skip to main content
DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Molecular latent space simulators

Abstract

Small integration time steps limit molecular dynamics (MD) simulations to millisecond time scales. Markov state models (MSMs) and equation-free approaches learn low-dimensional kinetic models from MD simulation data by performing configurational or dynamical coarse-graining of the state space. The learned kinetic models enable the efficient generation of dynamical trajectories over vastly longer time scales than are accessible by MD, but the discretization of configurational space and/or absence of a means to reconstruct molecular configurations precludes the generation of continuous all-atom molecular trajectories. We propose latent space simulators (LSS) to learn kinetic models for continuous all-atom simulation trajectories by training three deep learning networks to (i) learn the slow collective variables of the molecular system, (ii) propagate the system dynamics within this slow latent space, and (iii) generatively reconstruct molecular configurations. We demonstrate the approach in an application to Trp-cage miniprotein to produce novel ultra-long synthetic folding trajectories that accurately reproduce all-atom molecular structure, thermodynamics, and kinetics at six orders of magnitude lower cost than MD. The dramatically lower cost of trajectory generation enables greatly improved sampling and greatly reduced statistical uncertainties in estimated thermodynamic averages and kinetic rates.

Authors:
 [1];  [2]; ORCiD logo [1]
  1. Pritzker School of Molecular Engineering, University of Chicago, Chicago, USA
  2. Department of Physics, University of Illinois at Urbana-Champaign, Urbana, USA
Publication Date:
Research Org.:
Argonne National Lab. (ANL), Argonne, IL (United States)
Sponsoring Org.:
USDOE Office of Science (SC), Basic Energy Sciences (BES). Materials Sciences & Engineering Division
OSTI Identifier:
1657300
Alternate Identifier(s):
OSTI ID: 1651122
Grant/Contract Number:  
AC02-06CH11357
Resource Type:
Published Article
Journal Name:
Chemical Science
Additional Journal Information:
Journal Name: Chemical Science Journal Volume: 11 Journal Issue: 35; Journal ID: ISSN 2041-6520
Publisher:
Royal Society of Chemistry
Country of Publication:
United Kingdom
Language:
English
Subject:
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CHEMISTRY

Citation Formats

Sidky, Hythem, Chen, Wei, and Ferguson, Andrew L. Molecular latent space simulators. United Kingdom: N. p., 2020. Web. https://doi.org/10.1039/D0SC03635H.
Sidky, Hythem, Chen, Wei, & Ferguson, Andrew L. Molecular latent space simulators. United Kingdom. https://doi.org/10.1039/D0SC03635H
Sidky, Hythem, Chen, Wei, and Ferguson, Andrew L. Wed . "Molecular latent space simulators". United Kingdom. https://doi.org/10.1039/D0SC03635H.
@article{osti_1657300,
title = {Molecular latent space simulators},
author = {Sidky, Hythem and Chen, Wei and Ferguson, Andrew L.},
abstractNote = {Small integration time steps limit molecular dynamics (MD) simulations to millisecond time scales. Markov state models (MSMs) and equation-free approaches learn low-dimensional kinetic models from MD simulation data by performing configurational or dynamical coarse-graining of the state space. The learned kinetic models enable the efficient generation of dynamical trajectories over vastly longer time scales than are accessible by MD, but the discretization of configurational space and/or absence of a means to reconstruct molecular configurations precludes the generation of continuous all-atom molecular trajectories. We propose latent space simulators (LSS) to learn kinetic models for continuous all-atom simulation trajectories by training three deep learning networks to (i) learn the slow collective variables of the molecular system, (ii) propagate the system dynamics within this slow latent space, and (iii) generatively reconstruct molecular configurations. We demonstrate the approach in an application to Trp-cage miniprotein to produce novel ultra-long synthetic folding trajectories that accurately reproduce all-atom molecular structure, thermodynamics, and kinetics at six orders of magnitude lower cost than MD. The dramatically lower cost of trajectory generation enables greatly improved sampling and greatly reduced statistical uncertainties in estimated thermodynamic averages and kinetic rates.},
doi = {10.1039/D0SC03635H},
journal = {Chemical Science},
number = 35,
volume = 11,
place = {United Kingdom},
year = {2020},
month = {9}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record
https://doi.org/10.1039/D0SC03635H

