-
From Biased to Unbiased Dynamics: An Infinitesimal Generator Approach
Authors:
Timothée Devergne,
Vladimir Kostic,
Michele Parrinello,
Massimiliano Pontil
Abstract:
We investigate learning the eigenfunctions of evolution operators for time-reversal invariant stochastic processes, a prime example being the Langevin equation used in molecular dynamics. Many physical or chemical processes described by this equation involve transitions between metastable states separated by high potential barriers that can hardly be crossed during a simulation. To overcome this b…
▽ More
We investigate learning the eigenfunctions of evolution operators for time-reversal invariant stochastic processes, a prime example being the Langevin equation used in molecular dynamics. Many physical or chemical processes described by this equation involve transitions between metastable states separated by high potential barriers that can hardly be crossed during a simulation. To overcome this bottleneck, data are collected via biased simulations that explore the state space more rapidly. We propose a framework for learning from biased simulations rooted in the infinitesimal generator of the process and the associated resolvent operator. We contrast our approach to more common ones based on the transfer operator, showing that it can provably learn the spectral properties of the unbiased system from biased data. In experiments, we highlight the advantages of our method over transfer operator approaches and recent developments based on generator learning, demonstrating its effectiveness in estimating eigenfunctions and eigenvalues. Importantly, we show that even with datasets containing only a few relevant transitions due to sub-optimal biasing, our approach recovers relevant information about the transition mechanism.
△ Less
Submitted 13 June, 2024;
originally announced June 2024.
-
Transfer learning for atomistic simulations using GNNs and kernel mean embeddings
Authors:
John Falk,
Luigi Bonati,
Pietro Novelli,
Michele Parrinello,
Massimiliano Pontil
Abstract:
Interatomic potentials learned using machine learning methods have been successfully applied to atomistic simulations. However, accurate models require large training datasets, while generating reference calculations is computationally demanding. To bypass this difficulty, we propose a transfer learning algorithm that leverages the ability of graph neural networks (GNNs) to represent chemical envi…
▽ More
Interatomic potentials learned using machine learning methods have been successfully applied to atomistic simulations. However, accurate models require large training datasets, while generating reference calculations is computationally demanding. To bypass this difficulty, we propose a transfer learning algorithm that leverages the ability of graph neural networks (GNNs) to represent chemical environments together with kernel mean embeddings. We extract a feature map from GNNs pre-trained on the OC20 dataset and use it to learn the potential energy surface from system-specific datasets of catalytic processes. Our method is further enhanced by incorporating into the kernel the chemical species information, resulting in improved performance and interpretability. We test our approach on a series of realistic datasets of increasing complexity, showing excellent generalization and transferability performance, and improving on methods that rely on GNNs or ridge regression alone, as well as similar fine-tuning approaches.
△ Less
Submitted 20 January, 2024; v1 submitted 2 June, 2023;
originally announced June 2023.
-
Characterizing metastable states with the help of machine learning
Authors:
Pietro Novelli,
Luigi Bonati,
Massimiliano Pontil,
Michele Parrinello
Abstract:
Present-day atomistic simulations generate long trajectories of ever more complex systems. Analyzing these data, discovering metastable states, and uncovering their nature is becoming increasingly challenging. In this paper, we first use the variational approach to conformation dynamics to discover the slowest dynamical modes of the simulations. This allows the different metastable states of the s…
▽ More
Present-day atomistic simulations generate long trajectories of ever more complex systems. Analyzing these data, discovering metastable states, and uncovering their nature is becoming increasingly challenging. In this paper, we first use the variational approach to conformation dynamics to discover the slowest dynamical modes of the simulations. This allows the different metastable states of the system to be located and organized hierarchically. The physical descriptors that characterize metastable states are discovered by means of a machine learning method. We show in the cases of two proteins, Chignolin and Bovine Pancreatic Trypsin Inhibitor, how such analysis can be effortlessly performed in a matter of seconds. Another strength of our approach is that it can be applied to the analysis of both unbiased and biased simulations.
△ Less
Submitted 15 April, 2022;
originally announced April 2022.
-
A microscopic description of acid-base equilibrium
Authors:
Emanuele Grifoni,
GiovanniMaria Piccini,
Michele Parrinello
Abstract:
Acid-base reactions are ubiquitous in nature. Understanding their mechanisms is crucial in many fields, from biochemistry to industrial catalysis. Unfortunately, experiments only give limited information without much insight into the molecular behaviour. Atomistic simulations could complement experiments and shed precious light on microscopic mechanisms. The large free energy barriers connected to…
▽ More
Acid-base reactions are ubiquitous in nature. Understanding their mechanisms is crucial in many fields, from biochemistry to industrial catalysis. Unfortunately, experiments only give limited information without much insight into the molecular behaviour. Atomistic simulations could complement experiments and shed precious light on microscopic mechanisms. The large free energy barriers connected to proton dissociation however make the use of enhanced sampling methods mandatory. Here we perform an ab initio molecular dynamics (MD) simulation and enhance sampling with the help of methadynamics. This has been made possible by the introduction of novel descriptors or collective variables (CVs) that are based on a conceptually new outlook on acid-base equilibria. We test successfully our approach on three different aqueous solutions of acetic acid, ammonia, and bicarbonate. These are representative of acid, basic, and amphoteric behaviour.
△ Less
Submitted 22 January, 2019;
originally announced May 2019.
