Skip to main content

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
  • Published:

Bridging chemistry and artificial intelligence by a reaction description language

Nature Machine Intelligence volume 7, pages 782–793 (2025)Cite this article

Subjects

A preprint version of the article is available at ChemRxiv.

Abstract

With the fast-paced development of artificial intelligence, large language models are increasingly used to tackle various scientific challenges. A critical step in this process is converting domain-specific data into a sequence of tokens for language modelling. In chemistry, molecules are often represented by molecular linear notations, and chemical reactions are depicted as sequence pairs of reactants and products. However, this approach does not capture atomic and bond changes during reactions. Here, we present ReactSeq, a reaction description language that defines molecular editing operations for step-by-step chemical transformation. Based on ReactSeq, language models for retrosynthesis prediction may consistently excel in all benchmark tests, and demonstrate promising emergent abilities in the human-in-the-loop and explainable artificial intelligence. Moreover, ReactSeq has allowed us to obtain universal and reliable representations of chemical reactions, which enable navigation of the reaction space and aid in the recommendation of experimental procedures and prediction of reaction yields. We foresee that ReactSeq can serve as a bridge to narrow the gap between chemistry and artificial intelligence.

Access through your institution
Buy or subscribe

This is a preview of subscription content, access via your institution

Access options

Access through your institution

Buy this article

  • Purchase on SpringerLink
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Overview of this work.
Fig. 2: Illustration of ReactSeq.
Fig. 3: Interpretable retrosynthesis prediction with ReactSeq.
Fig. 4: The prompt-based learning with ReactSeq.
Fig. 5: Representations of chemical reactions and their applications.

Similar content being viewed by others

Data availability

All datasets used in this work are available at https://github.com/jiachengxiong/ReactSeq, https://drive.google.com/drive/folders/1a6NL5apcP_7isY3HccLjkSsjJGwp_FwD and via Zenodo at https://doi.org/10.5281/zenodo.13338263 (ref. 51). Source data are provided with this paper.

Code availability

All code for the generating and transforming of ReactSeq, as well as the code for model training and inference, is available at https://github.com/jiachengxiong/ReactSeq and via Zenodo at https://doi.org/10.5281/zenodo.13338263 (ref. 51). The webpage for our prompt learning model is available at https://huggingface.co/spaces/Oopstom/ReactSeq.

References

  1. Thirunavukarasu, A. J. et al. Large language models in medicine. Nat. Med. 29, 1930–1940 (2023).

    Article  Google Scholar 

  2. Boiko, D. A., MacKnight, R., Kline, B. & Gomes, G. Autonomous chemical research with large language models. Nature 624, 570–578 (2023).

    Article  Google Scholar 

  3. Rives, A. et al. Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences. Proc. Natl Acad. Sci. USA 118, e2016239118 (2021).

    Article  Google Scholar 

  4. Lin, Z. et al. Evolutionary-scale prediction of atomic-level protein structure with a language model. Science 379, 1123–1130 (2023).

    Article  MathSciNet  Google Scholar 

  5. Theodoris, C. V. et al. Transfer learning enables predictions in network biology. Nature 618, 616–624 (2023).

    Article  Google Scholar 

  6. Grisoni, F. Chemical language models for de novo drug design: challenges and opportunities. Curr. Opin. Struct. Biol. 79, 102527 (2023).

    Article  Google Scholar 

  7. Skinnider, M. A., Stacey, R. G., Wishart, D. S. & Foster, L. J. Chemical language models enable navigation in sparsely populated chemical space. Nat. Mach. Intell. 3, 759–770 (2021).

    Article  Google Scholar 

  8. Moret, M. et al. Leveraging molecular structure and bioactivity with chemical language models for de novo drug design. Nat. Commun. 14, 114 (2023).

    Article  Google Scholar 

  9. Weininger, D. SMILES, a chemical language and information system. 1. Introduction to methodology and encoding rules. J. Chem. Inf. Comput. Sci. 28, 31–36 (1988).

