Agentic End-to-End De Novo Protein Design for Tailored Dynamics Using a Language Diffusion Model

2025-02-14

AI Summary

Protein dynamics—such as normal vibrational modes—are critical for catalysis, allostery, and signal transduction; however, de novo design of proteins with targeted dynamics remains challenging due to the complex, highly degenerate sequence–structure–dynamics relationship. To address this, we introduce the first end-to-end protein generation framework conditioned on target normal modes. Our approach employs an embodied intelligent dual-model architecture—comprising a designer and a predictor—that enables bidirectional mapping between amino acid sequences and dynamic behavior. It integrates a language diffusion model, regularized modal analysis, all-atom molecular dynamics simulations, and surrogate-assisted iterative optimization. Experiments demonstrate that generated proteins exhibit backbone vibrational amplitude errors under 8%, sequence homology to natural proteins below 30%, structural stability, and functional dynamic signatures. This work establishes a paradigm shift from static-structure–centric design to dynamics-aware generative protein engineering.

Technology Category

• Natural Language Processing: Generation
• Intelligent Robots: Embodied AI
• Computer Vision: Diffusion Models for Vision

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

Proteins are dynamic molecular machines whose biological functions, spanning enzymatic catalysis, signal transduction, and structural adaptation, are intrinsically linked to their motions. Designing proteins with targeted dynamic properties, however, remains a challenge due to the complex, degenerate relationships between sequence, structure, and molecular motion. Here, we introduce VibeGen, a generative AI framework that enables end-to-end de novo protein design conditioned on normal mode vibrations. VibeGen employs an agentic dual-model architecture, comprising a protein designer that generates sequence candidates based on specified vibrational modes and a protein predictor that evaluates their dynamic accuracy. This approach synergizes diversity, accuracy, and novelty during the design process. Via full-atom molecular simulations as direct validation, we demonstrate that the designed proteins accurately reproduce the prescribed normal mode amplitudes across the backbone while adopting various stable, functionally relevant structures. Notably, generated sequences are de novo, exhibiting no significant similarity to natural proteins, thereby expanding the accessible protein space beyond evolutionary constraints. Our work integrates protein dynamics into generative protein design, and establishes a direct, bidirectional link between sequence and vibrational behavior, unlocking new pathways for engineering biomolecules with tailored dynamical and functional properties. This framework holds broad implications for the rational design of flexible enzymes, dynamic scaffolds, and biomaterials, paving the way toward dynamics-informed AI-driven protein engineering.

Problem

• Designing proteins with targeted dynamic properties.
• Integrating protein dynamics into generative protein design.
• Engineering biomolecules with tailored dynamical and functional properties.

Innovation

• Language Diffusion Model
• Agentic Dual-Model Architecture
• De Novo Protein Design