be themselves and be around others on similar paths. Victoria Burns there, joined, as you might have heard towards the end of the interview, by her cat Charlie. Community can be so important for each and every stigmatized subject we've touched on in this series, from coming out as LGBTQIA+, to navigating research with a disability. And this is equally important for the subject of next week's episode: navigating grief. Until then, this has been Off Limits: Academia's Taboos, a podcast from Nature Careers. Universities in exile: displaced scholars count the costs of starting afresh NASA's latest telescope is a feat of early-career leadership AI could transform research assessment — and some academics are worried Universities in exile: displaced scholars count the costs of starting afresh Universities in exile: displaced scholars count the costs of starting afresh Epstein files reveal deeper ties to scientists than previously known AI could transform research assessment — and some academics are worried Aligning with world-class university standards, CityUHK (Dongguan) has now launched a global recruitment campaign for faculty members The University invites individuals from diverse backgrounds to apply for faculty positions in this field The University invites individuals from diverse backgrounds to apply for faculty positions in this field You have full access to this article via your institution. An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday. Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript. Advertisement Nature Communications , Article number: (2026) Cite this article We are providing an unedited version of this manuscript to give early access to its findings. Before final publication, the manuscript will undergo further editing. Please note there may be errors present which affect the content, and all legal disclaimers apply. Fracturing is unavoidable and threatens the reliability and functionality of materials. Therefore, regulating the propagation of cracks in a predictable and ductile manner is of paramount importance. Herein, we exploit the properties of topological mechanical metamaterials (TMMs) as a versatile mechanism to guide cracks unidirectionally and turn fracturing of lattices made of brittle materials into ductile events. Inspired by quantum topological states, recent discoveries of TMMs have uncovered varieties of unconventional mechanical phenomena, ranging from one-way wave propagation to polar elasticity. We show that polarized floppy modes occurring in TMMs lead to strongly asymmetric stress fields localizing around the notch tips, leading to ductile one-way fracturing, in sharp contrast to classical theories of fractures in brittle materials. Our work demonstrates the universality of this fracture unidirectionality feature, which is protected by TMM's bulk topology, and provides robust solutions in programmable fracturing for a broad class of materials and structures. The source experimental data underlying Figs. 1E and 2B are provided in the Source Data file. The code used to generate the results is archived on Zenodo (https://doi.org/10.5281/zenodo.18434954) and is also available at Github. Supplementary information and movies are provided with this manuscript. Any additional information is available from the corresponding authors upon request. Source data are provided with this paper. Allegre, C. J., Le Mouel, J. & Provost, A. Scaling rules in rock fracture and possible implications for earthquake prediction. Google Scholar Peterlik, H., Roschger, P., Klaushofer, K. & Fratzl, P. From brittle to ductile fracture of bone. Mater. Google Scholar Chai, H., Lee, J. J. W., Constantino, P. J., Lucas, P. W. & Lawn, B. R. Remarkable resilience of teeth. Google Scholar Shen, T., Song, Z., Cai, S. & Vernerey, F. J. Nonsteady fracture of transient networks: the case of vitrimer. Google Scholar Tang, J., Li, J., Vlassak, J. J. & Suo, Z. Fatigue fracture of hydrogels. Google Scholar Li, X. et al. Mesoscale bicontinuous networks in self-healing hydrogels delay fatigue fracture. Google Scholar Barney, C. W. et al. Fracture of model end-linked networks. Google Scholar Lin, S. et al. Anti-fatigue-fracture hydrogels. Adv. Google Scholar Berthier, E. et al. Rigidity percolation control of the brittle-ductile transition in disordered networks. Mater. Google Scholar Fan, Z. & Wang, S. Q. Resolving stress state at crack tip to elucidate nature of elastomeric fracture. Google Scholar Slootman, J. et al. Quantifying rate-and temperature-dependent molecular damage in elastomer fracture. Google Scholar Chen, S. & Ravi-Chandar, K. Nonlinear poroviscoelastic behavior of gelatin-based hydrogel. Google Scholar Yang, Y. et al. Intrinsic toughening and stable crack propagation in hexagonal boron nitride. Google Scholar Zhang, X. et al. Atomistic measurement and modeling of intrinsic fracture toughness of two-dimensional materials. Google Scholar Lew, A. J., Yu, C. H., Hsu, Y. C. & Buehler, M. J. Deep