    Article  Google Scholar 

  10. Krenn, M., Häse, F., Nigam, A., Friederich, P. & Aspuru-Guzik, A. Self-referencing embedded strings (SELFIES): a 100% robust molecular string representation. Mach. Learn. Sci. Technol. 1, 045024 (2020).

    Article  Google Scholar 

  11. Kuenneth, C. & Ramprasad, R. polyBERT: a chemical language model to enable fully machine-driven ultrafast polymer informatics. Nat. Commun. 14, 4099 (2023).

    Article  Google Scholar 

  12. Krenn, M. et al. SELFIES and the future of molecular string representations. Patterns 3, 100588 (2022).

    Article  Google Scholar 

  13. Schwaller, P. et al. Molecular transformer: a model for uncertainty-calibrated chemical reaction prediction. ACS Cent. Sci. 5, 1572–1583 (2019).

    Article  Google Scholar 

  14. Liu, B. et al. Retrosynthetic reaction prediction using neural sequence-to-sequence models. ACS Cent. Sci. 3, 1103–1113 (2017).

    Article  Google Scholar 

  15. Sun, Y. & Sahinidis, N. V. Computer-aided retrosynthetic design: fundamentals, tools, and outlook. Curr. Opin. Chem. Eng. 35, 100721 (2022).

    Article  Google Scholar 

  16. Wang, X. et al. RetroPrime: a diverse, plausible and transformer-based method for single-step retrosynthesis predictions. Chem. Eng. J. 420, 129845 (2021).

    Article  Google Scholar 

  17. Thakkar, A. et al. Unbiasing retrosynthesis language models with disconnection prompts. ACS Cent. Sci. 9, 1488–1498 (2023).

    Article  Google Scholar 

  18. Huang, T. & Li, Y. Current progress, challenges, and future perspectives of language models for protein representation and protein design. The Innovation 4, 100446 (2023).

    Article  Google Scholar 

  19. Min, B. et al. Recent advances in natural language processing via large pre-trained language models: a survey. ACM Comput. Surv. 56, 1–40 (2023).

    Article  Google Scholar 

  20. Schwaller, P., Hoover, B., Reymond, J.-L., Strobelt, H. & Laino, T. Extraction of organic chemistry grammar from unsupervised learning of chemical reactions. Sci. Adv. 7, eabe4166 (2021).

    Article  Google Scholar 

  21. Schwaller, P. et al. Mapping the space of chemical reactions using attention-based neural networks. Nat. Mach. Intell. 3, 144–152 (2021).

    Article  Google Scholar 

  22. Strieth-Kalthoff, F. et al. Artificial intelligence for retrosynthetic planning needs both data and expert knowledge. J. Am. Chem. Soc. 146, 11005–11017 (2024).

    Google Scholar 

  23. Nugmanov, R. I. et al. CGRtools: Python library for molecule, reaction, and condensed graph of reaction processing. J. Chem. Inf. Model. 59, 2516–2521 (2019).

    Article  Google Scholar 

  24. Shi, C., Xu, M., Guo, H., Zhang, M. & Tang, J. A graph to graphs framework for retrosynthesis prediction. In Proc. 37th International Conference on Machine Learning (eds Blei, D. et al.) 8818–8827 (PMLR, 2020).

  25. Yan, C. et al. Retroxpert: decompose retrosynthesis prediction like a chemist. Adv. Neural Inf. Process. Syst. 33, 11248–11258 (2020).

  26. Somnath, V. R., Bunne, C., Coley, C., Krause, A. & Barzilay, R. Learning graph models for retrosynthesis prediction. Adv. Neural Inf. Process. Syst. 34, 9405–9415 (2021).

  27. Zhong, W., Yang, Z. & Chen, C. Y.-C. Retrosynthesis prediction using an end-to-end graph generative architecture for molecular graph editing. Nat. Commun. 14, 3009 (2023).