Save / Share:

Works referenced in this record:

Girsanov reweighting for metadynamics simulations
journal, August 2018

  • Donati, Luca; Keller, Bettina G.
  • The Journal of Chemical Physics, Vol. 149, Issue 7
  • DOI: 10.1063/1.5027728

Nonphysical sampling distributions in Monte Carlo free-energy estimation: Umbrella sampling
journal, February 1977


Enhanced Sampling in Molecular Dynamics Using Metadynamics, Replica-Exchange, and Temperature-Acceleration
journal, December 2013


Dynamical reweighting: Improved estimates of dynamical properties from simulations at multiple temperatures
journal, June 2011

  • Chodera, John D.; Swope, William C.; Noé, Frank
  • The Journal of Chemical Physics, Vol. 134, Issue 24
  • DOI: 10.1063/1.3592152

Scalable molecular dynamics with NAMD
journal, January 2005

  • Phillips, James C.; Braun, Rosemary; Wang, Wei
  • Journal of Computational Chemistry, Vol. 26, Issue 16, p. 1781-1802
  • DOI: 10.1002/jcc.20289

TweTriS: Twenty trillion-atom simulation
journal, January 2019

  • Tchipev, Nikola; Seckler, Steffen; Heinen, Matthias
  • The International Journal of High Performance Computing Applications, Vol. 33, Issue 5
  • DOI: 10.1177/1094342018819741

Model-Free Prediction of Large Spatiotemporally Chaotic Systems from Data: A Reservoir Computing Approach
journal, January 2018


VAMPnets for deep learning of molecular kinetics
journal, January 2018


Kinetic Distance and Kinetic Maps from Molecular Dynamics Simulation
journal, September 2015

  • Noé, Frank; Clementi, Cecilia
  • Journal of Chemical Theory and Computation, Vol. 11, Issue 10
  • DOI: 10.1021/acs.jctc.5b00553

Variational encoding of complex dynamics
journal, June 2018

  • Hernández, Carlos X.; Wayment-Steele, Hannah K.; Sultan, Mohammad M.
  • Physical Review E, Vol. 97, Issue 6
  • DOI: 10.1103/PhysRevE.97.062412

Commute Maps: Separating Slowly Mixing Molecular Configurations for Kinetic Modeling
journal, October 2016

  • Noé, Frank; Banisch, Ralf; Clementi, Cecilia
  • Journal of Chemical Theory and Computation, Vol. 12, Issue 11
  • DOI: 10.1021/acs.jctc.6b00762

Collective variable discovery and enhanced sampling using autoencoders: Innovations in network architecture and error function design
journal, August 2018

  • Chen, Wei; Tan, Aik Rui; Ferguson, Andrew L.
  • The Journal of Chemical Physics, Vol. 149, Issue 7
  • DOI: 10.1063/1.5023804

How Fast-Folding Proteins Fold
journal, October 2011


Rapid Exploration of Configuration Space with Diffusion-Map-Directed Molecular Dynamics
journal, July 2013

  • Zheng, Wenwei; Rohrdanz, Mary A.; Clementi, Cecilia
  • The Journal of Physical Chemistry B, Vol. 117, Issue 42
  • DOI: 10.1021/jp401911h

Simulating materials failure by using up to one billion atoms and the world's fastest computer: Work-hardening
journal, April 2002

  • Abraham, F. F.; Walkup, R.; Gao, H.
  • Proceedings of the National Academy of Sciences, Vol. 99, Issue 9
  • DOI: 10.1073/pnas.062054999

Deep Learning Unravels a Dynamic Hierarchy While Empowering Molecular Dynamics Simulations
journal, February 2020


High-Resolution Markov State Models for the Dynamics of Trp-Cage Miniprotein Constructed Over Slow Folding Modes Identified by State-Free Reversible VAMPnets
journal, August 2019

  • Sidky, Hythem; Chen, Wei; Ferguson, Andrew L.
  • The Journal of Physical Chemistry B, Vol. 123, Issue 38
  • DOI: 10.1021/acs.jpcb.9b05578

