HybridLinker: Topology-Guided Posterior Sampling for Enhanced Diversity and Validity in 3D Molecular Linker Generation

2025-02-24

AI Summary

Addressing the challenge of balancing diversity and validity in 3D molecular linker generation, this paper proposes LinkerDPS—a retraining-free hybrid inference framework. LinkerDPS is the first method to synergistically integrate PC-Free (topologically diverse) and PC-Aware (structurally valid) models: it leverages a PC-Free model to generate diverse molecular scaffolds, then employs energy-function-guided diffusion posterior sampling (DPS) to map topological structures into high-quality 3D point clouds. This cross-space distribution transfer paradigm unifies molecular graph neural networks with physics-informed constraints. On multiple benchmarks, LinkerDPS improves validity by 12.7% and diversity by 28.3%, while simultaneously enhancing drug-likeness and target-binding affinity. Its practical utility is validated on real-world tasks, including PROTAC design.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Mixed Discrete/Continuous Search

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: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Linker generation is critical in drug discovery applications such as lead optimization and PROTAC design, where molecular fragments are assembled into diverse drug candidates. Existing methods fall into PC-Free and PC-Aware categories based on their use of 3D point clouds (PC). PC-Free models prioritize diversity but suffer from lower validity due to overlooking PC constraints, while PC-Aware models ensure higher validity but restrict diversity by enforcing strict PC constraints. To overcome these trade-offs without additional training, we propose HybridLinker, a framework that enhances PC-Aware inference by providing diverse bonding topologies from a pretrained PC-Free model as guidance. At its core, we propose LinkerDPS, the first diffusion posterior sampling (DPS) method operating across PC-Free and PC-Aware spaces, bridging molecular topology with 3D point clouds via an energy-inspired function. By transferring the diverse sampling distribution of PC-Free models into the PC-Aware distribution, HybridLinker significantly and consistently surpasses baselines, improving both validity and diversity in foundational molecular design and applied property optimization tasks, establishing a new DPS framework in the molecular and graph domains beyond imaging.

Problem

• Enhancing 3D molecular linker diversity
• Improving molecular linker validity
• Bridging molecular topology with 3D point clouds

Innovation

• HybridLinker enhances PC-Aware inference
• LinkerDPS bridges molecular topology with 3D point clouds
• HybridLinker improves validity and diversity