RGFN: Synthesizable Molecular Generation Using GFlowNets
Abstract
In this paper, we propose an extension of the GFlowNet framework that operates directly in the space of chemical reactions, offering out-of-the-box synthesizability, while maintaining comparable quality of generated candidates. We demonstrate that with the proposed set of reactions and fragments, it is possible to obtain a search space of molecules orders of magnitude larger than existing screening libraries while offering low costs of synthesis. We also show that the approach scales to very large fragment libraries, further increasing the number of potential molecules. Our experiments showcase the effectiveness of the proposed approach across a range of oracle models.
Cite
Text
Koziarski et al. "RGFN: Synthesizable Molecular Generation Using GFlowNets." ICML 2024 Workshops: ML4LMS, 2024.Markdown
[Koziarski et al. "RGFN: Synthesizable Molecular Generation Using GFlowNets." ICML 2024 Workshops: ML4LMS, 2024.](https://mlanthology.org/icmlw/2024/koziarski2024icmlw-rgfn/)BibTeX
@inproceedings{koziarski2024icmlw-rgfn,
title = {{RGFN: Synthesizable Molecular Generation Using GFlowNets}},
author = {Koziarski, Michał and Rekesh, Andrei and Shevchuk, Dmytro and van der Sloot, Almer M. and Gaiński, Piotr and Bengio, Yoshua and Liu, Cheng-Hao and Tyers, Mike and Batey, Robert A.},
booktitle = {ICML 2024 Workshops: ML4LMS},
year = {2024},
url = {https://mlanthology.org/icmlw/2024/koziarski2024icmlw-rgfn/}
}