Showing 1–9 of 9 results for author: Bore, S L

DeePMD-kit v3: A Multiple-Backend Framework for Machine Learning Potentials

Authors: Jinzhe Zeng, Duo Zhang, Anyang Peng, Xiangyu Zhang, Sensen He, Yan Wang, Xinzijian Liu, Hangrui Bi, Yifan Li, Chun Cai, Chengqian Zhang, Yiming Du, Jia-Xin Zhu, Pinghui Mo, Zhengtao Huang, Qiyu Zeng, Shaochen Shi, Xuejian Qin, Zhaoxi Yu, Chenxing Luo, Ye Ding, Yun-Pei Liu, Ruosong Shi, Zhenyu Wang, Sigbjørn Løland Bore , et al. (22 additional authors not shown)

Abstract: In recent years, machine learning potentials (MLPs) have become indispensable tools in physics, chemistry, and materials science, driving the development of software packages for molecular dynamics (MD) simulations and related applications. These packages, typically built on specific machine learning frameworks such as TensorFlow, PyTorch, or JAX, face integration challenges when advanced applicat… ▽ More In recent years, machine learning potentials (MLPs) have become indispensable tools in physics, chemistry, and materials science, driving the development of software packages for molecular dynamics (MD) simulations and related applications. These packages, typically built on specific machine learning frameworks such as TensorFlow, PyTorch, or JAX, face integration challenges when advanced applications demand communication across different frameworks. The previous TensorFlow-based implementation of DeePMD-kit exemplified these limitations. In this work, we introduce DeePMD-kit version 3, a significant update featuring a multi-backend framework that supports TensorFlow, PyTorch, JAX, and PaddlePaddle backends, and demonstrate the versatility of this architecture through the integration of other MLPs packages and of Differentiable Molecular Force Field. This architecture allows seamless backend switching with minimal modifications, enabling users and developers to integrate DeePMD-kit with other packages using different machine learning frameworks. This innovation facilitates the development of more complex and interoperable workflows, paving the way for broader applications of MLPs in scientific research. △ Less

Submitted 27 February, 2025; v1 submitted 26 February, 2025; originally announced February 2025.

Journal ref: J. Chem. Theory Comput., 2025

Learning atomic forces from uncertainty-calibrated adversarial attacks

Authors: Henrique Musseli Cezar, Tilmann Bodenstein, Henrik Andersen Sveinsson, Morten Ledum, Simen Reine, Sigbjørn Løland Bore

Abstract: Adversarial approaches, which intentionally challenge machine learning models by generating difficult examples, are increasingly being adopted to improve machine learning interatomic potentials (MLIPs). While already providing great practical value, little is known about the actual prediction errors of MLIPs on adversarial structures and whether these errors can be controlled. We propose the Calib… ▽ More Adversarial approaches, which intentionally challenge machine learning models by generating difficult examples, are increasingly being adopted to improve machine learning interatomic potentials (MLIPs). While already providing great practical value, little is known about the actual prediction errors of MLIPs on adversarial structures and whether these errors can be controlled. We propose the Calibrated Adversarial Geometry Optimization (CAGO) algorithm to discover adversarial structures with user-assigned errors. Through uncertainty calibration, the estimated uncertainty of MLIPs is unified with real errors. By performing geometry optimization for calibrated uncertainty, we reach adversarial structures with the user-assigned target MLIP prediction error. Integrating with active learning pipelines, we benchmark CAGO, demonstrating stable MLIPs that systematically converge structural, dynamical, and thermodynamical properties for liquid water and water adsorption in a metal-organic framework within only hundreds of training structures, where previously many thousands were typically required. △ Less

Submitted 20 May, 2025; v1 submitted 25 February, 2025; originally announced February 2025.