    Article  Google Scholar 

  28. Wang, Y. et al. Retrosynthesis prediction with an interpretable deep-learning framework based on molecular assembly tasks. Nat. Commun. 14, 6155 (2023).

    Article  Google Scholar 

  29. Saebi, M. et al. On the use of real-world datasets for reaction yield prediction. Chem. Sci. 14, 4997–5005 (2023).

    Article  Google Scholar 

  30. Lu, J. & Zhang, Y. Unified deep learning model for multitask reaction predictions with explanation. J. Chem. Inf. Model. 62, 1376–1387 (2022).

    Article  Google Scholar 

  31. Wan, Y., Hsieh, C.-Y., Liao, B. & Zhang, S. Retroformer: pushing the limits of end-to-end retrosynthesis transformer. In Proc. 39th International Conference on Machine Learning (eds Chaudhuri, K. et al.) 22475–22490 (PMLR, 2022).

  32. Dong, J. et al. Ketones and aldehydes as alkyl radical equivalents for C-H functionalization of heteroarenes. Sci. Adv. 5, eaax9955 (2019).

    Article  Google Scholar 

  33. Peltzer, R. M., Gauss, J., Eisenstein, O. & Cascella, M. The Grignard reaction–unraveling a chemical puzzle. J. Am. Chem. Soc. 142, 2984–2994 (2020).

    Article  Google Scholar 

  34. Heravi, M. M., Hashemi, E. & Nazari, N. Negishi coupling: an easy progress for C–C bond construction in total synthesis. Mol. Divers. 18, 441–472 (2014).

    Article  Google Scholar 

  35. Kotha, S., Lahiri, K. & Kashinath, D. Recent applications of the Suzuki–Miyaura cross-coupling reaction in organic synthesis. Tetrahedron 58, 9633–9695 (2002).

    Article  Google Scholar 

  36. Zhou, J., Zhao, Z. & Shibata, N. Transition-metal-free silylboronate-mediated cross-couplings of organic fluorides with amines. Nat. Commun. 14, 1847 (2023).

    Article  Google Scholar 

  37. Vulovic, B., Cinderella, A. P. & Watson, D. A. Palladium-catalyzed cross-coupling of monochlorosilanes and Grignard reagents. ACS Catal. 7, 8113–8117 (2017).

    Article  Google Scholar 

  38. Xu, W. Q., Xu, X. H. & Qing, F. L. Synthesis and properties of CF3 (OCF3) CH‐substituted arenes and alkenes. Chin. J. Chem. 38, 847–854 (2020).

    Article  Google Scholar 

  39. Van der Maaten, L. & Hinton, G. Visualizing data using t-SNE. J. Mach. Learn. Res. 9, 2579–2605 (2008).

    Google Scholar 

  40. Probst, D., Schwaller, P. & Reymond, J.-L. Reaction classification and yield prediction using the differential reaction fingerprint DRFP. Digit. Discov. 1, 91–97 (2022).

    Article  Google Scholar 

  41. Schneider, N., Lowe, D. M., Sayle, R. A. & Landrum, G. A. Development of a novel fingerprint for chemical reactions and its application to large-scale reaction classification and similarity. J. Chem. Inf. Model. 55, 39–53 (2015).

    Article  Google Scholar 

  42. Kajino, M., Hasuoka, A. & Nishida, H. 1-heterocyclylsulfonyl, 2-aminomethyl, 5- (hetero-) aryl substituted 1-H-pyrrole derivatives as acid secretion inhibitors. Patent WO2007026916A1 (2007).

  43. Yu, Q.-Y., Zeng, H., Yao, K., Li, J.-Q. & Liu, Y. Novel and practical synthesis of vonoprazan fumarate. Synth. Commun. 47, 1169–1174 (2017).

    Article  Google Scholar 

  44. Chen, S. & Jung, Y. Deep retrosynthetic reaction prediction using local reactivity and global attention. JACS Au. 1, 1612–1620 (2021).