Variational Approach to Molecular Kinetics
journal, March 2014

  • Nüske, Feliks; Keller, Bettina G.; Pérez-Hernández, Guillermo
  • Journal of Chemical Theory and Computation, Vol. 10, Issue 4
  • DOI: 10.1021/ct4009156

Extended dynamic mode decomposition with dictionary learning: A data-driven adaptive spectral decomposition of the Koopman operator
journal, October 2017

  • Li, Qianxiao; Dietrich, Felix; Bollt, Erik M.
  • Chaos: An Interdisciplinary Journal of Nonlinear Science, Vol. 27, Issue 10
  • DOI: 10.1063/1.4993854

Fast recovery of free energy landscapes via diffusion-map-directed molecular dynamics
journal, January 2014

  • Preto, Jordane; Clementi, Cecilia
  • Phys. Chem. Chem. Phys., Vol. 16, Issue 36
  • DOI: 10.1039/C3CP54520B

Markov State Models: From an Art to a Science
journal, February 2018

  • Husic, Brooke E.; Pande, Vijay S.
  • Journal of the American Chemical Society, Vol. 140, Issue 7
  • DOI: 10.1021/jacs.7b12191

Equation-Free Multiscale Computation: Algorithms and Applications
journal, May 2009


Deep learning for universal linear embeddings of nonlinear dynamics
journal, November 2018


Machine learning for collective variable discovery and enhanced sampling in biomolecular simulation
journal, March 2020


MSMBuilder2: Modeling Conformational Dynamics on the Picosecond to Millisecond Scale
journal, August 2011

  • Beauchamp, Kyle A.; Bowman, Gregory R.; Lane, Thomas J.
  • Journal of Chemical Theory and Computation, Vol. 7, Issue 10
  • DOI: 10.1021/ct200463m

Everything you wanted to know about Markov State Models but were afraid to ask
journal, September 2010


Capabilities and limitations of time-lagged autoencoders for slow mode discovery in dynamical systems
journal, August 2019

  • Chen, Wei; Sidky, Hythem; Ferguson, Andrew L.
  • The Journal of Chemical Physics, Vol. 151, Issue 6
  • DOI: 10.1063/1.5112048

Machine Calculation of Thermodynamic Properties of a Simple Fluid at Supercritical Temperatures
journal, December 1967

  • McDonald, I. R.; Singer, K.
  • The Journal of Chemical Physics, Vol. 47, Issue 11
  • DOI: 10.1063/1.1701695

Hierarchical Time-Lagged Independent Component Analysis: Computing Slow Modes and Reaction Coordinates for Large Molecular Systems
journal, November 2016

  • Pérez-Hernández, Guillermo; Noé, Frank
  • Journal of Chemical Theory and Computation, Vol. 12, Issue 12
  • DOI: 10.1021/acs.jctc.6b00738

Improvements in Markov State Model Construction Reveal Many Non-Native Interactions in the Folding of NTL9
journal, March 2013

  • Schwantes, Christian R.; Pande, Vijay S.
  • Journal of Chemical Theory and Computation, Vol. 9, Issue 4
  • DOI: 10.1021/ct300878a

Markov models of molecular kinetics: Generation and validation
journal, May 2011

  • Prinz, Jan-Hendrik; Wu, Hao; Sarich, Marco
  • The Journal of Chemical Physics, Vol. 134, Issue 17
  • DOI: 10.1063/1.3565032

Intrinsic map dynamics exploration for uncharted effective free-energy landscapes
journal, June 2017

  • Chiavazzo, Eliodoro; Covino, Roberto; Coifman, Ronald R.
  • Proceedings of the National Academy of Sciences, Vol. 114, Issue 28
  • DOI: 10.1073/pnas.1621481114

Variational Approach for Learning Markov Processes from Time Series Data
journal, August 2019


A Variational Approach to Modeling Slow Processes in Stochastic Dynamical Systems
journal, January 2013

  • Noé, Frank; Nüske, Feliks
  • Multiscale Modeling & Simulation, Vol. 11, Issue 2
  • DOI: 10.1137/110858616