DeePMD-kit v2: A software package for Deep Potential models

Authors: Jinzhe Zeng, Duo Zhang, Denghui Lu, Pinghui Mo, Zeyu Li, Yixiao Chen, Marián Rynik, Li'ang Huang, Ziyao Li, Shaochen Shi, Yingze Wang, Haotian Ye, Ping Tuo, Jiabin Yang, Ye Ding, Yifan Li, Davide Tisi, Qiyu Zeng, Han Bao, Yu Xia, Jiameng Huang, Koki Muraoka, Yibo Wang, Junhan Chang, Fengbo Yuan , et al. (22 additional authors not shown)

Abstract: DeePMD-kit is a powerful open-source software package that facilitates molecular dynamics simulations using machine learning potentials (MLP) known as Deep Potential (DP) models. This package, which was released in 2017, has been widely used in the fields of physics, chemistry, biology, and material science for studying atomistic systems. The current version of DeePMD-kit offers numerous advanced… ▽ More DeePMD-kit is a powerful open-source software package that facilitates molecular dynamics simulations using machine learning potentials (MLP) known as Deep Potential (DP) models. This package, which was released in 2017, has been widely used in the fields of physics, chemistry, biology, and material science for studying atomistic systems. The current version of DeePMD-kit offers numerous advanced features such as DeepPot-SE, attention-based and hybrid descriptors, the ability to fit tensile properties, type embedding, model deviation, Deep Potential - Range Correction (DPRc), Deep Potential Long Range (DPLR), GPU support for customized operators, model compression, non-von Neumann molecular dynamics (NVNMD), and improved usability, including documentation, compiled binary packages, graphical user interfaces (GUI), and application programming interfaces (API). This article presents an overview of the current major version of the DeePMD-kit package, highlighting its features and technical details. Additionally, the article benchmarks the accuracy and efficiency of different models and discusses ongoing developments. △ Less

Submitted 18 April, 2023; originally announced April 2023.

Comments: 51 pages, 2 figures

ACM Class: J.2

Journal ref: J. Chem. Phys. 159, 054801 (2023)

On the equivalence of the hybrid particle-field and Gaussian core models

Authors: Morten Ledum, Samiran Sen, Sigbjørn Løland Bore, Michele Cascella

Abstract: Hybrid particle-field molecular dynamics is a molecular simulation strategy wherein particles couple to a density field instead of through ordinary pair potentials. Traditionally considered a mean-field theory, a momentum and energy-conserving hybrid particle-field formalism has recently been introduced, which was demonstrated to approach the Gaussian Core model potential in the grid-converged lim… ▽ More Hybrid particle-field molecular dynamics is a molecular simulation strategy wherein particles couple to a density field instead of through ordinary pair potentials. Traditionally considered a mean-field theory, a momentum and energy-conserving hybrid particle-field formalism has recently been introduced, which was demonstrated to approach the Gaussian Core model potential in the grid-converged limit. Here, we expand on and generalize the correspondence between the Hamiltonian hybrid particle-field method and particle-particle pair potentials. Using the spectral procedure suggested by Bore and Cascella, we establish compatibility to any local soft pair potential in the limit of infinitesimal grid spacing. Furthermore, we document how the mean-field regime often observed in hybrid particle-field simulations is due to the systems under consideration, and not an inherent property of the model. Considering the Gaussian filter form in particular, we demonstrate the ability of the Hamiltonian hybrid particle-field model of recovering all structural and dynamical properties of the Gaussian Core model, including solid phases, a first-order phase transition, and anomalous transport properties. We quantify the impact of the grid spacing on the correspondence, as well as the effect of the particle-field filtering length scale on the emergent particle-particle correlations. △ Less

Submitted 2 February, 2023; originally announced February 2023.

