-
Thermodynamically Informed Multimodal Learning of High-Dimensional Free Energy Models in Molecular Coarse Graining
Authors:
Blake R. Duschatko,
Xiang Fu,
Cameron Owen,
Yu Xie,
Albert Musaelian,
Tommi Jaakkola,
Boris Kozinsky
Abstract:
We present a differentiable formalism for learning free energies that is capable of capturing arbitrarily complex model dependencies on coarse-grained coordinates and finite-temperature response to variation of general system parameters. This is done by endowing models with explicit dependence on temperature and parameters and by exploiting exact differential thermodynamic relationships between th…
▽ More
We present a differentiable formalism for learning free energies that is capable of capturing arbitrarily complex model dependencies on coarse-grained coordinates and finite-temperature response to variation of general system parameters. This is done by endowing models with explicit dependence on temperature and parameters and by exploiting exact differential thermodynamic relationships between the free energy, ensemble averages, and response properties. Formally, we derive an approach for learning high-dimensional cumulant generating functions using statistical estimates of their derivatives, which are observable cumulants of the underlying random variable. The proposed formalism opens ways to resolve several outstanding challenges in bottom-up molecular coarse graining dealing with multiple minima and state dependence. This is realized by using additional differential relationships in the loss function to significantly improve the learning of free energies, while exactly preserving the Boltzmann distribution governing the corresponding fine-grain all-atom system. As an example, we go beyond the standard force-matching procedure to demonstrate how leveraging the thermodynamic relationship between free energy and values of ensemble averaged all-atom potential energy improves the learning efficiency and accuracy of the free energy model. The result is significantly better sampling statistics of structural distribution functions. The theoretical framework presented here is demonstrated via implementations in both kernel-based and neural network machine learning regression methods and opens new ways to train accurate machine learning models for studying thermodynamic and response properties of complex molecular systems.
△ Less
Submitted 29 May, 2024;
originally announced May 2024.
-
Chemical Transferability and Accuracy of Ionic Liquid Simulations with Machine Learning Interatomic Potentials
Authors:
Zachary A. H. Goodwin,
Malia B. Wenny,
Julia H. Yang,
Andrea Cepellotti,
Kyle Bystrom,
Blake R. Duschatko,
Anders Johansson,
Lixin Sun,
Simon Batzner,
Albert Musaelian,
Jarad A. Mason,
Boris Kozinsky,
Nicola Molinari
Abstract:
Ionic liquids (ILs) are an exciting class of electrolytes finding applications in many areas from energy storage to solvents, and have been touted as ``designer solvents'', as they can be mixed to precisely tailor the physiochemical properties. As using machine learning interatomic potentials (MLIPs) to simulate ILs is still relatively unexplored, several questions need to be answered to see if ML…
▽ More
Ionic liquids (ILs) are an exciting class of electrolytes finding applications in many areas from energy storage to solvents, and have been touted as ``designer solvents'', as they can be mixed to precisely tailor the physiochemical properties. As using machine learning interatomic potentials (MLIPs) to simulate ILs is still relatively unexplored, several questions need to be answered to see if MLIPs can be transformative for ILs. Since ILs are often not pure, but are either mixed together or contain additives, we first demonstrate that a MLIP can be trained to be chemically transferable, i.e., can be applied to compositions not directly train on, but with similar chemistries, whilst only being trained on a few compositions. We also investigate the accuracy of MLIPs for a novel IL, which we experimentally synthesis and characterise. Our MLIP trained on $\sim$200 DFT frames is in reasonable agreement with our experiments and good agreement against DFT, demonstrating the data-efficient power of this method.
△ Less
Submitted 27 May, 2024; v1 submitted 4 March, 2024;
originally announced March 2024.
-
Uncertainty Driven Active Learning of Coarse Grained Free Energy Models
Authors:
Blake R. Duschatko,
Jonathan Vandermause,
Nicola Molinari,
Boris Kozinsky
Abstract:
Coarse graining techniques play an essential role in accelerating molecular simulations of systems with large length and time scales. Theoretically grounded bottom-up models are appealing due to their thermodynamic consistency with the underlying all-atom models. In this direction, machine learning approaches hold great promise to fitting complex many-body data. However, training models may requir…
▽ More
Coarse graining techniques play an essential role in accelerating molecular simulations of systems with large length and time scales. Theoretically grounded bottom-up models are appealing due to their thermodynamic consistency with the underlying all-atom models. In this direction, machine learning approaches hold great promise to fitting complex many-body data. However, training models may require collection of large amounts of expensive data. Moreover, quantifying trained model accuracy is challenging, especially in cases of non-trivial free energy configurations, where training data may be sparse. We demonstrate a path towards uncertainty-aware models of coarse grained free energy surfaces. Specifically, we show that principled Bayesian model uncertainty allows for efficient data collection through an on-the-fly active learning framework and open the possibility of adaptive transfer of models across different chemical systems. Uncertainties also characterize models' accuracy of free energy predictions, even when training is performed only on forces. This work helps pave the way towards efficient autonomous training of reliable and uncertainty aware many-body machine learned coarse grain models.
△ Less
Submitted 28 October, 2022;
originally announced October 2022.
-
Tracking the quantized information transfer at the edge of a chiral Floquet phase
Authors:
Blake R. Duschatko,
Philipp T. Dumitrescu,
Andrew C. Potter
Abstract:
Two-dimensional arrays of periodically driven qubits can host inherently dynamical topological phases with anomalous chiral edge dynamics. These chiral Floquet phases are formally characterized by a dynamical topological invariant, the chiral unitary index. Introducing a quantity called the chiral mutual information, we show that this invariant can be precisely interpreted in terms of a quantized…
▽ More
Two-dimensional arrays of periodically driven qubits can host inherently dynamical topological phases with anomalous chiral edge dynamics. These chiral Floquet phases are formally characterized by a dynamical topological invariant, the chiral unitary index. Introducing a quantity called the chiral mutual information, we show that this invariant can be precisely interpreted in terms of a quantized chiral transfer of quantum information along the edge of the system, and devise a physical setup to measure it.
△ Less
Submitted 5 April, 2018;
originally announced April 2018.
