AI‐Physics‐Experiment Trinity for Integrated Protein Dynamics Modeling
2026/06/16 by S T Chen, Minying Low, Peng Xiu +2 · 1 voice
Biochemistry, Genetics and Molecular Biology · Materials Science · #Protein Structure and Dynamics #Machine Learning in Materials Science #Enzyme Structure and Function
paper · doi:10.1002/advs.76023
openalex publication_date 2026/06/16 · openalex created_date 2026/06/17 · openalex updated_date 2026/07/25
Abstract
Proteins exist as conformational ensembles, with dynamic transitions governing biological processes. Deciphering these dynamics demands integrating experimental data, physics-based simulations, and artificial intelligence (AI)-each with distinct strengths and limitations. Experiments deliver direct structural and dynamic benchmarks but are often constrained by insufficient spatiotemporal resolution or difficulties with providing information on transiently and weakly populated states. Physics-based methods may generate atomic-scale trajectories via force fields yet face sampling bottlenecks, force field sensitivity, and the curse of dimensionality. AI, particularly deep learning and generative modeling approaches, facilitates the efficient prediction of protein structures and conformational ensembles, as well as dimensionality reduction, yet is hindered by limited interpretability and transferability, and a scarcity of high-quality ground-truth data for training and benchmarking models of dynamics. This review outlines core principles of standalone approaches and highlights integrative strategies: experimental constraints guide physics-driven refinement; AI enhances experimental processing and ensemble generation; physics imparts plausibility to AI, while AI accelerates simulation sampling and force field optimization. We elaborate on this synergy, emphasizing physics-based modeling's glue-like role in reconciling heterogeneous datasets. Finally, we summarize persistent challenges and discuss future directions for integrated modeling of protein dynamics.
Citations
- Experimental Parameterization of an Energy Function for the Simulation of Unfolded Proteins
- T<scp>RANSITION</scp>P<scp>ATH</scp>S<scp>AMPLING</scp>: Throwing Ropes Over Rough Mountain Passes, in the Dark
- PLUMED 2: New feathers for an old bird
- Transition-Path Theory and Path-Finding Algorithms for the Study of Rare Events
- THEORY OF PROTEIN FOLDING: The Energy Landscape Perspective
- Relation between native ensembles and experimental structures of proteins
- Principles of protein structural ensemble determination
- Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
- Coarse graining molecular dynamics with graph neural networks
- LAMMPS - a flexible simulation tool for particle-based materials modeling at the atomic, meso, and continuum scales
- КОМПЬЮТЕРНАЯ ТОМОГРАФИЯ – ОСНОВА ОПТИМИЗАЦИИ МИНИТОРАКОТОМНОГО ДОСТУПА
- Robust deep learning–based protein sequence design using ProteinMPNN
- Evolutionary-scale prediction of atomic-level protein structure with a language model
- The Open Force Field Initiative: Open Software and Open Science for Molecular Modeling
- A coarse‐grained model for disordered and multi‐domain proteins
- Structure-Based Experimental Datasets for Benchmarking Protein Simulation Force Fields [Article v1.0]
- AF-CALVADOS: AlphaFold-guided simulations of multi-domain proteins at the proteome level
- Extending Conformational Ensemble Prediction to Multidomain Proteins and Protein Complex
- Unlocking hidden biomolecular conformational landscapes in diffusion models at inference time
- Breaking the Timescale Barrier: Generative Discovery of Conformational Free-Energy Landscapes and Transition Pathways
- ODesign: A World Model for Biomolecular Interaction Design
- HollowFlow: Efficient Sample Likelihood Evaluation using Hollow Message Passing
- CryoDyna: Multiscale end-to-end modeling of cryo-EM macromolecule dynamics with physics-aware neural network
- Transferable Generative Models Bridge Femtosecond to Nanosecond Time-Step Molecular Dynamics
- MarS-FM: Generative Modeling of Molecular Dynamics via Markov State Models
- AI-based Methods for Simulating, Sampling, and Predicting Protein Ensembles
- Accelerating Protein Molecular Dynamics Simulation with DeepJump
- Enhanced Sampling in the Age of Machine Learning: Algorithms and Applications
- Beyond Ensembles: Simulating All-Atom Protein Dynamics in a Learned Latent Space