Hamiltonian and Alias-Free Hybrid Particle-Field Molecular Dynamics

Authors: Sigbjørn Løland Bore, Michele Cascella

Abstract: Hybrid particle-field molecular dynamics combines standard molecular potentials with density-field models into a computationally efficient methodology that is well-adapted for the study of mesoscale soft matter systems. Here, we introduce a new formulation based on filtered densities and a particle-mesh formalism that allows for Hamiltonian dynamics and alias-free force computation. This is achiev… ▽ More Hybrid particle-field molecular dynamics combines standard molecular potentials with density-field models into a computationally efficient methodology that is well-adapted for the study of mesoscale soft matter systems. Here, we introduce a new formulation based on filtered densities and a particle-mesh formalism that allows for Hamiltonian dynamics and alias-free force computation. This is achieved by introducing a length scale for the particle-field interactions independent of the numerical grid used to represent the density fields, enabling systematic convergence of the forces upon grid refinement. Our scheme generalises the original particle-field molecular dynamics implementations presented in the literature, finding them as limit conditions. The accuracy of this new formulation is benchmarked by considering simple monoatomic systems described by the standard hybrid particle-field potentials. We find that by controlling the time step and grid size, conservation of energy and momenta, as well as disappearance of alias, is obtained. Increasing the particle-field interaction length scale permits the use of larger time steps and coarser grids. This promotes the use of multiple time step strategies over the quasi-instantaneous approximation, which is found to not conserve energy and momenta equally well. Finally, our investigations of the structural and dynamic properties of simple monoatomic systems show a consistent behavior between the present formulation and Gaussian Core models. △ Less

Submitted 3 July, 2020; originally announced July 2020.

Can Polarity-Inverted Surfactants Self-Assemble in Nonpolar Solvents

Authors: M. Carrer, T. Skrbic, S. L. Bore, G. Milano, M. Cascella, A. Giacometti

Abstract: We investigate the self-assembly process of a surfactant with inverted polarity in water and cyclohexane using both all-atom and coarse grained hybrid particle-field molecular dynamics simulations. Unlike conventional surfactants, the molecule under study, proposed in a recent experiment, is formed by a rigid and compact hydrophobic adamantane moiety, and a long and floppy triethylene glycol tail.… ▽ More We investigate the self-assembly process of a surfactant with inverted polarity in water and cyclohexane using both all-atom and coarse grained hybrid particle-field molecular dynamics simulations. Unlike conventional surfactants, the molecule under study, proposed in a recent experiment, is formed by a rigid and compact hydrophobic adamantane moiety, and a long and floppy triethylene glycol tail. In water, we report the formation of stable inverted micelles with the adamantane heads grouping together into a hydrophobic core, and the tails forming hydrogen bonds with water. By contrast, microsecond simulations do not provide evidence of stable micelle formation in cyclohexane. Validating the computational results by comparison with experimental diffusion constant and small-angle X-ray scattering intensity, we show that at laboratory thermodynamic conditions the mixture resides in the supercritical region of the phase diagram, where aggregated and free surfactant states co-exist in solution. Our simulations also provide indications about how to escape this region, to produce thermodynamically stable micellar aggregates. △ Less

Submitted 2 July, 2020; originally announced July 2020.

Comments: 14 pages, 10 Figures, accepted for publication (2020)

Automated Determination of Hybrid Particle-Field Parameters by Machine Learning

Authors: Morten Ledum, Sigbjørn Løland Bore, Michele Cascella

Abstract: The hybrid particle-field molecular dynamics method is an efficient alternative to standard particle-based coarse grained approaches. In this work, we propose an automated protocol for optimisation of the effective parameters that define interaction energy density functional, based on Bayesian optimization. The machine-learning protocol makes use of an arbitrary fitness function defined upon a set… ▽ More The hybrid particle-field molecular dynamics method is an efficient alternative to standard particle-based coarse grained approaches. In this work, we propose an automated protocol for optimisation of the effective parameters that define interaction energy density functional, based on Bayesian optimization. The machine-learning protocol makes use of an arbitrary fitness function defined upon a set of observables of relevance, which are optimally matched by an iterative process. Employing phospholipid bilayers as test systems, we demonstrate that the parameters obtained through our protocol are able to reproduce reference data better than currently employed sets derived by Flory-Huggins models. The optimisation procedure is robust and yields physically sound values. Moreover, we show that the parameters are satisfactorily transferable among chemically analogous species. Our protocol is general, and does not require heuristic a posteriori rebalancing. Therefore it is particularly suited for optimisation of reliable hybrid particle-field potentials of complex chemical mixtures, and extends the applicability corresponding simulations to all those systems for which calibration of the density functionals may not be done via simple theoretical models. △ Less

Submitted 4 April, 2020; originally announced April 2020.

