An AI-native experimental laboratory for autonomous biomolecular operations | Nature

Abstract Developing artificial intelligence (AI)-native experimental laboratories for autonomous scientific exploration with minimal human intervention represents a long-standing goal. Although autonomous experimental systems are emerging 1 , 2 , 3 , 4 , especially in chemical and materials science, they remain largely confined to domains with singular objectives, well-defined workflows and simple instrumentation. This limitation may reflect an ‘add-on’ architecture that uses AI as a task-specific tool rather than providing it with a global understanding of experiments, especially under engineering-oriented circumstances. Here we present AutoDNA, an AI-native autonomous biomolecular engineering laboratory targeting multi-objective experiments across diverse instruments. Through the co-design of models, experiments and instruments, AutoDNA realizes the co-evolution of AI and the automation system within an end-to-end multi-user framework. The system autonomously orchestrates over 20 types of instruments across more than 11,000 steps to execute core nucleic acid operations including synthesis, transcription, amplification and sequencing, while human intervention is limited to hardware maintenance tasks, such as reagent and labware replenishment. Across four representative applications, AutoDNA matched expert-optimized enzymatic DNA synthesis yields, accelerated concurrent protocol execution by 3.2-fold, completed an end-to-end DNA data storage cycle with exact recovery of the original digital content, and identified and validated a clinically detected KRAS G13D mutation. AutoDNA establishes a blueprint towards science-as-a-service, accelerating biological discovery and advancing various applications in synthetic biology and information storage. This is a preview of subscription content, access via your institution Access options Access through your institution Access Nature and 54 other Nature Portfolio journals Get Nature+, our best-value online-access subscription 32.70 € / 30 days cancel any time Learn more Subscribe to this journal Receive 51 print issues and online access 185.98 € per year only 3.65 € per issue Learn more Buy this article Purchase on SpringerLink Instant access to the full article PDF. 39,95 € Prices may be subject to local taxes which are calculated during checkout Fig. 1: The architecture of AutoDNA, an AI-native autonomous laboratory. Fig. 2: End-to-end autonomous workflow for the nucleic acid test service using AutoDNA. Fig. 3: Autonomous workflow for optimizing enzymatic DNA synthesis using AutoDNA. Fig. 4: Management of concurrent experiment requests by AutoDNA. Fig. 5: An end-to-end DNA data storage service supported by AutoDNA. Fig. 6: End-to-end cancer detection using AutoDNA. Explore related subjects Discover the latest articles and news in related subjects. Chemistry Computer science DNA nanotechnology Genome Editing Techniques in Plant and Animal Systems Data availability The data supporting the findings of this study are available in the Article and its Supplementary Information . The sequencing data generated in the enzymatic DNA synthesis and DNA data storage experiments, together with the agent-related outputs, are available at Zenodo ( https://doi.org/10.5281/zenodo.20624108 ) 53 . Reference genome hg38 was downloaded from https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_000001405.40/ . Further data are available from the corresponding authors upon reasonable request. Source data are provided with this paper. Code availability The code is available at GitHub ( https://github.com/IPADS-X/AutoDNA ). The version used in this study is available at Zenodo ( https://doi.org/10.5281/zenodo.20624108 ) 53 . References Burger, B. et al. A mobile robotic chemist. Nature 583 , 237–241 (2020). Article CAS PubMed Google