scholarly journals Multiscale Reweighted Stochastic Embedding: Deep Learning of Collective Variables for Enhanced Sampling

2021 ◽  
Author(s):  
Jakub Rydzewski ◽  
Omar Valsson
Keyword(s):  
Deep Learning ◽  
Enhanced Sampling ◽  
Collective Variables

scholarly journals Neural networks-based variationally enhanced sampling

2019 ◽  
Vol 116 (36) ◽  
pp. 17641-17647 ◽  
Author(s):  
Luigi Bonati ◽  
Yue-Yu Zhang ◽  
Michele Parrinello
Keyword(s):  
Neural Networks ◽  
Free Energy ◽  
Atomistic Simulation ◽  
Sampling Method ◽  
Machine Learning Techniques ◽  
Enhanced Sampling ◽  
Collective Variables ◽  
Learning Techniques ◽  
Minimization Technique ◽  
Energy Surfaces

Sampling complex free-energy surfaces is one of the main challenges of modern atomistic simulation methods. The presence of kinetic bottlenecks in such surfaces often renders a direct approach useless. A popular strategy is to identify a small number of key collective variables and to introduce a bias potential that is able to favor their fluctuations in order to accelerate sampling. Here, we propose to use machine-learning techniques in conjunction with the recent variationally enhanced sampling method [O. Valsson, M. Parrinello, Phys. Rev. Lett. 113, 090601 (2014)] in order to determine such potential. This is achieved by expressing the bias as a neural network. The parameters are determined in a variational learning scheme aimed at minimizing an appropriate functional. This required the development of a more efficient minimization technique. The expressivity of neural networks allows representing rapidly varying free-energy surfaces, removes boundary effects artifacts, and allows several collective variables to be handled.

scholarly journals Data-Driven Collective Variables for Enhanced Sampling

2020 ◽  
Vol 11 (8) ◽  
pp. 2998-3004 ◽  
Author(s):  
Luigi Bonati ◽  
Valerio Rizzi ◽  
Michele Parrinello
Keyword(s):  
Data Driven ◽  
Enhanced Sampling ◽  
Collective Variables

scholarly journals Deep learning the slow modes for rare events sampling

2021 ◽  
Vol 118 (44) ◽  
pp. e2113533118
Author(s):  
Luigi Bonati ◽  
GiovanniMaria Piccini ◽  
Michele Parrinello
Keyword(s):  
Sampling Method ◽  
Rare Events ◽  
Conformational Transition ◽  
Transfer Operator ◽  
Atomistic Simulations ◽  
Enhanced Sampling ◽  
Collective Variables ◽  
Long Time ◽  
Computational Resources

The development of enhanced sampling methods has greatly extended the scope of atomistic simulations, allowing long-time phenomena to be studied with accessible computational resources. Many such methods rely on the identification of an appropriate set of collective variables. These are meant to describe the system’s modes that most slowly approach equilibrium under the action of the sampling algorithm. Once identified, the equilibration of these modes is accelerated by the enhanced sampling method of choice. An attractive way of determining the collective variables is to relate them to the eigenfunctions and eigenvalues of the transfer operator. Unfortunately, this requires knowing the long-term dynamics of the system beforehand, which is generally not available. However, we have recently shown that it is indeed possible to determine efficient collective variables starting from biased simulations. In this paper, we bring the power of machine learning and the efficiency of the recently developed on the fly probability-enhanced sampling method to bear on this approach. The result is a powerful and robust algorithm that, given an initial enhanced sampling simulation performed with trial collective variables or generalized ensembles, extracts transfer operator eigenfunctions using a neural network ansatz and then accelerates them to promote sampling of rare events. To illustrate the generality of this approach, we apply it to several systems, ranging from the conformational transition of a small molecule to the folding of a miniprotein and the study of materials crystallization.

scholarly journals Enhanced Sampling of Protein Conformational Transitions via Dynamically Optimized Collective Variables

2018 ◽  
Author(s):  
Z. Faidon Brotzakis ◽  
Michele Parrinello
Keyword(s):  
Degrees Of Freedom ◽  
Conformational Transition ◽  
Conformational Dynamics ◽  
Computational Cost ◽  
Md Simulations ◽  
Conformational Transitions ◽  
Enhanced Sampling ◽  
Collective Variables ◽  
Simulation Based ◽  
Combined Data

AbstractProtein conformational transitions often involve many slow degrees of freedom. Their knowledge would give distinctive advantages since it provides chemical and mechanistic insight and accelerates the convergence of enhanced sampling techniques that rely on collective variables. In this study, we implemented a recently developed variational approach to conformational dynamics metadynamics to the conformational transition of the moderate size protein, L99A T4 Lysozyme. In order to find the slow modes of the system we combined data coming from NMR experiments as well as short MD simulations. A Metadynamics simulation based on these information reveals the presence of two intermediate states, at an affordable computational cost.

scholarly journals Discover, Sample and Refine: Exploring Chemistry with Enhanced Sampling Techniques

