Enhancing Diffusion-Based Sampling with Molecular Collective Variables

2025-10-13

AI Summary

Diffusion models suffer from low sampling efficiency and inadequate coverage of thermodynamic basins in molecular conformational sampling. To address this, we propose the first diffusion-based framework integrating collective variable (CV)-guided enhanced sampling: reaction coordinates are defined via CV projection; a serialized repulsive potential drives barrier-crossing exploration; and importance reweighting enables energy-driven, training-data-free sampling. Our method supports reactive conformation generation (e.g., bond cleavage/formation) and accurate free energy difference estimation. On peptide benchmarks, it faithfully reproduces polymorphic conformational distributions and high-fidelity free energy profiles, achieving sampling efficiency markedly superior to conventional diffusion and molecular dynamics methods—approaching first-principles accuracy.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Learning with Manifolds

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

Diffusion-based samplers learn to sample complex, high-dimensional distributions using energies or log densities alone, without training data. Yet, they remain impractical for molecular sampling because they are often slower than molecular dynamics and miss thermodynamically relevant modes. Inspired by enhanced sampling, we encourage exploration by introducing a sequential bias along bespoke, information-rich, low-dimensional projections of atomic coordinates known as collective variables (CVs). We introduce a repulsive potential centered on the CVs from recent samples, which pushes future samples towards novel CV regions and effectively increases the temperature in the projected space. Our resulting method improves efficiency, mode discovery, enables the estimation of free energy differences, and retains independent sampling from the approximate Boltzmann distribution via reweighting by the bias. On standard peptide conformational sampling benchmarks, the method recovers diverse conformational states and accurate free energy profiles. We are the first to demonstrate reactive sampling using a diffusion-based sampler, capturing bond breaking and formation with universal interatomic potentials at near-first-principles accuracy. The approach resolves reactive energy landscapes at a fraction of the wall-clock time of standard sampling methods, advancing diffusion-based sampling towards practical use in molecular sciences.

Problem

• Improving molecular sampling efficiency using collective variables
• Enhancing exploration of thermodynamically relevant molecular configurations
• Enabling reactive sampling with bond breaking and formation

Innovation

• Introducing sequential bias along collective variables
• Using repulsive potential to explore novel CV regions
• Enabling reactive sampling with diffusion-based methods