Skip to content
DraconiticResearch
Search Sources API
194,230 studies31 sourcesUpdated 55 min ago
Peer-reviewedCC BYNature Communications

Automated transition state generation for mechanistic exploration in organic synthesis

  • Li‐Cheng Xu1,2
  • Junyi An3,4
  • Weiqi Liu5,4
  • Yunfei Shi3,4
  • Chong‐Lei Ji6
  • Fenglei Cao7,8
  • Yuan Qi9,10
10 affiliations
  1. 1Golab Inc, Shanghai, China. [email protected]
  2. 2Shanghai Academy of Artificial Intelligence for Science, Shanghai, China. [email protected]
  3. 3Golab Inc, Shanghai, China
  4. 4Shanghai Academy of Artificial Intelligence for Science, Shanghai, China
  5. 5Artificial Intelligence Innovation and Incubation Institute, Fudan University, Shanghai, China
  6. 6School of Physical Science and Technology, ShanghaiTech University, Shanghai, China
  7. 7Golab Inc, Shanghai, China. [email protected]
  8. 8Shanghai Academy of Artificial Intelligence for Science, Shanghai, China. [email protected]
  9. 9Artificial Intelligence Innovation and Incubation Institute, Fudan University, Shanghai, China. [email protected]
  10. 10Shanghai Academy of Artificial Intelligence for Science, Shanghai, China. [email protected]
Open at sourcePDFDOICite

Abstract

Accurate generation of intricate 3D molecular structures is a fundamental challenge in computational chemistry. Transition states (TS), the transient structures that dictate reaction kinetics, exemplify this challenge, as their calculation remains a major bottleneck in elucidating reaction mechanisms. While generative AI has shown promise in automating TS generation, existing methods are largely confined to simple systems, struggling with the complex structures like those in transition metal catalysis. Here we show a unified framework for general-purpose TS generation, comprising the UniTS-Lib library of 4,391 high-quality structures spanning 42 elements and diverse chemical transformations, coupled with the UniTS-Gen diffusion model that generates 3D TS configurations from 2D reactant graphs using a custom-designed higher-degree equivariant network. Validation demonstrates UniTS-Gen's superior accuracy and robust generalization to unseen chemical systems. We show that UniTS-Gen provides reliable initial guesses and accelerates discovery by locating kinetically favored conformations. This work provides a scalable and transferable solution for automating mechanistic studies in organic synthesis and beyond.

Related

  1. Commit: Automated Scale-Up Crystallisation DataFactory for model-based pharmaceutical process development: a data-driven, mechanistic and hybrid model case study
    Thomas Pickles, Cameron J. Brown
    ChemRxivPreprintCC BYFull textMaterials & polymersChemRxiv
  2. Molecular representation shapes the balance between target fidelity and exploration in flow based polymer generation
    Tianren Zhang
    arXivPreprintarXiv licenceFull textPhysical chemistryCondensed matterarXiv
  3. Exposure States, Environmental Fate, and Functional Effects of Microplastics in Soil: Bibliometric Mapping and Critical Mechanistic Synthesis
    Yuhe Li, Su Yan, Xinting Lv +4
    PolymersPeer-reviewedCC BYFull textMaterials & polymersEurope PMC
  4. Silver Nanoparticle Synthesis in a Plant-Derived Gel and Exploration of its Redox and Catalytic Properties
    Chaiti Malla, Chitraniva Datta, Manish Kumar +2
    ChemRxivPreprintCC BYFull textMaterials & polymersChemRxiv
  5. Extension of the Pnictinous Acid Series: Rational Synthesis of a Stibinous Acid and Exploration of Its Reactivity.
    Bi Youan E Tra, Addis Getahun, Timothy C Johnstone
    Inorganic chemistryPeer-reviewedCC BY-NC-NDFull textInorganic & organometallic chemistryEurope PMC

Cite

BibTeXRIS
BibTeX
@article{xu2026automated,
  title = {{Automated transition state generation for mechanistic exploration in organic synthesis}},
  author = {Li‐Cheng Xu and Junyi An and Weiqi Liu and Yunfei Shi and Chong‐Lei Ji and Fenglei Cao and Yuan Qi},
  journal = {Nature Communications},
  year = {2026},
  month = aug,
  doi = {10.1038/s41467-026-77230-8},
  url = {https://doi.org/10.1038/s41467-026-77230-8},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Automated transition state generation for mechanistic exploration in organic synthesis
AU  - Li‐Cheng Xu
AU  - Junyi An
AU  - Weiqi Liu
AU  - Yunfei Shi
AU  - Chong‐Lei Ji
AU  - Fenglei Cao
AU  - Yuan Qi
PY  - 2026
DA  - 2026/08/27
JO  - Nature Communications
DO  - 10.1038/s41467-026-77230-8
UR  - https://doi.org/10.1038/s41467-026-77230-8
AB  - Accurate generation of intricate 3D molecular structures is a fundamental challenge in computational chemistry. Transition states (TS), the transient structures that dictate reaction kinetics, exemplify this challenge, as their calculation remains a major bottleneck in elucidating reaction mechanisms. While generative AI has shown promise in automating TS generation, existing methods are largely confined to simple systems, struggling with the complex structures like those in transition metal catalysis. Here we show a unified framework for general-purpose TS generation, comprising the UniTS-Lib library of 4,391 high-quality structures spanning 42 elements and diverse chemical transformations, coupled with the UniTS-Gen diffusion model that generates 3D TS configurations from 2D reactant graphs using a custom-designed higher-degree equivariant network. Validation demonstrates UniTS-Gen's superior accuracy and robust generalization to unseen chemical systems. We show that UniTS-Gen provides reliable initial guesses and accelerates discovery by locating kinetically favored conformations. This work provides a scalable and transferable solution for automating mechanistic studies in organic synthesis and beyond.
SN  - 2041-1723
ER  -