learning model to predict fracture mechanisms of graphene. npj 2D Mater. Google Scholar Odde, D. J., Ma, L., Briggs, A. H., DeMarco, A. & Kirschner, M. W. Microtubule bending and breaking in living fibroblast cells. Cell Sci. Google Scholar Morley, C. D. et al. Quantitative characterization of 3D bioprinted structural elements under cell-generated forces. Google Scholar Long, R., Hui, C. Y., Gong, J. P. & Bouchbinder, E. The fracture of highly deformable soft materials: A tale of two length scales. Annu. Condens. Matter Phys. Google Scholar Font-Clos, F. et al. Predicting the failure of two-dimensional silica glasses. Google Scholar Deng, B., Wang, S., Hartquist, C. & Zhao, X. Nonlocal intrinsic fracture energy of polymerlike networks. Google Scholar Kammer, D. S. et al. Earthquake energy dissipation in a fracture mechanics framework. Google Scholar Heisser, R. H., Patil, V. P., Stoop, N., Villermaux, E. & Dunkel, J. Controlling fracture cascades through twisting and quenching. Google Scholar Fuentealba, J. F., Hamm, E. & Roman, B. Intertwined multiple spiral fracture in perforated sheets. Google Scholar Zhang, L., Rocklin, D. Z., Sander, L. M. & Mao, X. Fiber networks below the isostatic point: Fracture without stress concentration. Mater. Google Scholar Driscoll, M. M. et al. The role of rigidity in controlling material failure. Google Scholar Ma, X. & Wei, Y. Programming fracture patterns of thin films. Google Scholar Wang, M., Shi, S. & Fineberg, J. Tensile cracks can shatter classical speed limits. Google Scholar Pons, A. J. & Karma, A. Helical crack-front instability in mixed-mode fracture. Google Scholar Vandenberghe, N., Vermorel, R. & Villermaux, E. Star-shaped crack pattern of broken windows. Google Scholar Marthelot, J. et al. Self-replicating cracks: a collaborative fracture mode in thin films. Google Scholar Takei, A., Roman, B., Bico, J., Hamm, E. & Melo, F. Forbidden directions for the fracture of thin anisotropic sheets: an analogy with the wulff plot. Google Scholar Bailey, N. P. & Sethna, J. P. Macroscopic measure of the cohesive length scale: fracture of notched single-crystal silicon. Google Scholar Hamm, E., Sivak, I. & Roman, B. Nature of crack path instabilities in thin sheets cut by blunt objects. Google Scholar Fuentealba, J. F., Marthelot, J., Roman, B. & Melo, F. Collaborative oscillatory fracture. Google Scholar Martins, A. F. et al. Locally-curved geometry generates bending cracks in the African elephant skin. Google Scholar Barthelat, F., Tang, H., Zavattieri, P., Li, C. M. & Espinosa, H. On the mechanics of mother-of-pearl: a key feature in the material hierarchical structure. Google Scholar Lew, A. J. et al. A Molecular-scale understanding of misorientation toughening in corals and seashells. Adv. Mater. Google Scholar Domino, L. et al. Fracture metamaterials with on-demand crack paths enabled by bending. arXiv preprint arXiv:240519061. Brodnik, N. et al. Fracture diodes: directional asymmetry of fracture toughness. Google Scholar Souto, A., Zhang, J., Aragón, A. M., Velikov, K. P. & Coulais, C. Edible mechanical metamaterials with designed fracture for mouthfeel control. Soft Matter 18, 2910–9 (2022). Google Scholar Mirkhalaf, M., Dastjerdi, A. K. & Barthelat, F. Overcoming the brittleness of glass through bio-inspiration and micro-architecture. Google Scholar Noah, P. M., Koning, V., Vitelli, V. & Irvine, W. T. M. Fracture in sheets draped on curved surfaces. Mater. Google Scholar Klein, Y. & Sharon, E. Buckling-fracture transition and the geometrical charge of a crack. Google Scholar Anderson, T. L. Fracture Mechanics: Fundamentals and Applications (CRC Press, 2005). Kane, C. L. & Lubensky, T. C. Topological boundary modes in isostatic lattices. Google Scholar Wang, P., Lu, L. & Bertoldi, K. Topological phononic crystals with one-way elastic edge waves. Google Scholar Nash, L. M. et al. Topological mechanics of gyroscopic metamaterials. Google Scholar Lubensky, T., Kane, C., Mao, X., Souslov, A. & Sun, K. Phonons and elasticity in critically coordinated lattices. Google Scholar Mao, X. & Lubensky, T. C. Maxwell lattices and topological mechanics. Matter Phys. Google Scholar Zhou, D., Zhang, L. & Mao, X. Topological edge floppy modes in disordered fiber networks. Google Scholar Sun, K. & Mao, X. Continuum theory for topological edge soft modes. Google Scholar Cheng, W. et al. Backscattering-free edge states below all bands in two-dimensional auxetic media. Google Scholar Rocklin, D. Z., Zhou, S., Sun, K. & Mao, X. Transformable topological mechanical metamaterials. Google Scholar Xiu, H. et al. Topological transformability and reprogrammability of multistable mechanical metamaterials. Google Scholar Jolly, J. C. et al. Soft mechanical metamaterials with transformable topology protected by stress caching. Google Scholar Paulose, J., Meeussen, A. S. & Vitelli, V. Selective buckling via states of self-stress in topological metamaterials. Google Scholar Zhang, L. & Mao, X. Fracturing