Transferable Boltzmann Generators

2024-06-20

arXiv.org

AI Summary

Addressing the challenges of equilibrium sampling in molecular systems and poor generalization across chemical space, this paper introduces the first transferable Boltzmann generator framework. Built upon a flow-matching-based normalizing flow architecture, the method jointly models molecular graph representations and coordinate-free geometric invariances, enabling zero-shot prediction of Boltzmann distributions and efficient approximate sampling for unseen molecules. Compared to molecule-specific models, it significantly improves sampling efficiency and importance reweighting accuracy. Experiments on dipeptide systems demonstrate that the model achieves high-fidelity conformational sampling and accurate energy distribution modeling—without any training data for the target molecule—thereby validating its strong cross-molecular generalization capability. This work establishes a novel paradigm for designing universal molecular samplers in computational chemistry, advancing the goal of scalable, data-efficient, and physically grounded generative modeling of molecular ensembles.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Search and Optimization: Sampling/Simulation-based Search
• Natural Language Processing: Generation

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications

Abstract

The generation of equilibrium samples of molecular systems has been a long-standing problem in statistical physics. Boltzmann Generators are a generative machine learning method that addresses this issue by learning a transformation via a normalizing flow from a simple prior distribution to the target Boltzmann distribution of interest. Recently, flow matching has been employed to train Boltzmann Generators for small molecular systems in Cartesian coordinates. We extend this work and propose a first framework for Boltzmann Generators that are transferable across chemical space, such that they predict zero-shot Boltzmann distributions for test molecules without being retrained for these systems. These transferable Boltzmann Generators allow approximate sampling from the target distribution of unseen systems, as well as efficient reweighting to the target Boltzmann distribution. The transferability of the proposed framework is evaluated on dipeptides, where we show that it generalizes efficiently to unseen systems. Furthermore, we demonstrate that our proposed architecture enhances the efficiency of Boltzmann Generators trained on single molecular systems.

Problem

• Statistical Physics
• Molecular Sampling
• Chemical Environments

Innovation

• Boltzmann Generator
• Chemical Environment Adaptability
• Efficient Sampling