GDiffRetro: Retrosynthesis Prediction with Dual Graph Enhanced Molecular Representation and Diffusion Generation

2025-01-14

AI Summary

Existing retrosynthetic prediction methods neglect molecular 3D geometry, leading to inaccurate reaction center identification and 3D-implausible reactant generation. To address this, we propose a two-stage framework. In the first stage, we design a dual-graph molecular representation—integrating primal and dual graphs—augmented with molecular face encoding to precisely localize reaction centers. In the second stage, we introduce a novel conditional 3D diffusion model for synthon-to-reactant generation, incorporating geometry-aware denoising under joint topological and spatial constraints to produce chemically valid, full-atom reactants. Evaluated on multiple benchmarks, our method significantly outperforms semi-template-based approaches: it achieves a 4.2% absolute improvement in Top-1 accuracy and establishes new state-of-the-art performance in both 3D conformational validity and reaction feasibility.

Technology Category

• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

Retrosynthesis prediction focuses on identifying reactants capable of synthesizing a target product. Typically, the retrosynthesis prediction involves two phases: Reaction Center Identification and Reactant Generation. However, we argue that most existing methods suffer from two limitations in the two phases: (i) Existing models do not adequately capture the ``face'' information in molecular graphs for the reaction center identification. (ii) Current approaches for the reactant generation predominantly use sequence generation in a 2D space, which lacks versatility in generating reasonable distributions for completed reactive groups and overlooks molecules' inherent 3D properties. To overcome the above limitations, we propose GDiffRetro. For the reaction center identification, GDiffRetro uniquely integrates the original graph with its corresponding dual graph to represent molecular structures, which helps guide the model to focus more on the faces in the graph. For the reactant generation, GDiffRetro employs a conditional diffusion model in 3D to further transform the obtained synthon into a complete reactant. Our experimental findings reveal that GDiffRetro outperforms state-of-the-art semi-template models across various evaluative metrics.

Problem

• retrosynthetic prediction
• molecular shape
• 3D molecular properties

Innovation

• 3D Reaction Prediction
• Diffusion Models
• Retrosynthetic Analysis