Scholar Boiko, D. A., MacKnight, R., Kline, B. & Gomes, G. Autonomous chemical research with large language models. Nature 624 , 570–578 (2023). Article CAS PubMed PubMed Central Google Scholar Dai, T. et al. Autonomous mobile robots for exploratory synthetic chemistry. Nature 635 , 890–897 (2024). Article PubMed PubMed Central Google Scholar Ha, T. et al. AI-driven robotic chemist for autonomous synthesis of organic molecules. Sci. Adv. 9 , eadj0461 (2023). Article CAS PubMed PubMed Central Google Scholar Zhang, Z. et al. A multimodal robotic platform for multi-element electrocatalyst discovery. Nature 647 , 390–396 (2025). Article CAS PubMed Google Scholar Jia, Y. et al. Robot-assisted mapping of chemical reaction hyperspaces and networks. Nature 645 , 922–931 (2025). Article CAS PubMed PubMed Central Google Scholar Steiner, S. et al. Organic synthesis in a modular robotic system driven by a chemical programming language. Science 363 , eaav2211 (2019). Article CAS PubMed Google Scholar Mehr, S. H. M., Craven, M., Leonov, A. I., Keenan, G. & Cronin, L. A universal system for digitization and automatic execution of the chemical synthesis literature. Science 370 , 101–108 (2020). Article CAS PubMed Google Scholar Abolhasani, M. & Kumacheva, E. The rise of self-driving labs in chemical and materials sciences. Nat. Synth. 2 , 483–492 (2023). Article CAS Google Scholar Seifrid, M. et al. Autonomous chemical experiments: challenges and perspectives on establishing a self-driving lab. Acc. Chem. Res. 55 , 2454–2466 (2022). Article CAS PubMed PubMed Central Google Scholar Darvish, K. et al. ORGANA: a robotic assistant for automated chemistry experimentation and characterization. Matter 8 , 101897 (2025). Article Google Scholar Song, T. et al. A multiagent-driven robotic AI chemist enabling autonomous chemical research on demand. J. Am. Chem. Soc. 147 , 12534–12545 (2025). Article CAS PubMed Google Scholar Touvron, H. et al. LLaMA: open and efficient foundation language models. Preprint at https://doi.org/10.48550/arXiv.2302.13971 (2023). Introducing ChatGPT. OpenAI https://openai.com/index/chatgpt/ (2022). Guo, D. et al. DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning. Nature 645 , 633–638 (2025). Article CAS PubMed PubMed Central Google Scholar Introducing Claude 3.5 Sonnet. Anthropic https://www.anthropic.com/news/claude-3-5-sonnet (2024). Brown, T. B. et al. Language models are few-shot learners. In Proc. Advances in Neural Information Processing Systems , Vol. 33 (eds Larochelle, H. et al.) 1877–1901 (Curran Associates, 2020). Swanson, K., Wu, W., Bulaong, N. L., Pak, J. E. & Zou, J. The virtual lab of AI agents designs new SARS-CoV-2 nanobodies. Nature 646 , 716–723 (2025). Article CAS PubMed Google Scholar Yu, B., Baker, F. N., Chen, Z., Ning, X. & Sun, H. LlaSMol: advancing large language models for chemistry with a large-scale, comprehensive, high-quality instruction tuning dataset. In Proc. First Conference on Language Modeling https://openreview.net/forum?id=lY6XTF9tPv (2024). Jablonka, K. M., Schwaller, P., Ortega-Guerrero, A. & Smit, B. Leveraging large language models for predictive chemistry. Nat. Mach. Intell. 6 , 161–169 (2024). Article Google Scholar Antunes, L. M., Butler, K. T. & Grau-Crespo, R. Crystal structure generation with autoregressive large language modeling. Nat. Commun. 15 , 10570 (2024). Article CAS PubMed PubMed Central Google Scholar Frey, N. C. et al. Neural scaling of deep chemical models. Nat. Mach. Intell. 5 , 1297–1305 (2023). Article Google Scholar Charting the Future of Biotechnology https://www.biotech.senate.gov/final-report/chapters/ (National Security Commission of Emerging Biotechnology, 2025). Lu, X. et al. Enzymatic DNA synthesis by engineering terminal deoxynucleotidyl transferase. ACS Catal. 12 , 2988–2997 (2022). Article CAS Google Scholar Meiser, L. C. et al. Reading and writing digital data in DNA. Nat. Protoc. 