Hybrid Particle-Field Molecular Dynamics Under Constant Pressure

Authors: Sigbjørn Løland Bore, Hima Bindu Kolli, Antonio De Nicola, Maksym Byshkin, Toshihiro Kawakatsu, Giuseppe Milano, Michele Cascella

Abstract: Hybrid particle-field methods are computationally efficient approaches for modelling soft matter systems. So far applications of these methodologies have been limited to constant volume conditions. Here, we reformulate particle-field interactions to represent systems coupled to constant external pressure. First, we show that the commonly used particle-field energy functional can be modified to mod… ▽ More Hybrid particle-field methods are computationally efficient approaches for modelling soft matter systems. So far applications of these methodologies have been limited to constant volume conditions. Here, we reformulate particle-field interactions to represent systems coupled to constant external pressure. First, we show that the commonly used particle-field energy functional can be modified to model and parameterize the isotropic contributions to the pressure tensor without interfering with the microscopic forces on the particles. Second, we employ a square gradient particle-field interaction term to model non-isotropic contributions to the pressure tensor, such as in surface tension phenomena. This formulation is implemented within the hybrid particle-field molecular dynamics approach and is tested on a series of model systems. Simulations of a homogeneous water box demonstrate that it is possible to parameterize the equation of state to reproduce any target density for a given external pressure. Moreover, the same parameterization is transferable to systems of similar coarse-grained mapping resolution. Finally, we evaluate the feasibility of the proposed approach on coarse-grained models of phospholipids, finding that the term between water and the lipid hydrocarbon tails is alone sufficient to reproduce the experimental area per lipid in constant-pressure simulations, and to produce a qualitatively correct lateral pressure profile. △ Less

Submitted 12 March, 2020; originally announced March 2020.

Comments: 24 pages, 7 figures

Journal ref: J. Chem. Phys. 152, 184908 (2020)

Coupling spin to velocity: collective motion of Hamiltonian polar particles

Authors: Sigbjørn Løland Bore, Michael Schindler, Khanh-Dang Nguyen Thu Lam, Eric Bertin, Olivier Dauchot

Abstract: We propose a conservative two-dimensional particle model in which particles carry a continuous and classical spin. The model includes standard ferromagnetic interactions between spins of two different particles, and a nonstandard coupling between spin and velocity of the same particle inspired by the coupling observed in self-propelled hard discs. Because of this coupling Galilean invariance is br… ▽ More We propose a conservative two-dimensional particle model in which particles carry a continuous and classical spin. The model includes standard ferromagnetic interactions between spins of two different particles, and a nonstandard coupling between spin and velocity of the same particle inspired by the coupling observed in self-propelled hard discs. Because of this coupling Galilean invariance is broken and the conserved linear momentum associated to translation invariance is not proportional to the velocity of the center of mass. Also, the dynamics is not invariant under a global rotation of the spins alone. This, in principle, leaves room for collective motion and thus raises the question whether collective motion can arise in Hamiltonian systems. We study the statistical mechanics of such a system, and show that, in the fully connected (or mean-field) case, a transition to collective motion does exist in spite of momentum conservation. Interestingly, the velocity of the center of mass, which in the absence of Galilean invariance, is a relevant variable, also feeds back on the magnetization properties, as it acts as an external magnetic field that smoothens the transition. Molecular dynamics simulations of finite size systems indeed reveal a rich phase diagram, with a transition from a disordered to a homogeneous polar phase, but also more complex inhomogeneous phases with local order interrupted by topological defects. △ Less

Submitted 24 March, 2016; v1 submitted 15 December, 2015; originally announced December 2015.

Journal ref: J. Stat. Mech. (2016) 033305