Piloting Structure-Based Drug Design via Modality-Specific Optimal Schedule

Keyue Qiu, Yuxuan Song, Zhehuan Fan, Peidong Liu, Zhe Zhang, Mingyue Zheng, Hao Zhou, Wei-Ying Ma
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:50619-50644, 2025.

Abstract

Structure-Based Drug Design (SBDD) is crucial for identifying bioactive molecules. Recent deep generative models are faced with challenges in geometric structure modeling. A major bottleneck lies in the twisted probability path of multi-modalities—continuous 3D positions and discrete 2D topologies—which jointly determine molecular geometries. By establishing the fact that noise schedules decide the Variational Lower Bound (VLB) for the twisted probability path, we propose VLB-Optimal Scheduling (VOS) strategy in this under-explored area, which optimizes VLB as a path integral for SBDD. Our model effectively enhances molecular geometries and interaction modeling, achieving state-of-the-art PoseBusters passing rate of 95.9% on CrossDock, more than 10% improvement upon strong baselines, while maintaining high affinities and robust intramolecular validity evaluated on held-out test set. Code is available at https://github.com/AlgoMole/MolCRAFT.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-qiu25g, title = {Piloting Structure-Based Drug Design via Modality-Specific Optimal Schedule}, author = {Qiu, Keyue and Song, Yuxuan and Fan, Zhehuan and Liu, Peidong and Zhang, Zhe and Zheng, Mingyue and Zhou, Hao and Ma, Wei-Ying}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {50619--50644}, 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/qiu25g/qiu25g.pdf}, url = {https://proceedings.mlr.press/v267/qiu25g.html}, abstract = {Structure-Based Drug Design (SBDD) is crucial for identifying bioactive molecules. Recent deep generative models are faced with challenges in geometric structure modeling. A major bottleneck lies in the twisted probability path of multi-modalities—continuous 3D positions and discrete 2D topologies—which jointly determine molecular geometries. By establishing the fact that noise schedules decide the Variational Lower Bound (VLB) for the twisted probability path, we propose VLB-Optimal Scheduling (VOS) strategy in this under-explored area, which optimizes VLB as a path integral for SBDD. Our model effectively enhances molecular geometries and interaction modeling, achieving state-of-the-art PoseBusters passing rate of 95.9% on CrossDock, more than 10% improvement upon strong baselines, while maintaining high affinities and robust intramolecular validity evaluated on held-out test set. Code is available at https://github.com/AlgoMole/MolCRAFT.} }
Endnote
%0 Conference Paper %T Piloting Structure-Based Drug Design via Modality-Specific Optimal Schedule %A Keyue Qiu %A Yuxuan Song %A Zhehuan Fan %A Peidong Liu %A Zhe Zhang %A Mingyue Zheng %A Hao Zhou %A Wei-Ying Ma %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-qiu25g %I PMLR %P 50619--50644 %U https://proceedings.mlr.press/v267/qiu25g.html %V 267 %X Structure-Based Drug Design (SBDD) is crucial for identifying bioactive molecules. Recent deep generative models are faced with challenges in geometric structure modeling. A major bottleneck lies in the twisted probability path of multi-modalities—continuous 3D positions and discrete 2D topologies—which jointly determine molecular geometries. By establishing the fact that noise schedules decide the Variational Lower Bound (VLB) for the twisted probability path, we propose VLB-Optimal Scheduling (VOS) strategy in this under-explored area, which optimizes VLB as a path integral for SBDD. Our model effectively enhances molecular geometries and interaction modeling, achieving state-of-the-art PoseBusters passing rate of 95.9% on CrossDock, more than 10% improvement upon strong baselines, while maintaining high affinities and robust intramolecular validity evaluated on held-out test set. Code is available at https://github.com/AlgoMole/MolCRAFT.
APA
Qiu, K., Song, Y., Fan, Z., Liu, P., Zhang, Z., Zheng, M., Zhou, H. & Ma, W.. (2025). Piloting Structure-Based Drug Design via Modality-Specific Optimal Schedule. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:50619-50644 Available from https://proceedings.mlr.press/v267/qiu25g.html.

Related Material