15 , 86–101 (2020). Article CAS PubMed Google Scholar Daher, R. K., Stewart, G., Boissinot, M. & Bergeron, M. G. Recombinase polymerase amplification for diagnostic applications. Clin. Chem. 62 , 947–958 (2016). Article CAS PubMed PubMed Central Google Scholar Wang, D. Prizes must recognize machine contributions to discovery. Nature 646 , 261 (2025). Article CAS PubMed Google Scholar Gao, S. et al. Empowering biomedical discovery with AI agents. Cell 187 , 6125–6151 (2024). Article CAS PubMed Google Scholar Ramos, M. C., Collison, C. J. & White, A. D. A review of large language models and autonomous agents in chemistry. Chem. Sci. 16 , 2514–2572 (2025). Article CAS PubMed Google Scholar Learning to Reason with LLMs . OpenAI https://openai.com/index/learning-to-reason-with-llms/ (2024). Ouyang, L. et al. Training language models to follow instructions with human feedback. In Proc. Advances in Neural Information Processing Systems , Vol. 35 (eds Koyejo, S. et al.) 27730–27744 (Curran Associates, 2022). Wang, L., Zhang, X., Su, H. & Zhu, J. A comprehensive survey of continual learning: theory, method and application. IEEE Trans. Pattern Anal. Mach. Intell. 46 , 5362–5383 (2024). Article PubMed Google Scholar Tanaka, T. & Letsinger, R. L. Syringe method for stepwise chemical synthesis of oligonucleotides. Nucleic Acids Res. 10 , 3249–3259 (1982). Article CAS PubMed PubMed Central Google Scholar Hoose, A., Vellacott, R., Storch, M., Freemont, P. S. & Ryadnov, M. G. DNA synthesis technologies to close the gene writing gap. Nat. Rev. Chem. 7 , 144–161 (2023). Article CAS PubMed PubMed Central Google Scholar Lee, H. et al. Photon-directed multiplexed enzymatic DNA synthesis for molecular digital data storage. Nat. Commun. 11 , 5246 (2020). Article CAS PubMed PubMed Central Google Scholar Lewis, P. et al. Retrieval-augmented generation for knowledge-intensive NLP tasks. In Proc. Advances in Neural Information Processing Systems , Vol. 33 (eds Larochelle, H. et al.) 9459–9474 (Curran Associates, 2020). Bayley, O., Savino, E., Slattery, A. & Noël, T. Autonomous chemistry: navigating self-driving labs in chemical and material sciences. Matter 7 , 2382–2398 (2024). Article CAS Google Scholar Bran, A. M. et al. Augmenting large language models with chemistry tools. Nat. Mach. Intell. 6 , 525–535 (2024). Article Google Scholar Ruan, Y. et al. An automatic end-to-end chemical synthesis development platform powered by large language models. Nat. Commun. 15 , 10160 (2024). Article CAS PubMed PubMed Central Google Scholar Balandat, M. et al. BoTorch: a framework for efficient Monte-Carlo Bayesian optimization. In Proc. Advances in Neural Information Processing Systems , Vol. 33 (eds Larochelle, H. et al.) 21524–21538 (Curran Associates, 2020). Li, K. et al. Highly ordered DNA framework interface enables efficient enzymatic oligonucleotide synthesis. Adv. Sci. 12 , e05868 (2025). Article CAS Google Scholar Li, K. et al. DNA-DISK: automated end-to-end data storage via enzymatic single-nucleotide DNA synthesis and sequencing on digital microfluidics. Proc. Natl Acad. Sci. USA 121 , e2410164121 (2024). Article CAS PubMed PubMed Central Google Scholar Takahashi, C. N., Nguyen, B. H., Strauss, K. & Ceze, L. Demonstration of end-to-end automation of DNA data storage. Sci. Rep. 9 , 4998 (2019). Article PubMed PubMed Central Google Scholar Zhirnov, V., Zadegan, R. M., Sandhu, G. S., Church, G. M. & Hughes, W. L. Nucleic acid memory. Nat. Mater. 