2021 ◽  
Author(s):  
Umberto Raucci ◽  
Valerio Rizzi ◽  
Michele Parrinello
Keyword(s):  
Free Energy ◽  
Ad Hoc ◽  
Spectral Graph Theory ◽  
Ab Initio Molecular Dynamics ◽  
Enhanced Sampling ◽  
Free Energy Surface ◽  
Collective Variables ◽  
User Input ◽  
Specific Reactions

Over the last few decades enhanced sampling methods have made great strides. Here, we exploit this progress and propose a modular workflow for blind reaction discovery and characterization of reaction paths. Central to our strategy is the use of the recently developed explore variant of the on-the-fly probability enhanced sampling method. Like metadynamics, this method is based on the identification of appropriate collective variables. Our first step is the discovery of new chemical reactions and it is performed biasing a one dimensional collective variable derived from spectral graph theory. Once new reaction pathways are detected, we construct ad-hoc tailored neural-network based collective variables to improve sampling of specific reactions and finally we refine the results using free energy perturbation theory. Our workflow has been successfully applied to both intramolecular and intermolecular reactions. Without any chemical hypothesis, we discovered several possible products, computed the free energy surface at semiempirical level, and finally refined it with a more accurate Hamiltonian. Our workflow requires minimal user input, and thanks to its modularity and flexibility, can extend the scope of ab initio molecular dynamics for the exploration and characterization of reaction space.

scholarly journals Achieving Reversible Ligand-Protein Unbinding with Deep Learning and Molecular Dynamics through RAVE

2018 ◽  
Author(s):  
João Marcelo Lamim Ribeiro ◽  
Pratyush Tiwary
Keyword(s):  
Molecular Dynamics ◽  
Free Energy ◽  
Deep Learning ◽  
Rare Events ◽  
Binding Free Energy ◽  
Piecewise Linear ◽  
Chem Phys ◽  
Enhanced Sampling ◽  
Order Of Magnitude ◽  
Low Dimensional

AbstractIn this work we demonstrate how to leverage our recent iterative deep learning–all atom molecular dynamics (MD) technique “Reweighted autoencoded variational Bayes for enhanced sampling (RAVE)” (Ribeiro, Bravo, Wang, Tiwary, J. Chem. Phys. 149, 072301 (2018)) for sampling protein-ligand unbinding mechanisms and calculating absolute binding affinities when plagued with difficult to sample rare events. RAVE iterates between rounds of MD and deep learning, and unlike other enhanced sampling methods, it stands out in simultaneously learning both a low-dimensional physically interpretable reaction coordinate (RC) and associated free energy. Here, we introduce a simple but powerful extension to RAVE which allows learning a position-dependent RC expressed as a superposition of piecewise linear RCs valid in different metastable states. With this approach, we retain the original physical interpretability of a RAVE-derived RC while making it applicable to a wider range of complex systems. We demonstrate how in its multi-dimensional form introduced here, RAVE can efficiently simulate the unbinding of the tightly bound benzene-lysozyme (L99A variant) complex, in all atom-precision and with minimal use of human intuition except for the choice of a larger dictionary of order parameters. These simulations had a 100 % success rate, and took between 3–50 nanoseconds for a process that takes on an average close to few hundred milliseconds, thereby reflecting a seven order of magnitude acceleration relative to straightforward MD. Furthermore, without any time-dependent biasing, the trajectories display clear back–and– forth movement between various metastable intermediates, demonstrating the reliability of the RC and its probability distribution learnt in RAVE. Our binding free energy is in good agreement with other reported simulation results. We thus believe that RAVE, especially in its multi-dimensional variant introduced here, will be a useful tool for simulating the dissociation process of practical biophysical systems with rare events in an automated manner with minimal use of human intuition.

scholarly journals The role of water in host-guest interaction

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Valerio Rizzi ◽  
Luigi Bonati ◽  
Narjes Ansari ◽  
Michele Parrinello
Keyword(s):  
Ligand Binding ◽  
Degrees Of Freedom ◽  
Enhanced Sampling ◽  
Free Energies ◽  
Collective Variables ◽  
Complex Behaviour ◽  
Field Quality ◽  
Non Local ◽  
Binding Free Energies

AbstractOne of the main applications of atomistic computer simulations is the calculation of ligand binding free energies. The accuracy of these calculations depends on the force field quality and on the thoroughness of configuration sampling. Sampling is an obstacle in simulations due to the frequent appearance of kinetic bottlenecks in the free energy landscape. Very often this difficulty is circumvented by enhanced sampling techniques. Typically, these techniques depend on the introduction of appropriate collective variables that are meant to capture the system’s degrees of freedom. In ligand binding, water has long been known to play a key role, but its complex behaviour has proven difficult to fully capture. In this paper we combine machine learning with physical intuition to build a non-local and highly efficient water-describing collective variable. We use it to study a set of host-guest systems from the SAMPL5 challenge. We obtain highly accurate binding free energies and good agreement with experiments. The role of water during the binding process is then analysed in some detail.

Sign in / Sign up

Export Citation Format

Share Document