EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

PFM: perturbed flow matching for structure-based drug design

Yankai Yu iD, Guikun Xu, Zhuyang Xie iD, Yan Yang iD, Yongquan Jiang, Xiaobo Zhou, Kang Li

Bioinformatics · 10.11.2026

DOI: 10.1093/bioinformatics/btag811 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Bioinformatics, online erschienen am 10.11.2026. Verfasst von Yankai Yu und 6 weiteren Autor:innen. Beteiligte Einrichtung: School of Computing and Artificial Intelligence, Southwest Jiaotong University, Chengdu, Sichuan 611756, China. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Motivation Target-specific 3D molecular generation is a promising approach for structure-based drug design. However, existing diffusion-based methods are generally restricted to marginal distributions with closed-form solutions, limiting the flexibility of generative process design, especially for the forward noising process. As a result, these methods may be less effective in modeling the complex distributions underlying pocket-conditioned molecular generation. Results We develop Perturbed Flow Matching (PFM), a flow-matching-based framework for protein pocket–conditioned 3D molecular generation. PFM employs a perturbed conditional probability path that incorporates both pocket binding-site information and coupled atom type–coordinate information, enabling more effective generation of molecules in protein binding environments. On CrossDocked2020, PFM achieves competitive structural quality and the best average Vina score among the evaluated unguided methods (-7.12); PFM-G further improves the score to -8.70 with tailored classifier-free guidance. Experiments on additional molecular datasets further confirm the effectiveness and generalizability of the proposed framework. Availability and implementation https://github.com/kurisu92725/PFM

Autor:innen und Institutionen

  • School of Computing and Artificial Intelligence, Southwest Jiaotong University, Chengdu, Sichuan 611756, China.
  • School of Artificial Intelligence, Shanghai Jiao Tong University, Shanghai 200030, China.
  • College of Information Engineering, Sichuan Agricultural University, Ya'an, Sichuan 625014, China.
  • Center for Computational Systems Medicine, McWilliams School of Biomedical Informatics, The University of Texas Health Science Center at Houston, Houston, TX 77030, USA.
  • West China Biomedical Big Data Center, West China Hospital, Sichuan University, Chengdu, Sichuan 610041, China.
  • Med-X Center for Informatics, Sichuan University, Chengdu, Sichuan 610041, China.

ORCID-Profile: Yankai Yu, Zhuyang Xie, Yan Yang

Zitieren

APA

Yu, Y., Xu, G., Xie, Z., Yang, Y., Jiang, Y., Zhou, X., & Li, K. (2026). PFM: perturbed flow matching for structure-based drug design. Bioinformatics. https://doi.org/10.1093/bioinformatics/btag811

Vancouver

Yu Y, Xu G, Xie Z, Yang Y, Jiang Y, Zhou X, et al. PFM: perturbed flow matching for structure-based drug design. Bioinformatics. 2026. doi:10.1093/bioinformatics/btag811

BibTeX
@article{Yu2026,
  author = {Yu, Yankai and Xu, Guikun and Xie, Zhuyang and Yang, Yan and Jiang, Yongquan and Zhou, Xiaobo and Li, Kang},
  title = {PFM: perturbed flow matching for structure-based drug design},
  journal = {Bioinformatics},
  year = {2026},
  doi = {10.1093/bioinformatics/btag811}
}
Kostenlos starten →