Aligning Protein Conformation Ensemble Generation with Physical Feedback

2025/05/30 by Jiarui Lu, Lu, Jiarui, Xiaoyin Chen +11 · 1 voice · 2 citations
Biochemistry, Genetics and Molecular Biology · Computer Science · #Bioinformatics and Genomic Networks #Biomolecules (q-bio.BM) #Cell Image Analysis Techniques #FOS: Biological sciences #FOS: Computer and information sciences #Genetics, Bioinformatics, and Biomedical Research #Machine Learning (cs.LG) #cs.LG #q-bio.BM

paper · pdf · doi:10.48550/arxiv.2505.24203

openalex publication_date 2025/05/30 · arxiv published 2025/05/30 · arxiv updated 2025/05/30 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Protein dynamics play a crucial role in protein biological functions and properties, and their traditional study typically relies on time-consuming molecular dynamics (MD) simulations conducted in silico. Recent advances in generative modeling, particularly denoising diffusion models, have enabled efficient accurate protein structure prediction and conformation sampling by learning distributions over crystallographic structures. However, effectively integrating physical supervision into these data-driven approaches remains challenging, as standard energy-based objectives often lead to intractable optimization. In this paper, we introduce Energy-based Alignment (EBA), a method that aligns generative models with feedback from physical models, efficiently calibrating them to appropriately balance conformational states based on their energy differences. Experimental results on the MD ensemble benchmark demonstrate that EBA achieves state-of-the-art performance in generating high-quality protein ensembles. By improving the physical plausibility of generated structures, our approach enhances model predictions and holds promise for applications in structural biology and drug discovery.

Citations

• Scalable emulation of protein equilibrium ensembles with generative deep learning
• Structure Language Models for Protein Conformation Generation
• Generative Modeling of Molecular Dynamics Trajectories
• Aligning Target-Aware Molecule Diffusion Models with Exact Energy Optimization
• Accurate structure prediction of biomolecular interactions with AlphaFold 3
• Antigen-Specific Antibody Design via Direct Energy-based Preference Optimization
• Protein Conformation Generation via Force-Guided SE(3) Diffusion Models
• AlphaFold Meets Flow Matching for Generating Protein Ensembles
• Diffusion Model Alignment Using Direct Preference Optimization
• OpenMM 8: Molecular Dynamics Simulation with Machine Learning Potentials
• Str2Str: A Score-based Framework for Zero-shot Protein Conformation Sampling
• Direct Preference Optimization: Your Language Model is Secretly a Reward Model
• Training Diffusion Models with Reinforcement Learning
• EigenFold: Generative Protein Structure Prediction with Diffusion Models
• Two for One: Diffusion Models and Force Fields for Coarse-Grained Molecular Dynamics
• A theory of continuous generative flow networks
• Training language models to follow instructions with human feedback
• High-Resolution Image Synthesis with Latent Diffusion Models
• High-Resolution Image Synthesis with Latent Diffusion Models
• Path Integral Sampler: a stochastic control approach for sampling
• Highly accurate protein structure prediction with AlphaFold
• Flow Network based Generative Models for Non-Iterative Diverse Candidate Generation
• Denoising Diffusion Implicit Models
• Learning to summarize from human feedback
• Denoising Diffusion Probabilistic Models
• Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning
• Neural Ordinary Differential Equations
• Reliability and Learnability of Human Bandit Feedback for Sequence-to-Sequence Reinforcement Learning
• Deep Unsupervised Learning using Nonequilibrium Thermodynamics

Cited by

• FlowBack-Adjoint: Physics-Aware and Energy-Guided Conditional Flow-Matching for All-Atom Protein Backmapping
• Geometric deep learning assists protein engineering. Opportunities and Challenges

Discussions

• Jiarui Lu, Xiaoyin Chen, Stephen Zhewen Lu, Aur\'elie Lozano, Vijil Chenthamarakshan, Payel Das, Jian Tang: Aligning Protein Conformation Ensemble Generation with Physical Feedback https://arxiv.org/a [bsky, 3 points, 1 comments]

Related

• Generative modeling of protein ensembles guided by crystallographic electron densities
• AlphaFold Meets Flow Matching for Generating Protein Ensembles
• Fusing Neural and Physical: Augment Protein Conformation Sampling with Tractable Simulations
• EPO: Diverse and Realistic Protein Ensemble Generation via Energy Preference Optimization
• Scalable emulation of protein equilibrium ensembles with generative deep learning