    Article  Google Scholar 

  45. Tetko, I. V., Karpov, P., Van Deursen, R. & Godin, G. State-of-the-art augmented NLP transformer models for direct and single-step retrosynthesis. Nat. Commun. 11, 5575 (2020).

    Article  Google Scholar 

  46. Zhong, Z. et al. Root-aligned SMILES: a tight representation for chemical reaction prediction. Chem. Sci. 13, 9023–9034 (2022).

    Article  Google Scholar 

  47. Irwin, R., Dimitriadis, S., He, J. & Bjerrum, E. J. Chemformer: a pre-trained transformer for computational chemistry. Mach. Learn. Sci. Technol. 3, 015022 (2022).

    Article  Google Scholar 

  48. Zdrazil, B. et al. The ChEMBL Database in 2023: a drug discovery platform spanning multiple bioactivity data types and time periods. Nucleic Acids Res. 52, D1180–D1192 (2024).

    Article  Google Scholar 

  49. Tingle, B. I. et al. ZINC-22—a free multi-billion-scale database of tangible compounds for ligand discovery. J. Chem. Inf. Model. 63, 1166–1176 (2023).

    Article  Google Scholar 

  50. Chilingaryan, G. et al. BartSmiles: generative masked language models for molecular representations. J. Chem. Inf. Model. 64, 5832–5843 (2024).

    Article  Google Scholar 

  51. zw-SIMM & Xiong, J. jiachengxiong/ReactSeq: ReactSeq (1.0). Zenodo https://doi.org/10.5281/zenodo.13338263 (2024).

  52. Coley, C. W., Rogers, L., Green, W. H. & Jensen, K. F. Computer-assisted retrosynthesis based on molecular similarity. ACS Cent. Sci. 3, 1237–1245 (2017).

    Article  Google Scholar 

  53. Segler, M. H. & Waller, M. P. Neural‐symbolic machine learning for retrosynthesis and reaction prediction. Chem-Eur. J. 23, 5966–5971 (2017).

    Article  Google Scholar 

  54. Dai, H., Li, C., Coley, C., Dai, B. & Song, L. Retrosynthesis prediction with conditional graph logic network. Adv. Neural Inf. Process. Syst. 32, 8872–8882 (2019).

  55. Sacha, M. et al. Molecule edit graph attention network: modeling chemical reactions as sequences of graph edits. J. Chem. Inf. Model. 61, 3273–3284 (2021).

    Article  Google Scholar 

  56. Chen, Z., Ayinde, O. R., Fuchs, J. R., Sun, H. & Ning, X. G2Retro as a two-step graph generative models for retrosynthesis prediction. Commun. Chem. 6, 102 (2023).

    Article  Google Scholar 

  57. Yao, L. et al. Node-aligned graph-to-graph: elevating template-free deep learning approaches in single-step retrosynthesis. JACS Au. 4, 992–1003 (2024).

    Article  Google Scholar 

  58. Liu, X. et al. RetroCaptioner: beyond attention in end-to-end retrosynthesis transformer via contrastively captioned learnable graph representation. Bioinformatics 40, btae561 (2024).

    Article  Google Scholar 

  59. Zheng, S., Rao, J., Zhang, Z., Xu, J. & Yang, Y. Predicting retrosynthetic reactions using self-corrected transformer neural networks. J. Chem. Inf. Model. 60, 47–55 (2019).

    Article  Google Scholar 

  60. Sun, R., Dai, H., Li, L., Kearnes, S. & Dai, B. Towards understanding retrosynthesis by energy-based models. Adv. Neural Inf. Process. Syst. 34, 10186–10194 (2021).

