Enhancing Ligand Validity and Affinity in Structure-Based Drug Design with Multi-Reward Optimization

Seungbeom Lee, Munsun Jo, Jungseul Ok, Dongwoo Kim
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:33129-33142, 2025.

Abstract

Deep learning-based Structure-based drug design aims to generate ligand molecules with desirable properties for protein targets. While existing models have demonstrated competitive performance in generating ligand molecules, they primarily focus on learning the chemical distribution of training datasets, often lacking effective steerability to ensure the desired chemical quality of generated molecules. To address this issue, we propose a multi-reward optimization framework that fine-tunes generative models for attributes, such as binding affinity, validity, and drug-likeness, together. Specifically, we derive direct preference optimization for a Bayesian flow network, used as a backbone for molecule generation, and integrate a reward normalization scheme to adopt multiple objectives. Experimental results show that our method generates more realistic ligands than baseline models while achieving higher binding affinity, expanding the Pareto front empirically observed in previous studies.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-lee25k, title = {Enhancing Ligand Validity and Affinity in Structure-Based Drug Design with Multi-Reward Optimization}, author = {Lee, Seungbeom and Jo, Munsun and Ok, Jungseul and Kim, Dongwoo}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {33129--33142}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/lee25k/lee25k.pdf}, url = {https://proceedings.mlr.press/v267/lee25k.html}, abstract = {Deep learning-based Structure-based drug design aims to generate ligand molecules with desirable properties for protein targets. While existing models have demonstrated competitive performance in generating ligand molecules, they primarily focus on learning the chemical distribution of training datasets, often lacking effective steerability to ensure the desired chemical quality of generated molecules. To address this issue, we propose a multi-reward optimization framework that fine-tunes generative models for attributes, such as binding affinity, validity, and drug-likeness, together. Specifically, we derive direct preference optimization for a Bayesian flow network, used as a backbone for molecule generation, and integrate a reward normalization scheme to adopt multiple objectives. Experimental results show that our method generates more realistic ligands than baseline models while achieving higher binding affinity, expanding the Pareto front empirically observed in previous studies.} }
Endnote
%0 Conference Paper %T Enhancing Ligand Validity and Affinity in Structure-Based Drug Design with Multi-Reward Optimization %A Seungbeom Lee %A Munsun Jo %A Jungseul Ok %A Dongwoo Kim %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-lee25k %I PMLR %P 33129--33142 %U https://proceedings.mlr.press/v267/lee25k.html %V 267 %X Deep learning-based Structure-based drug design aims to generate ligand molecules with desirable properties for protein targets. While existing models have demonstrated competitive performance in generating ligand molecules, they primarily focus on learning the chemical distribution of training datasets, often lacking effective steerability to ensure the desired chemical quality of generated molecules. To address this issue, we propose a multi-reward optimization framework that fine-tunes generative models for attributes, such as binding affinity, validity, and drug-likeness, together. Specifically, we derive direct preference optimization for a Bayesian flow network, used as a backbone for molecule generation, and integrate a reward normalization scheme to adopt multiple objectives. Experimental results show that our method generates more realistic ligands than baseline models while achieving higher binding affinity, expanding the Pareto front empirically observed in previous studies.
APA
Lee, S., Jo, M., Ok, J. & Kim, D.. (2025). Enhancing Ligand Validity and Affinity in Structure-Based Drug Design with Multi-Reward Optimization. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:33129-33142 Available from https://proceedings.mlr.press/v267/lee25k.html.

Related Material