Beyond Ensembles: Simulating All-Atom Protein Dynamics in a Learned Latent Space
Not on the map. Why?
Abstract (arXiv)
Simulating the long-timescale dynamics of biomolecules is a central challenge in computational science. While enhanced sampling methods can accelerate these simulations, they rely on pre-defined collective variables that are often difficult to identify, restricting their ability to model complex switching mechanisms between metastable states. A recent generative model, LD-FPG, demonstrated that this problem could be bypassed by learning to sample the static equilibrium ensemble as all-atom deformations from a reference structure, establishing a powerful method for all-atom ensemble generation. However, while this approach successfully captures a system's probable conformations, it does not model the temporal evolution between them. We introduce the Graph Latent Dynamics Propagator (GLDP), a modular component for simulating dynamics within the learned latent space of LD-FPG. We then compare three classes of propagators: (i) score-guided Langevin dynamics, (ii) Koopman-based linear operators, and (iii) autoregressive neural networks. Within a unified encoder-propagator-decoder framework, we evaluate long-horizon stability, backbone and side-chain ensemble fidelity, and temporal kinetics via TICA. Benchmarks on systems ranging from small peptides to mixed-topology proteins and large GPCRs reveal that autoregressive neural networks deliver the most robust long rollouts and coherent physical timescales; score-guided Langevin best recovers side-chain thermodynamics when the score is well learned; and Koopman provides an interpretable, lightweight baseline that tends to damp fluctuations. These results clarify the trade-offs among propagators and offer practical guidance for latent-space simulators of all-atom protein dynamics.
Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 42 of its 75 references that are in the map. First posted on arXiv 2 Sep 2025.
Builds on
42 of 75 references.
References · 42
- 2020Denoising Diffusion Probabilistic Modelscited by 21,076
- 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 9,122
- 2022Flow Matching for Generative Modelingcited by 5,702
- 2018Neural Ordinary Differential Equationscited by 3,779
- 2021E(n) Equivariant Graph Neural Networkscited by 965
- 2020Learning to Simulate Complex Physics with Graph Networkscited by 880
- 2019Hamiltonian Neural Networkscited by 712
- 2020SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networkscited by 607
- 2021Diffusion Schr"odinger Bridge with Applications to Score-Based Generative Modelingcited by 602
- 2022GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generationcited by 516
- 2017Deep learning for universal linear embeddings of nonlinear dynamicscited by 494
- 2020Lagrangian Neural Networkscited by 398
30 more
- 2022DiffDock: Diffusion Steps, Twists, and Turns for Molecular Dockingcited by 313
- 2021Modern Koopman Theory for Dynamical Systemscited by 276
- 2022TorchMD-NET: Equivariant Transformers for Neural Network based Molecular Potentialscited by 182
- 2021Neural SDEs as Infinite-Dimensional GANscited by 161
- 2017VAMPnets: Deep learning of molecular kineticscited by 136
- 2020Forecasting Sequential Data using Consistent Koopman Autoencoderscited by 115
- 2023FAENet: Frame Averaging Equivariant GNN for Materials Modelingcited by 73
- 2023Two for One: Diffusion Models and Force Fields for Coarse-Grained Molecular Dynamicscited by 72
- 2017Time-lagged autoencoders: Deep learning of slow collective variables for molecular kineticscited by 65
- 2024Generative Modeling of Molecular Dynamics Trajectoriescited by 51
- 2024Transferable Boltzmann Generatorscited by 46
- 2024Equivariant Graph Neural Operator for Modeling 3D Dynamicscited by 38
- 2022Flow-matching -- efficient coarse-graining of molecular dynamics without forcescited by 36
- 2022DiffMD: A Geometric Diffusion Model for Molecular Dynamics Simulationscited by 24
- 2025Consistent Sampling and Simulation: Molecular Dynamics with Energy-Based Diffusion Modelscited by 23
- 2017Variational Encoding of Complex Dynamicscited by 21
- 2025Action-Minimization Meets Generative Modeling: Efficient Transition Path Sampling with the Onsager-Machlup Functionalcited by 20
- 2018Deep Generative Markov State Modelscited by 19
- 2023Score dynamics: scaling molecular dynamics with picoseconds timestep via conditional diffusion modelcited by 16
- 2024Temporally Consistent Koopman Autoencoders for Forecasting Dynamical Systemscited by 15
- 2024Geometric Trajectory Diffusion Modelscited by 15
- 2020Molecular Latent Space Simulatorscited by 14
- 2024F3low: Frame-to-Frame Coarse-grained Molecular Dynamics with SE(3) Guided Flow Matchingcited by 9
- 2022GraphVAMPNet, using graph neural networks and variational approach to markov processes for dynamical modeling of biomoleculescited by 8
- 2025ConfRover: Simultaneous Modeling of Protein Conformation and Dynamics via Autoregressioncited by 8
- 2023Koopman Invertible Autoencoder: Leveraging Forward and Backward Dynamics for Temporal Modelingcited by 6
- 2024A Multi-Grained Symmetric Differential Equation Model for Learning Protein-Ligand Binding Dynamicscited by 5
- 2024Graph Neural Network-State Predictive Information Bottleneck (GNN-SPIB) approach for learning molecular thermodynamics and kineticscited by 5
- 2022Path sampling of recurrent neural networks by incorporating known physicscited by 4
- 2025Generative Modeling of Full-Atom Protein Conformations using Latent Diffusion on Graph Embeddingscited by 2
Built on by
Citing papers per year: 2026: 3.
3 papers.