Optimal Data-Driven Estimation of Generalized Markov State Models for Non-Equilibrium Dynamics
journal, February 2018


Deflation reveals dynamical structure in nondominant reaction coordinates
journal, August 2019

  • Husic, Brooke E.; Noé, Frank
  • The Journal of Chemical Physics, Vol. 151, Issue 5
  • DOI: 10.1063/1.5099194

Time-lagged autoencoders: Deep learning of slow collective variables for molecular kinetics
journal, June 2018

  • Wehmeyer, Christoph; Noé, Frank
  • The Journal of Chemical Physics, Vol. 148, Issue 24
  • DOI: 10.1063/1.5011399

Data-Driven Model Reduction and Transfer Operator Approximation
journal, January 2018


GPU-accelerated molecular modeling coming of age
journal, September 2010

  • Stone, John E.; Hardy, David J.; Ufimtsev, Ivan S.
  • Journal of Molecular Graphics and Modelling, Vol. 29, Issue 2
  • DOI: 10.1016/j.jmgm.2010.06.010

Nonlinear discovery of slow molecular modes using state-free reversible VAMPnets
journal, June 2019

  • Chen, Wei; Sidky, Hythem; Ferguson, Andrew L.
  • The Journal of Chemical Physics, Vol. 150, Issue 21
  • DOI: 10.1063/1.5092521

Introduction to Markov state modeling with the PyEMMA software [Article v1.0]
journal, January 2019

  • Wehmeyer, Christoph; Scherer, Martin K.; Hempel, Tim
  • Living Journal of Computational Molecular Science, Vol. 1, Issue 1
  • DOI: 10.33011/livecoms.1.1.5965

Strong scaling of general-purpose molecular dynamics simulations on GPUs
journal, July 2015

  • Glaser, Jens; Nguyen, Trung Dac; Anderson, Joshua A.
  • Computer Physics Communications, Vol. 192
  • DOI: 10.1016/j.cpc.2015.02.028

Equation-free: The computer-aided analysis of complex multiscale systems
journal, January 2004

  • Kevrekidis, Ioannis G.; Gear, C. William; Hummer, Gerhard
  • AIChE Journal, Vol. 50, Issue 7
  • DOI: 10.1002/aic.10106

Nonlinear generalized Langevin equations
journal, November 1973


Transport, Collective Motion, and Brownian Motion
journal, March 1965


Simulating materials failure by using up to one billion atoms and the world's fastest computer: Brittle fracture
journal, April 2002

  • Abraham, F. F.; Walkup, R.; Gao, H.
  • Proceedings of the National Academy of Sciences, Vol. 99, Issue 9
  • DOI: 10.1073/pnas.062012699

Equation-Free, Coarse-Grained Multiscale Computation: Enabling Mocroscopic Simulators to Perform System-Level Analysis
journal, January 2003

  • Gear, C. William; Hyman, James M.; Kevrekidid, Panagiotis G.
  • Communications in Mathematical Sciences, Vol. 1, Issue 4
  • DOI: 10.4310/CMS.2003.v1.n4.a5

Perspective: Computer simulations of long time dynamics
journal, February 2016

  • Elber, Ron
  • The Journal of Chemical Physics, Vol. 144, Issue 6
  • DOI: 10.1063/1.4940794

Unconstrained enhanced sampling for free energy calculations of biomolecules: a review
journal, July 2016


NSF’s Inaugural Software Institutes: The Science Gateways Community Institute and the Molecular Sciences Software Institute
journal, September 2018


Identification of slow molecular order parameters for Markov model construction
journal, July 2013

  • Pérez-Hernández, Guillermo; Paul, Fabian; Giorgino, Toni
  • The Journal of Chemical Physics, Vol. 139, Issue 1
  • DOI: 10.1063/1.4811489

Perspective: Computational chemistry software and its advancement as illustrated through three grand challenge cases for molecular science
journal, November 2018

  • Krylov, Anna; Windus, Theresa L.; Barnes, Taylor
  • The Journal of Chemical Physics, Vol. 149, Issue 18
  • DOI: 10.1063/1.5052551