- BioMD: All-atom Generative Model for Biomolecular Dynamics Simulation
- Molecular Simulations with a Pretrained Neural Network and Universal Pairwise Force Fields
- Following the Committor Flow: A Data-Driven Discovery of Transition Pathways
- La-Proteina: Atomistic Protein Generation via Partially Latent Flow Matching
- Generation of protein dynamics by machine learning
- A Scalable and Quantum-Accurate Foundation Model for Biomolecular Force Field via Linearly Tensorized Quadrangle Attention
- Generative Modeling of Full-Atom Protein Conformations using Latent Diffusion on Graph Embeddings
- Consistent Sampling and Simulation: Molecular Dynamics with Energy-Based Diffusion Models
- Multiscale guidance of protein structure prediction with heterogeneous cryo-EM data
- ConfRover: Simultaneous Modeling of Protein Conformation and Dynamics via Autoregression
- UniSim: A Unified Simulator for Time-Coarsened Dynamics of Biomolecules
- Software package for simulations using the coarse-grained CALVADOS model
- Scalable emulation of protein equilibrium ensembles with generative deep learning
- P2DFlow: A Protein Ensemble Generative Model with SE(3) Flow Matching
- EquiJump: Protein Dynamics Simulation via SO(3)-Equivariant Stochastic Interpolants
- Generative Modeling of Molecular Dynamics Trajectories
- Reaction Coordinates are Optimal Channels of Energy Flow
- Transferable Boltzmann Generators
- Towards a General Time Series Forecasting Model with Unified Representation and Adaptive Transfer
- Accurate structure prediction of biomolecular interactions with AlphaFold 3
- F3low: Frame-to-Frame Coarse-grained Molecular Dynamics with SE(3) Guided Flow Matching
- Protein Conformation Generation via Force-Guided SE(3) Diffusion Models
- Computing the Committor with the Committor: an Anatomy of the Transition State Ensemble
- Navigating protein landscapes with a machine-learned transferable coarse-grained model
- Gaussian Approximation Potentials: theory, software implementation and application examples
- DiAMoNDBack: Diffusion-denoising Autoregressive Model for Non-Deterministic Backmapping of Cα Protein Traces
- De novo design of protein structure and function with RFdiffusion
- Str2Str: A Score-based Framework for Zero-shot Protein Conformation Sampling
- FRETpredict: a Python package for FRET efficiency predictions using rotamer libraries
- ViSNet: an equivariant geometry-enhanced graph neural network with vector-scalar interactive message passing for molecules
- Protein structure generation via folding diffusion
- Protein structure generation via folding diffusion
- A Graph Neural Network Approach to Automated Model Building in Cryo-EM Maps
- Learning Local Equivariant Representations for Large-Scale Atomistic Dynamics
- Collective variable discovery in the age of machine learning: reality, hype and everything in between
- Highly accurate protein structure prediction with AlphaFold
- Martini 3: a general purpose force field for coarse-grained molecular dynamics
- DeepEMhancer: a deep learning solution for cryo-EM volume post-processing
- Adversarial Reverse Mapping of Equilibrated Condensed-Phase Molecular\n Structures
- Machine learning approaches for analyzing and enhancing molecular dynamics simulations
- Machine learning approaches for analyzing and enhancing molecular dynamics simulations
- How to learn from inconsistencies: Integrating molecular simulations with experimental data
- A practical guide to the simultaneous determination of protein structure\n and dynamics using metainference
- Reweighted Autoencoded Variational Bayes for Enhanced Sampling (RAVE)
- Metainference: A Bayesian Inference Method for Heterogeneous Systems
- The Protein-Folding Problem, 50 Years On
- CHARMM general force field: A force field for drug‐like molecules compatible with the CHARMM all‐atom additive biological force fields
- Comparison of multiple Amber force fields and development of improved protein backbone parameters
- Scalable molecular dynamics with NAMD
- GROMACS: Fast, flexible, and free
- Topological and energetic factors: what determines the structural details of the transition state ensemble and "on-route" intermediates for protein folding? An investigation for small globular proteins
- Funnels, pathways, and the energy landscape of protein folding: A synthesis
- Nonphysical sampling distributions in Monte Carlo free-energy estimation: Umbrella sampling
Discussions
Related