Machine Learning of Coarse-Grained Molecular Dynamics Force Fields

By Jiang Wang, Christoph Wehmeyer, Frank No'e, Cecilia Clementi. ACS Central Science, 2018

484 citations recorded by the source corpus.

Atomistic or ab initio molecular dynamics simulations are widely used to predict thermodynamics and kinetics and relate them to molecular structure. A common approach to go beyond the time- and length-scales accessible with such computationally expensive simulations is the definition of coarse-grained molecular models. Existing coarse-graining approaches define an effective interaction potential to match defined properties of high-resolution models or experimental data. In this paper, we reformulate coarse-graining as a supervised machine learning problem. We use statistical learning theory to decompose the coarse-graining error and cross-validation to select and compare the performance of different models. We introduce CGnets, a deep learning approach, that learns coarse-grained free energy functions and can be trained by a force-matching scheme. CGnets maintain all physically relevant invariances and allow one to incorporate prior physics knowledge to avoid sampling of unphysical structures. We show that CGnets can capture all-atom explicit-solvent free energy surfaces with models using only a few coarse-grained beads and no solvent, while classical coarse-graining methods fail to capture crucial features of the free energy surface. Thus, CGnets are able to capture multibody terms that emerge from the dimensionality reduction.

Subfields: Electrical and Electronic Engineering, Materials Chemistry, Molecular Biology

View Paper: https://www.semanticscholar.org/paper/f74efb7f8b18ac1c6bdbafc5a6966e2917624a7b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Electrical and Electronic Engineering

In Materials Chemistry

In Molecular Biology

Citations in the corpus

Deep Learning in Protein Structural Modeling and Design (2020, background): “Wang et al.(138) developed CGNet and learned the coarse-grained model of the mini protein, chignolin, in which the atoms of a residue aremapped to the corresponding Ca atom.”
Scalable emulation of protein equilibrium ensembles with generative deep learning (2025, unknown): “Machine-learned coarse-grained MD models have an opportunity to achieve similar accuracy as all-atom MD at 2-3 orders of magnitude lower computational cost [10, 11] but are still under development.”
Physics-driven coarse-grained model for biomolecular phase separation with near-quantitative accuracy (2021, background): “…with the data from which they were parameterized, but, because in such high-dimensional problems many solutions are similarly good, careful curation is required to obtain statistically meaningful results for a specific system, and even these may still not be transferable to similar molecules 37 .”
Hamiltonian Neural Networks (2019, methodology): “There are a few previous works which do this [42, 35, 28], but their scope and implementation details diverge from this work and from one another.”
Data-driven approximation of the Koopman generator: Model reduction, system identification, and control (2019, methodology): “The data set is the same as in reference [51] and comprises one million snapshots of Langevin dynamics saved every 1 ps.”
Discovering Symbolic Models from Deep Learning with Inductive Biases (2020, methodology): “This is a Hamiltonian version of the Lagrangian Graph Network in [38], and is similar to [39].”
Boltzmann generators: Sampling equilibrium states of many-body systems with deep learning (2019, unknown): “Alternatively,Boltzmanngeneratorscouldbeused tosamplelower-dimensionalfree-energysurfaceslearnedfromall-atommodels( 35 , 36 ).”