15 , 366–370 (2016). Article CAS PubMed PubMed Central Google Scholar Church, G. M., Gao, Y. & Kosuri, S. Next-generation digital information storage in DNA. Science 337 , 1628 (2012). Article CAS PubMed Google Scholar Cherry, K. M. & Qian, L. Supervised learning in DNA neural networks. Nature 645 , 639–647 (2025). Article CAS PubMed PubMed Central Google Scholar Zhang, C. et al. Parallel molecular data storage by printing epigenetic bits on DNA. Nature 634 , 824–832 (2024). Article CAS PubMed PubMed Central Google Scholar Yang, S. et al. DNA as a universal chemical substrate for computing and data storage. Nat. Rev. Chem. 8 , 179–194 (2024). Article PubMed Google Scholar Gottweis, J. et al. Accelerating scientific discovery with Co-Scientist. Nature 655 , 487–496 (2026). Article CAS PubMed PubMed Central Google Scholar Ghareeb, A. E. et al. A multi-agent system for automating scientific discovery. Nature 655 , 497–505 (2026). Article CAS PubMed PubMed Central Google Scholar Yao, S. et al. ReAct: synergizing reasoning and acting in language models. In Proc. 11th International Conference on Learning Representations https://openreview.net/forum?id=WE_vluYUL-X (2023). Skarlinski, M. D. et al. Language agents achieve superhuman synthesis of scientific knowledge. Preprint at https://doi.org/10.48550/arXiv.2409.13740 (2024). Wu, M. et al. AutoDNA code and data for autonomous biomolecular operations. Zenodo https://doi.org/10.5281/zenodo.20624108 (2026). Download references Acknowledgements During the preparation of this work, ChatGPT (OpenAI) and DeepSeek were used to assist with language editing. All AI-assisted content was reviewed and edited by the authors. No generative AI tools were used to generate or modify experimental data or scientific conclusions. Funding This work was supported by the National Natural Science Foundation of China (T2188102, 62132014 and 62422209), the Shanghai Pilot Program for Basic Research, the Shanghai Municipal Science and Technology Major Project, the Fundamental and Interdisciplinary Disciplines Breakthrough Plan of the Ministry of Education of China (JYB2025XDXM113 and JYB2025XDXM602) and the New Cornerstone Science Foundation. Author information Author notes These authors contributed equally: Mingyu Wu, Zhaoguo Wang, Jiabin Wang Authors and Affiliations Institute of Parallel and Distributed Systems, School of Computer Science, Institute of Artificial Intelligence Operating Systems, Engineering Research Center for Domain-specific Operating Systems (Ministry of Education), Zhang Jiang Institute for Advanced Study, Shanghai Jiao Tong University, Shanghai, China Mingyu Wu, Zhaoguo Wang, Zhiyuan Dong, Jingkai Yang, Tianyu Huang & Haibo Chen State Key Laboratory of Synergistic Chem-Bio Synthesis, School of Chemistry and Chemical Engineering, New Cornerstone Science Laboratory, Frontiers Science Center for Transformative Molecules, Zhang Jiang Institute for Advanced Study and National Center for Translational Medicine, Shanghai Jiao Tong University, Shanghai, China Jiabin Wang, Qingting Li, Chi Zhang, Lei Zhao, Mingqiang Li, Xiaolei Zuo, Fei Wang & Chunhai Fan Jiaxing Key Laboratory of Biosemiconductors (A); Xiangfu Laboratory, Jiashan, China Jiabin Wang Yuanzi Zhishu Technology Co. Ltd, Shanghai, China Zhiyuan Dong School of Biomedical Engineering, Department of Gynecologic Oncology, The International Peace Maternity and Child Health Hospital, Shanghai Jiao Tong University, Shanghai, China Ping Song Institute of Molecular Medicine, Shanghai Key Laboratory for Nucleic Acids Chemistry and Nanomedicine, Renji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China Xiaolei Zuo Authors Mingyu Wu View author publications Search author on: PubMed Google Scholar Zhaoguo Wang View author publications Search author on: PubMed Google Scholar Jiabin Wang View author publications Search author on: PubMed Google Scholar Zhiyuan Dong View author publications Search author on: PubMed Google Scholar Jingkai Yang View author publications Search author on: PubMed Google Scholar Qingting Li View author publications Search author on: PubMed Google Scholar Tianyu Huang View author publications Search author on: PubMed Google Scholar Chi Zhang View author publications Search author on: PubMed Google Scholar Ping Song View author