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

2025-05-12

AI Summary

To address the challenge of multimodal modeling—integrating continuous 3D conformations and discrete 2D topologies—in structure-based drug design, this paper introduces VLB-Optimal Scheduling (VOS), a novel noise scheduling paradigm. VOS is the first to theoretically characterize how noise scheduling affects the variational lower bound (VLB) of multimodal probabilistic diffusion paths, overcoming limitations of conventional unimodal scheduling. The method unifies geometric deep learning, multimodal variational inference, and path integral optimization to jointly model atomic coordinates and molecular graphs. Evaluated on the CrossDock benchmark, our approach achieves a PoseBusters pass rate of 95.9%, outperforming strong baselines by over 10%. It further maintains high-affinity prediction accuracy and ensures >99.8% intramolecular structural validity, demonstrating robustness and chemical feasibility.

Technology Category

• Search and Optimization: Mixed Discrete/Continuous Search
• Planning, Routing, and Scheduling: Planning with Language Models
• Machine Learning: Multimodal Learning

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

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.

Problem

• Challenges in geometric structure modeling for drug design
• Twisted probability path of multi-modalities in molecular geometries
• Optimizing noise schedules for better molecular interaction modeling

Innovation

• VLB-Optimal Scheduling (VOS) strategy
• Optimizes VLB as path integral
• Enhances molecular geometry modeling