SchNetPack 2.0: A neural network toolbox for atomistic machine learning

Schütt, Kristof T.; Hessmann, Stefaan S. P.; Gebauer, Niklas W. A.; Lederer, Jonas; Gastegger, Michael

doi:10.1063/5.0138367

Cited by 20 publications

(20 citation statements)

References 66 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With this single-point energy approach, fewer structures need to be taken to computationally demanding DFT calculation as opposed to filtering straight on the semiempirical energy ordering. With very large diverse databases or training more parameters (e.g., forces), we also recommend using other training data set reduction methods and/or using JKML coupled with SchNetPack , for training a NN potential.…”

Section: Application and Discussionmentioning

confidence: 99%

“…Our ML-oriented subpackage, JKML, offers an interface between the JKQC-constructed database files (e.g., those stored in ACDB 2.0) and two ML programs, quantum machine learning (QML) and SchNetPack. , In the procedure, XYZ coordinates are extracted and together with the property of interest (e.g., electronic energy, forces, or mobility) are stored in a database. Subsequently, JKML uses QML or SchNetPack to perform the training, validation, and testing of the predicted property or its difference from a reference state.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Computational Tools for Handling Molecular Clusters: Configurational Sampling, Storage, Analysis, and Machine Learning

Kubečka,

Besel,

Neefjes

et al. 2023

ACS Omega

View full text Add to dashboard Cite

Computational modeling of atmospheric molecular clusters requires a comprehensive understanding of their complex configurational spaces, interaction patterns, stabilities against fragmentation, and even dynamic behaviors. To address these needs, we introduce the Jammy Key framework, a collection of automated scripts that facilitate and streamline molecular cluster modeling workflows. Jammy Key handles file manipulations between varieties of integrated third-party programs. The framework is divided into three main functionalities: (1) Jammy Key for configurational sampling (JKCS) to perform systematic configurational sampling of molecular clusters, (2) Jammy Key for quantum chemistry (JKQC) to analyze commonly used quantum chemistry output files and facilitate database construction, handling, and analysis, and (3) Jammy Key for machine learning (JKML) to manage machine learning methods in optimizing molecular cluster modeling. This automation and machine learning utilization significantly reduces manual labor, greatly speeds up the search for molecular cluster configurations, and thus increases the number of systems that can be studied. Following the example of the Atmospheric Cluster Database (ACDB) of Elm (ACS Omega, 4, 10965–10984, 2019), the molecular clusters modeled in our group using the Jammy Key framework have been stored in an improved online GitHub repository named ACDB 2.0. In this work, we present the Jammy Key package alongside its assorted applications, which underline its versatility. Using several illustrative examples, we discuss how to choose appropriate combinations of methodologies for treating particular cluster types, including reactive, multicomponent, charged, or radical clusters, as well as clusters containing flexible or multiconformer monomers or heavy atoms. Finally, we present a detailed example of using the tools for atmospheric acid–base clusters.

show abstract

Section: Application and Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Computational Tools for Handling Molecular Clusters: Configurational Sampling, Storage, Analysis, and Machine Learning

Kubečka,

Besel,

Neefjes

et al. 2023

ACS Omega

View full text Add to dashboard Cite

show abstract

“…After several iterations, the features encode the relevant information about the chemical environment of each atom. In this work, we use SchNet 7,9,28 to construct atomic feature representations. In general, however, MoINN is applicable to any other representation learning scheme.…”

Section: Methodsmentioning

confidence: 99%

“…In this context, machine learning (ML) methods have become increasingly popular as a means to circumvent costly quantum mechanical calculations. 1–37 One class of such ML methods are message passing neural networks (MPNNs), 38 which provide molecular property predictions based on end-to-end learned representations of atomic environments.…”

Section: Introductionmentioning

confidence: 99%

Automatic identification of chemical moieties

Lederer,

Gastegger,

Schütt

et al. 2023

Phys. Chem. Chem. Phys.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Other settings are provided in Section . SchNet was accessed via the SchNetPack v1.0.0 () master branch and PaiNN through the same repository, but through the developmental (“dev”) branch, which is currently the base for SchNetPack v2.0 …”

Section: Methodsmentioning

confidence: 99%

Machine Learning Interatomic Potentials for Reactive Hydrogen Dynamics at Metal Surfaces Based on Iterative Refinement of Reaction Probabilities

Stark,

Westermayr,

Douglas-Gallardo

et al. 2023

J. Phys. Chem. C

View full text Add to dashboard Cite

The reactive chemistry of molecular hydrogen at surfaces, notably dissociative sticking and hydrogen evolution, plays a crucial role in energy storage and fuel cells. Theoretical studies can help to decipher underlying mechanisms and reaction design, but studying dynamics at surfaces is computationally challenging due to the complex electronic structure at interfaces and the high sensitivity of dynamics to reaction barriers. In addition, ab initio molecular dynamics, based on density functional theory, is too computationally demanding to accurately predict reactive sticking or desorption probabilities, as it requires averaging over tens of thousands of initial conditions. High-dimensional machine learning-based interatomic potentials are starting to be more commonly used in gas-surface dynamics, yet robust approaches to generate reliable training data and assess how model uncertainty affects the prediction of dynamic observables are not well established. Here, we employ ensemble learning to adaptively generate training data while assessing model performance with full uncertainty quantification (UQ) for reaction probabilities of hydrogen scattering on different copper facets. We use this approach to investigate the performance of two message-passing neural networks, SchNet and PaiNN. Ensemble-based UQ and iterative refinement allow us to expose the shortcomings of the invariant pairwise-distance-based feature representation in the SchNet model for gas-surface dynamics.

show abstract

SchNetPack 2.0: A neural network toolbox for atomistic machine learning

Cited by 20 publications

References 66 publications

Computational Tools for Handling Molecular Clusters: Configurational Sampling, Storage, Analysis, and Machine Learning

Computational Tools for Handling Molecular Clusters: Configurational Sampling, Storage, Analysis, and Machine Learning

Automatic identification of chemical moieties

Machine Learning Interatomic Potentials for Reactive Hydrogen Dynamics at Metal Surfaces Based on Iterative Refinement of Reaction Probabilities

Contact Info

Product

Resources

About