EquiBoost: An Equivariant Boosting Approach to Molecular Conformation Generation

2025-01-09

AI Summary

Addressing the longstanding trade-off among accuracy, diversity, and efficiency in 3D molecular conformation generation, this work proposes the first non-diffusion, boosting-based iterative optimization framework. The method employs stacked equivariant graph transformers as weak learners, integrating 3D geometry-aware message passing with RMSD-driven per-iteration residual correction to progressively refine conformations. Crucially, it pioneers the application of ensemble learning—specifically boosting—to conformation generation, eliminating computationally expensive diffusion processes. On the GEOM-QM9 benchmark, our approach achieves a 18.3% improvement in the AMR (Average Minimum RMSD) metric over current state-of-the-art diffusion models, while simultaneously attaining higher conformational diversity and a 3.2× speedup in inference time. This demonstrates that boosting-based iterative refinement offers a highly efficient and accurate alternative to diffusion-based paradigms in 3D molecular structure prediction.

Technology Category

• Machine Learning: Structured Learning
• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Molecular conformation generation plays key roles in computational drug design. Recently developed deep learning methods, particularly diffusion models have reached competitive performance over traditional cheminformatical approaches. However, these methods are often time-consuming or require extra support from traditional methods. We propose EquiBoost, a boosting model that stacks several equivariant graph transformers as weak learners, to iteratively refine 3D conformations of molecules. Without relying on diffusion techniques, EquiBoost balances accuracy and efficiency more effectively than diffusion-based methods. Notably, compared to the previous state-of-the-art diffusion method, EquiBoost improves generation quality and preserves diversity, achieving considerably better precision of Average Minimum RMSD (AMR) on the GEOM datasets. This work rejuvenates boosting and sheds light on its potential to be a robust alternative to diffusion models in certain scenarios.

Problem

• Molecular Shape Generation
• Drug Design
• Computational Efficiency

Innovation

• EquiBoost
• Non-Diffusion Model
• Graph Transformer