publications Search author on: PubMed Google Scholar Lei Zhao View author publications Search author on: PubMed Google Scholar Mingqiang Li View author publications Search author on: PubMed Google Scholar Xiaolei Zuo View author publications Search author on: PubMed Google Scholar Fei Wang View author publications Search author on: PubMed Google Scholar Chunhai Fan View author publications Search author on: PubMed Google Scholar Haibo Chen View author publications Search author on: PubMed Google Scholar Contributions C.F. and H.C. conceived the research. Z.W. led and supervised the research. M.W. and Z.W. designed the multi-agent architecture and constructed the AutoDNA experimental platform. J.W. and Z.D. contributed to most aspects of the experimental design and wet-lab preparation. J.Y., T.H. and Z.D. implemented the agents and LLM workflows. C.Z. conducted the manual DNA synthesis control experiments. P.S. and X.Z. contributed to the experimental design for cancer detection and provided clinical samples. M.W. and Z.W. analysed the data and wrote the paper. J.W., L.Z., M.L., Q.L., F.W., C.F. and H.C. participated in the data analysis and discussions. Z.W., F.W., C.F. and H.C. reviewed and edited the Article. Corresponding authors Correspondence to Zhaoguo Wang , Fei Wang , Chunhai Fan or Haibo Chen . Ethics declarations Competing interests The authors declare no competing interests. Peer review Peer review information Nature thanks Milad Abolhasani and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available. Additional information Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Extended data figures and tables Extended Data Fig. 1 The automated hardware platform of AutoDNA. a , The platform contains a series of workstations with distinct functionalities. b , A robotic arm (AUBO-i10) is responsible for transferring reagents among workstations. Extended Data Fig. 2 Engineering design of the AutoDNA system for automated Nanopore sequencing. a , Automated operation of Nanopore Mk1C is enabled by a high-precision robotic arm guided by a dual-camera machine vision system. b , The original cover of the sequencing flow cell is replaced with a custom-designed one engineered for robotic gripping. c , The robotic arm’s gripper can be directly used for the installation and removal of the custom cover. d , A high-precision camera (GTS-8016) is used to precisely locate the flow cell’s sampling port (~2 mm in diameter). e , An automated pipetting module (Keyto-SP16) dispenses liquid directly into the identified sampling port of the flow cell. Extended Data Fig. 3 The user interface of AutoDNA. The main panel displays the output of the current step, corresponding to the step highlighted in the Agent Workflow panel on the right. Extended Data Fig. 4 Comparison of AutoDNA and manual NAT performance. a , b , Representative real-time fluorescence curves from one of three independent experiments for the NAT assay executed by the AutoDNA ( a ) and by manual operation ( b ). Source data Extended Data Fig. 5 Normalized fluorescence kinetics in the strand-displacement assay for siRNA validation. a , Commercially synthesized siRNA as the positive control and the corresponding no-template control (NTC). b , siRNA generated by AutoDNA and the corresponding NTC. c , An incorrect siRNA sequence as the negative control and the corresponding NTC. Solid lines represent mean values and shaded areas indicate s.d. from three independent experiments (n = 3 per curve). Source data Extended Data Fig. 6 PAGE analysis results during iterative optimizations. Each iteration included four separately synthesized samples. Known-length oligonucleotides are indicated as size references. For gel source data, see Supplementary Fig. 1 . Source data Extended Data Fig. 7 Timeline comparison of concurrent and sequential execution time. Four enzymatic DNA synthesis protocols