Generalized Neural-Network Representation of High-Dimensional Potential-Energy Surfaces

Behler, Jörg; Parrinello, Michele

doi:10.1103/physrevlett.98.146401

Cited by 3,442 publications

(3,869 citation statements)

References 17 publications

Supporting

Mentioning

3,814

Contrasting

Unclassified

Order By: Relevance

“…' [97,98] More sophisticated statistical learning methods have been applied to the training of exchange correlation functionals in DFT, [99,100] or to parameterizing interatomic force fields. [101][102][103][104][105][106] Support vector machines have been shown to quantify basis-set incompleteness. [107] Gaussian kernel-based ML for the design of accurate and reactive forcefields without predetermined functional form was introduced by Bartok et al [108] Contributions by Curtarolo, Hautier, and Ceder combine data-mining with mean-field electronic structure theory.…”

Section: Methodsmentioning

confidence: 99%

First principles view on chemical compound space: Gaining rigorous atomistic control of molecular properties

Lilienfeld

2013

Int J of Quantum Chemistry

133

160

View full text Add to dashboard Cite

A well‐defined notion of chemical compound space (CCS) is essential for gaining rigorous control of properties through variation of elemental composition and atomic configurations. Here, we give an introduction to an atomistic first principles perspective on CCS. First, CCS is discussed in terms of variational nuclear charges in the context of conceptual density functional and molecular grand‐canonical ensemble theory. Thereafter, we revisit the notion of compound pairs, related to each other via “alchemical” interpolations involving fractional nuclear charges in the electronic Hamiltonian. We address Taylor expansions in CCS, property nonlinearity, improved predictions using reference compound pairs, and the ounce‐of‐gold prize challenge to linearize CCS. Finally, we turn to machine learning of analytical structure property relationships in CCS. These relationships correspond to inferred, rather than derived through variational principle, solutions of the electronic Schrödinger equation. © 2013 Wiley Periodicals, Inc.

show abstract

Section: Methodsmentioning

confidence: 99%

First principles view on chemical compound space: Gaining rigorous atomistic control of molecular properties

Lilienfeld

2013

Int J of Quantum Chemistry

133

160

View full text Add to dashboard Cite

show abstract

“…MLIPs have been applied to a wide range of materials, regardless of chemical bonding nature of the materials. Recently, frameworks applicable to periodic systems have been proposed [9][10][11].…”

Section: Construction Of Mlip For Elemental Metalsmentioning

confidence: 99%

Descriptors for Machine Learning of Materials Data

2018

View full text Add to dashboard Cite

Descriptors, which are representations of compounds, play an essential role in machine learning of materials data. Although many representations of elements and structures of compounds are known, these representations are difficult to use as descriptors in their unchanged forms. This chapter shows how compounds in a dataset can be represented as descriptors and applied to machine-learning models for materials datasets.Keywords Machine-learning interatomic potential ⋅ Lattice thermal conductivity ⋅ Recommender system ⋅ Gaussian process ⋅ Bayesian optimization IntroductionRecent developments of data-centric approaches should accelerate the progress in materials science dramatically. Thanks to the recent advances in computational power and techniques, the results from numerous density functional theory (DFT) calculations with predictive performances have been stored as databases. A combination of such databases and an efficient machine-learning approach should realize prediction and classification models of target physical properties. Consequently, machine-learning techniques are becoming ubiquitous. They are used to explore materials and structures from a huge number of candidates and to extract meaningful information and patterns from existing data.A key factor in controlling the performance of a machine-learning approach is how compounds are represented in a data set. Representations of compounds are called "descriptors" or "features". To perform machine-learning modeling, available descriptors must be determined according to the evaluation cost of the target property

show abstract

“…[11] Trained to reference datasets, ML models can predict energies, forces, and other molecular properties. [12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27] They have been 3 used to discover materials [28][29][30][31][32][33][34][35][36][37] and study dynamical processes such as charge and exciton transfer. [38][39][40][41] Most related to this work are ML models of existing charge models, [9,[42][43][44] which are orders of magnitude faster than ab initio calculation.…”

Section: Mskmentioning

confidence: 99%

Discovering a Transferable Charge Assignment Model Using Machine Learning

Sifain¹,

Lubbers²,

Nebgen³

et al. 2018

Preprint

View full text Add to dashboard Cite

Partial atomic charge assignment is of immense practical value to force field parametrization, molecular docking, and cheminformatics. Machine learning has emerged as a powerful tool for modeling chemistry at unprecedented computational speeds given ground-truth values, but for the task of charge assignment, the choice of ground-truth may not be obvious. In this letter, we use machine learning to discover a charge model by training a neural network to molecular dipole moments using a large, diverse set of CHNO molecular conformations. The new model, called Affordable Charge Assignment (ACA), is computationally inexpensive and predicts dipoles of out-of-sample molecules accurately. Furthermore, dipole-inferred ACA charges are transferable to dipole and even quadrupole moments of much larger molecules than those used for training. We apply ACA to long dynamical trajectories of biomolecules and successfully produce their infrared spectra. Additionally, we compare ACA with existing charge models and find that ACA assigns similar charges to Charge Model 5, but with a greatly reduced computational cost. Graphical TOC Entry Extensibility Tests Molecular Size Training DatasetKeywords machine learning, neural networks, quantum chemisty 2 Electrostatic interactions contribute strongly to the forces within and between molecules.These interactions depend on the charge density field ⇢(r), which is computationally demanding to compute. Simplified models of the charge density, such as atom-centered monopoles, are commonly employed. These partial atomic charges result in faster computation as well as provide a qualitative understanding of the underlying chemistry. [1][2][3][4] However, the decomposition of charge density into atomic charges is, by itself, an ambiguous task. Additional principles are necessary to make the charge assignment task well-defined. Here we show that a Machine Learning model, trained only on the dipole moments of small molecules, discovers a charge model that is transferable to quadrupole predictions and extensible to much larger molecules.Existing popular charge models have also been designed to reproduce observables of the electrostatic potential. The Merz-Singh-Kollman (MSK) [5,6] charge model exactly replicates the dipole moment and approximates the electrostatic potential on many points surrounding the molecule, resulting in high-quality electrostatic properties exterior to the molecule. However, MSK suffers from basis set sensitivity, particularly for "buried atoms" located inside large molecules. [7][8][9] Charge model 5 (CM5) [8] is an extension of Hirshfeld analysis, [10] with additional parametrization in order to approximately reproduce ab initio and experimental dipoles of 614 gas-phase dipoles. Unlike MSK, Hirshfeld and CM5 are nearly independent of basis set. [9] This insensitivity allows CM5 to use a single set of model parameters. The corresponding tradeoff is that its charges do not reproduce electrostatic fields as well as MSK.A limitation of these conventional charge models is...

show abstract

Generalized Neural-Network Representation of High-Dimensional Potential-Energy Surfaces

Cited by 3,442 publications

References 17 publications

First principles view on chemical compound space: Gaining rigorous atomistic control of molecular properties

First principles view on chemical compound space: Gaining rigorous atomistic control of molecular properties

Descriptors for Machine Learning of Materials Data

Discovering a Transferable Charge Assignment Model Using Machine Learning

Contact Info

Product

Resources

About