Download references

Acknowledgements

We gratefully acknowledge financial support from the National Natural Science Foundation of China (grant nos. T2225002 and 82273855, M.Z.), the National Key Research and Development Program of China (grant nos. 2022YFC3400504 and 2023YFC2305904, M.Z.), the Strategic Priority Research Program of the Chinese Academy of sciences (grant no. XDB0830200, M.Z.), the open fund of state key laboratory of Pharmaceutical Biotechnology, Nanjing University, China (grant no. KF-202301, M.Z.) and Shanghai Post-doctoral Excellence Program (grant no. 2024707, J.X.).

Author information

Author notes

  1. These authors contributed equally: Jiacheng Xiong, Wei Zhang.

Authors and Affiliations

  1. Drug Discovery and Design Center, State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai, China

    Jiacheng Xiong, Wei Zhang, Yinquan Wang, Yuqi Shi, Mingyan Xu, Manjia Li, Zunyun Fu, Xiangtai Kong, Yitian Wang & Mingyue Zheng

  2. University of Chinese Academy of Sciences, Beijing, China

    Jiacheng Xiong, Wei Zhang, Yuqi Shi, Mingyan Xu, Xiangtai Kong, Yitian Wang & Mingyue Zheng

  3. Department of Medicinal Chemistry, School of Pharmacy, Fudan University, Shanghai, China

    Yinquan Wang

  4. School of Physical Science and Technology, ShanghaiTech University, Shanghai, China

    Jiatao Huang

  5. ProtonUnfold Technology Co. Ltd, Suzhou, China

    Zhaoping Xiong

Authors
  1. Jiacheng Xiong
    View author publications

    Search author on:PubMed Google Scholar

  2. Wei Zhang
    View author publications

    Search author on:PubMed Google Scholar

  3. Yinquan Wang
    View author publications

    Search author on:PubMed Google Scholar

  4. Jiatao Huang
    View author publications

    Search author on:PubMed Google Scholar

  5. Yuqi Shi
    View author publications

    Search author on:PubMed Google Scholar

  6. Mingyan Xu
    View author publications

    Search author on:PubMed Google Scholar

  7. Manjia Li
    View author publications

    Search author on:PubMed Google Scholar

  8. Zunyun Fu
    View author publications

    Search author on:PubMed Google Scholar

  9. Xiangtai Kong
    View author publications

    Search author on:PubMed Google Scholar

  10. Yitian Wang
    View author publications

    Search author on:PubMed Google Scholar

  11. Zhaoping Xiong
    View author publications

    Search author on:PubMed Google Scholar

  12. Mingyue Zheng
    View author publications

    Search author on:PubMed Google Scholar

Contributions

J.X. proposed the idea, conducted computational experiments and drafted the initial paper together with W.Z. Yinquan W., J.H., W.Z., M.X. and M.L. carried out the synthetic experiments. Y.S. developed the web application. Z.F. and J.H. contributed to the case analysis. Z.F., X.K. and M.Z. helped check and improve the paper. Yitian W. and Z.X. participated in the analysis of results. M.Z. led the project and designed the study. All authors read and approved the final paper.

Corresponding author

Correspondence to Mingyue Zheng.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Machine Intelligence thanks Jannis Born, Xiangliang Zhang and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Multistep retrosynthesis predictions by our model for (a) Vonoprazan, (b) Mitapivat, (c) Daridorexant.

The reaction centers and leaving groups are highlighted in different colors at different reaction steps.

Supplementary information

Supplementary Information

Supplementary Figs 1–26, Tables 1–12, Additional Results and Discussion, and Methods.

Reporting Summary

Source data

Source Data Fig. 3

Raw data for Fig. 3.

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

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Xiong, J., Zhang, W., Wang, Y. et al. Bridging chemistry and artificial intelligence by a reaction description language. Nat Mach Intell 7, 782–793 (2025). https://doi.org/10.1038/s42256-025-01032-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue date:

  • DOI: https://doi.org/10.1038/s42256-025-01032-8

Search

Advanced search

Quick links

Nature Briefing AI and Robotics

Sign up for the Nature Briefing: AI and Robotics newsletter — what matters in AI and robotics research, free to your inbox weekly.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing: AI and Robotics