were executed concurrently after static code merging (Concurrent-1–4) or sequentially (Sequence). Coloured segments show the timing and duration of individual instrument operations, and arrows indicate protocol completion. The timeline shows a representative run from two independent experiments. Source data Extended Data Fig. 8 Real-time qPCR amplification curves for MUT, WT, and no-template control (NTC) samples using commercially synthesized primers. a , Representative real-time qPCR amplification curves from one of four independent experiments. The MUT sample shows an earlier and stronger fluorescence increase compared with WT, while NTC remains flat throughout most of the amplification cycles, indicating good specificity and low background signal. b , Corresponding Cq values for MUT, WT and NTC. MUT exhibits the lowest Cq value, followed by WT, whereas NTC shows a substantially delayed signal with a Cq value close to the end of the tested cycle range. Data are presented as mean ± s.d. from four independent experiments (n = 4 per sample), with three technical replicates per sample in each independent experiment. Each data point represents the mean of the three technical replicates within one independent experiment. Source data Supplementary information Supplementary Information (download PDF ) Supplementary Methods, Notes, Figs. 1–32, Tables 1–39 and References. Reporting Summary (download PDF ) Supplementary Data 1 (download XLSX ) Supplementary Data 1 contains the curated literature dataset used to construct the literature-tuned Bayesian optimization baseline, together with field definitions and a record of data curation operations, including value filling, sequence descriptor calculation, wash-count mapping, scaffold annotation and yield conversion. Peer Review File (download PDF ) Supplementary Video 1 (download MP4 ) This video demonstrates the end-to-end workflow of the nucleic acid test experiment. Supplementary Video 2 (download MP4 ) This video demonstrates how AutoDNA autonomously operates the nanopore sequencing device. Supplementary Video 3 (download MP4 ) This video shows how AutoDNA schedules and handles several experimental requests concurrently. Supplementary Video 4 (download MP4 ) This video provides a close-up view of reagent transfer within (intra-instrument) and between (inter-instrument) workstations. Supplementary Video 5 (download MP4 ) This video demonstrates AutoDNA autonomously completing a single step of the enzymatic synthesis process. Supplementary Video 6 (download MP4 ) This video demonstrates how AutoDNA recovers from two types of hardware error: pipette–tip pickup error and pipette malfunction error. Source data Source Data Fig. 2 (download XLSX ) Source Data Fig. 3 (download XLSX ) Source Data Fig. 4 (download XLSX ) Source Data Fig. 5 (download XLSX ) Source Data Fig. 6 (download XLSX ) Source Data Extended Data Fig. 4 (download XLSX ) Source Data Extended Data Fig. 5 (download XLSX ) Source Data Extended Data Fig. 6 (download XLSX ) Source Data Extended Data Fig. 7 (download XLSX ) Source Data Extended Data Fig. 8 (download XLSX ) Rights and permissions Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law. Reprints and permissions About this article Cite this article Wu, M., Wang, Z., Wang, J. et al. An AI-native experimental laboratory for autonomous biomolecular operations. Nature (2026). https://doi.org/10.1038/s41586-026-11228-6 Download citation Received : 01 October 2025 Accepted : 30 September 2026 Published : 04 November 2026 Version of record : 04 November 2026 DOI : https://doi.org/10.1038/s41586-026-11228-6 Share this article Anyone you share the following link with will be able to read this content: Get shareable link Sorry, a shareable link is not currently available for this article. Copy shareable link to clipboard Provided by the Springer Nature SharedIt content-sharing initiative