Let Physics Guide Your Protein Flows: Topology-aware Unfolding and Generation

2025-09-29

AI Summary

Existing diffusion models for protein structure generation often produce physically implausible conformations due to physically ungrounded noise schedules. Method: We propose a physics-guided nonlinear noising scheme that embeds classical mechanical constraints—such as bond lengths, bond angles, and secondary-structure continuity—into an SE(3)-equivariant flow-matching framework, enabling the first physically driven, topology-preserving protein unfolding and generation. Our method jointly encodes sequence information and backbone geometry, supporting high-fidelity, SE(3)-equivariant 3D conformation generation conditioned on amino acid sequences. Results: In unconditional generation, our approach achieves state-of-the-art performance, significantly improving structural diversity, physical validity, and designability. Moreover, it accurately folds monomeric sequences into native-like conformations. This establishes a new paradigm for programmable protein design grounded in physical principles.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Generation

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies

Abstract

Protein structure prediction and folding are fundamental to understanding biology, with recent deep learning advances reshaping the field. Diffusion-based generative models have revolutionized protein design, enabling the creation of novel proteins. However, these methods often neglect the intrinsic physical realism of proteins, driven by noising dynamics that lack grounding in physical principles. To address this, we first introduce a physically motivated non-linear noising process, grounded in classical physics, that unfolds proteins into secondary structures (e.g., alpha helices, linear beta sheets) while preserving topological integrity--maintaining bonds, and preventing collisions. We then integrate this process with the flow-matching paradigm on SE(3) to model the invariant distribution of protein backbones with high fidelity, incorporating sequence information to enable sequence-conditioned folding and expand the generative capabilities of our model. Experimental results demonstrate that the proposed method achieves state-of-the-art performance in unconditional protein generation, producing more designable and novel protein structures while accurately folding monomer sequences into precise protein conformations.

Problem

• Physically realistic protein unfolding preserving topological integrity
• Integrating physics-based flows with SE(3) for backbone distribution modeling
• Enhancing designability and accuracy in protein generation and folding

Innovation

• Physics-based nonlinear noising for protein unfolding
• SE(3) flow-matching for backbone distribution modeling
• Sequence-conditioned folding with topological integrity preservation