EUVIMEDEUVIMED entdecken
Crossref

CRAFT: A Chemical Reaction-Aware Transformer for Synthesizable Molecular Optimization via Multimodal Learning

Runfu Yu, Tongmao Ma iD, Zian Song, Qiyao Yin, Shuang Wang iD

Journal of Medicinal Chemistry · 08.10.2026 · 69 (19) · 21919-21936

DOI: 10.1021/acs.jmedchem.6c00920 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Medicinal Chemistry, online erschienen am 23.09.2026. Verfasst von Runfu Yu und 4 weiteren Autor:innen. Beteiligte Einrichtung: College of Computer Science and Technology, China University of Petroleum, Qingdao266580, China. Hauptthemen laut MEDLINE: Drug Design. Die Arbeit nennt 53 Literaturangaben. Gefördert von: Youth Innovation Technology Project of Higher School in Shandong Province, Natural Science Foundation of Shandong Province, Young Talent of Lifting engineering for Science and Technology in Shandong.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Ensuring synthetic accessibility remains a major challenge in de novo drug design, as generative models can produce high-scoring molecules that are difficult to synthesize. To address this issue, we propose CRAFT, a reaction-aware molecular optimization framework that integrates Monte Carlo Tree Search (MCTS) with a multimodal strategy network. Unlike sequence-only baselines, CRAFT explicitly encodes two-dimensional molecular topology and atom-level reaction roles, enabling chemically feasible transformations during multi-step optimization. Comprehensive evaluations on DRD2, AKT1, and CXCR4 demonstrate that CRAFT consistently outperforms state-of-the-art methods, generating more high-scoring candidates while improving reaction-constrained synthetic accessibility and molecular diversity. CRAFT also rediscovers scaffolds homologous to known active ligands and identifies novel chemical entities beyond the training distribution, as supported by molecular docking simulations. Moreover, the model provides transparent forward-synthesis pathways for generated candidates, helping bridge in silico molecular design and practical synthesis.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Drug DesignMonte Carlo MethodMolecular Docking SimulationLigandsReceptors, CXCR4Receptors, Dopamine D2Proto-Oncogene Proteins c-akt

Stichwörter der Autor:innen

Monte Carlo Tree searchmolecular optimizationreaction-based generationsynthetic accessibility

Substanzen

LigandsReceptors, CXCR4Receptors, Dopamine D2Proto-Oncogene Proteins c-akt

Autor:innen und Institutionen

  • College of Computer Science and Technology, China University of Petroleum, Qingdao266580, China.
  • Shandong Key Laboratory of Intelligent Oil & Gas Industrial Software, Qingdao266580, China.

ORCID-Profile: Tongmao Ma, Shuang Wang

Förderung

  • Youth Innovation Technology Project of Higher School in Shandong Province
  • Natural Science Foundation of Shandong Province
  • Young Talent of Lifting engineering for Science and Technology in Shandong
  • National Natural Science Foundation of China

Zitieren

APA

Yu, R., Ma, T., Song, Z., Yin, Q., & Wang, S. (2026). CRAFT: A Chemical Reaction-Aware Transformer for Synthesizable Molecular Optimization via Multimodal Learning. Journal of Medicinal Chemistry, 69(19), 21919-21936. https://doi.org/10.1021/acs.jmedchem.6c00920

Vancouver

Yu R, Ma T, Song Z, Yin Q, Wang S. CRAFT: A Chemical Reaction-Aware Transformer for Synthesizable Molecular Optimization via Multimodal Learning. Journal of Medicinal Chemistry. 2026;69(19):21919-21936. doi:10.1021/acs.jmedchem.6c00920

BibTeX
@article{Yu2026,
  author = {Yu, Runfu and Ma, Tongmao and Song, Zian and Yin, Qiyao and Wang, Shuang},
  title = {CRAFT: A Chemical Reaction-Aware Transformer for Synthesizable Molecular Optimization via Multimodal Learning},
  journal = {Journal of Medicinal Chemistry},
  year = {2026},
  volume = {69},
  number = {19},
  pages = {21919--21936},
  doi = {10.1021/acs.jmedchem.6c00920}
}
Kostenlos starten →