of topological Maxwell lattices. Google Scholar Widstrand, C., Hu, C., Mao, X., Labuz, J. & Gonella, S. Stress focusing and damage protection in topological Maxwell metamaterials. Solids Struct. Google Scholar Widstrand, C., Mao, X. & Gonella, S. Robustness of stress focusing in soft lattices under topology-switching deformation. Google Scholar Ma, J., Zhou, D., Sun, K., Mao, X. & Gonella, S. Edge modes and asymmetric wave transport in topological lattices: Experimental characterization at finite frequencies. Google Scholar Charara, M., McInerney, J., Sun, K., Mao, X. & Gonella, S. Omnimodal topological polarization of bilayer networks: Analysis in the Maxwell limit and experiments on a 3D-printed prototype. Google Scholar de Waal, L., Chouzouris, M. & Dias, M. A. Architecting mechanisms of damage in topological metamaterials. Google Scholar de Waal, L., Chouzouris, M. & Dias, M. A. Cracking down on fracture to functionalize damage. Google Scholar Griffith, A. The phenomena of rupture and flow in solids. Google Scholar Lawn, B. Fracture of Brittle Solids (Cambridge University Press, 1993). Inglis, C. Stresses in a plate due to the presence of cracks and sharp corners. Google Scholar Zehnder, A. T. Lecture Notes on Fracture Mechanics. 20–22 (Cornell University. Charara, M., Sun, K., Mao, X. & Gonella, S. Topological flexural modes in polarized bilayer lattices. Google Scholar Janssen M, Zuidema J, Wanhill R. Fracture Mechanics: Fundamentals and Applications (CRC Press, 2004). Wang, X., Sarkar, S., Gonella, S. & Mao, X. Topological mechanical metamaterial for robust and ductile one-way fracturing. Available from: https://doi.org/10.5281/zenodo.18434954 (2026). Fleck, N. A. & Qiu, X. The damage tolerance of elastic–brittle, two-dimensional isotropic lattices. Google Scholar Jia, Z, Liu, F, Jiang, X, Wang, L Engineering lattice metamaterials for extreme property, programmability, and multifunctionality. Shekhawat, A., Zapperi, S. & Sethna, J. P. From damage percolation to crack nucleation through finite size criticality. Google Scholar Bouzid, M. & Del Gado, E. Network topology in soft gels: Hardening and softening materials. Google Scholar Qian, C. et al. Nanoscale phonon dynamics in self-assembled nanoparticle lattices. Liu, K., Zegard, T., Pratapa, P. P. & Paulino, G. H. Unraveling tensegrity tessellations for metamaterials with tunable stiffness and bandgaps. Google Scholar Gu, H., Pavier, M. & Shterenlikht, A. Experimental study of modulus, strength and toughness of 2D triangular lattices. Solids Struct. Google Scholar Wang, X., Guest, S. D. & Kamien, R. D. Keeping it together: interleaved kirigami extension assembly. Google Scholar Hutchinson, J. W. & Suo, Z. Mixed mode cracking in layered materials. Google Scholar Abaqus Analysis User's Manual (Dassault Systems, 2017);. Zhou, D., Zhang, L. & Mao, X. Topological boundary floppy modes in quasicrystals. Google Scholar Download references This work is supported by the National Science Foundation CMMI-2026794 (X.W. and CMMI-202700 (S.G.) and by the Office of Naval Research MURI N00014-20-1-2479 (X.W., S.S., and X.M.). We thank Fan Liu, Qi Zhang, Shi-Qing Wang, Mohammad Charara, Kai Sun, and Nick Kotov for fruitful discussions. We thank Andy Poli and Ellen Arruda's lab for their great help in setting up the experiments. We are grateful to the Advanced Research Computing at University of Michigan for access to the software and computational resources used in the simulations. Department of Physics, University of Michigan, Ann Arbor, MI, USA Xinyu Wang, Siddhartha Sarkar & Xiaoming Mao Department of Civil, Environmental, and Geo- Engineering, University of Minnesota, Minneapolis, MN, USA Stefano Gonella Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar Search author on:PubMed Google Scholar All the authors designed the project. conducted theoretical derivation. and S.S. performed numerical simulations. conducted experiments. All authors analyzed the data, discussed the results and contributed to writing and revising the manuscript. Correspondence to Xiaoming Mao. The authors declare no competing interests. Nature Communications thanks Corentin Coulais and Xiaoyan Li for their contribution to the peer review of this work. A peer review file is available. Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/. Reprints and permissions Wang, X., Sarkar, S., Gonella, S. et al. Topological mechanical metamaterial for robust and ductile one-way fracturing. Nat Commun (2026). Download citation Received: 04 November 2024 Accepted: 21 January 2026 Published: 06 February 2026 DOI: https://doi.org/10.1038/s41467-026-69026-7 Anyone you share the following link with will be able to read this content: Sorry, a shareable link is not currently available for this article. Provided by the Springer Nature SharedIt content-sharing initiative Nature Communications (Nat Commun) ISSN 2041-1723 (online) © 2026 Springer Nature Limited Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.