Machine learning prediction of accurate atomization energies of organic molecules from low-fidelity quantum chemical calculations

Ward, Logan; Blaiszik, Ben; Foster, Ian; Assary, Rajeev S.; Narayanan, Badri; Curtiss, Larry

doi:10.1557/mrc.2019.107

Title: Machine learning prediction of accurate atomization energies of organic molecules from low-fidelity quantum chemical calculations

Abstract

Recent studies illustrate how machine learning (ML) can be used to bypass a core challenge of molecular modeling: the trade-off between accuracy and computational cost. Here, we assess multiple ML approaches for predicting the atomization energy of organic molecules. Our resulting models learn the difference between low-fidelity, B3LYP, and high-accuracy, G4MP2, atomization energies and predict the G4MP2 atomization energy to 0.005 eV (mean absolute error) for molecules with less than nine heavy atoms (training set of 117,232 entries, test set 13,026) and 0.012 eV for a small set of 66 molecules with between 10 and 14 heavy atoms. As a result, our two best models, which have different accuracy/speed trade-offs, enable the efficient prediction of G4MP2-level energies for large molecules and are available through a simple web interface.

Authors:

^[1];

^[1]; Assary, Rajeev S. ^[2]; Narayanan, Badri ^[3]; Curtiss, Larry ^[2]

Argonne National Lab. (ANL), Lemont, IL (United States); Univ. of Chicago, Chicago, IL (United States)
Argonne National Lab. (ANL), Lemont, IL (United States)
Argonne National Lab. (ANL), Lemont, IL (United States); Univ. of Louisville, Louisville, KY (United States)

Publication Date:: 2019-08-27

Research Org.:: Argonne National Lab. (ANL), Argonne, IL (United States)

Sponsoring Org.:: USDOE Office of Science (SC), Basic Energy Sciences (BES); USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR)

OSTI Identifier:: 1578167

Grant/Contract Number:: AC02-06CH11357

Resource Type:: Accepted Manuscript

Journal Name:: MRS Communications

Additional Journal Information:: Journal Volume: 9; Journal Issue: 3; Journal ID: ISSN 2159-6859

Publisher:: Materials Research Society - Cambridge University Press

Country of Publication:: United States

Language:: English

Subject:: 37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CHEMISTRY

Citation Formats


                    Ward, Logan, Blaiszik, Ben, Foster, Ian, Assary, Rajeev S., Narayanan, Badri, and Curtiss, Larry. Machine learning prediction of accurate atomization energies of organic molecules from low-fidelity quantum chemical calculations.  United States: N. p., 2019. 
Web.  doi:10.1557/mrc.2019.107.

Copy to clipboard


                    Ward, Logan, Blaiszik, Ben, Foster, Ian, Assary, Rajeev S., Narayanan, Badri, & Curtiss, Larry. Machine learning prediction of accurate atomization energies of organic molecules from low-fidelity quantum chemical calculations.  United States.  https://doi.org/10.1557/mrc.2019.107

Copy to clipboard


                    Ward, Logan, Blaiszik, Ben, Foster, Ian, Assary, Rajeev S., Narayanan, Badri, and Curtiss, Larry. Tue .  
"Machine learning prediction of accurate atomization energies of organic molecules from low-fidelity quantum chemical calculations".  United States.  https://doi.org/10.1557/mrc.2019.107.  https://www.osti.gov/servlets/purl/1578167.

Copy to clipboard


                    
@article{osti_1578167,

  title        = {Machine learning prediction of accurate atomization energies of organic molecules from low-fidelity quantum chemical calculations},

  author       = {Ward, Logan and Blaiszik, Ben and Foster, Ian and Assary, Rajeev S. and Narayanan, Badri and Curtiss, Larry},

  abstractNote = {Recent studies illustrate how machine learning (ML) can be used to bypass a core challenge of molecular modeling: the trade-off between accuracy and computational cost. Here, we assess multiple ML approaches for predicting the atomization energy of organic molecules. Our resulting models learn the difference between low-fidelity, B3LYP, and high-accuracy, G4MP2, atomization energies and predict the G4MP2 atomization energy to 0.005 eV (mean absolute error) for molecules with less than nine heavy atoms (training set of 117,232 entries, test set 13,026) and 0.012 eV for a small set of 66 molecules with between 10 and 14 heavy atoms. As a result, our two best models, which have different accuracy/speed trade-offs, enable the efficient prediction of G4MP2-level energies for large molecules and are available through a simple web interface.},

  doi          = {10.1557/mrc.2019.107},

  journal      = {MRS Communications},

  number       = 3,

  volume       = 9,

  place        = {United States},

  year         = {2019},

  month        = {8}

}

Copy to clipboard

Journal Article:

Free Publicly Available Full Text

Accepted Manuscript (DOE)

Publisher's Version of Record

https://doi.org/10.1557/mrc.2019.107

Other availability

Search WorldCat to find libraries that may hold this journal

Citation Metrics:

Cited by: 5 works

Citation information provided by
Web of Science

Save / Share:

Export Metadata

Save to My Library

Works referenced in this record:

Boosting Quantum Machine Learning Models with a Multilevel Combination Technique: Pople Diagrams Revisited
journal, December 2018

Zaspel, Peter; Huang, Bing; Harbrecht, Helmut
Journal of Chemical Theory and Computation, Vol. 15, Issue 3
DOI: 10.1021/acs.jctc.8b00832

The Materials Data Facility: Data Services to Advance Materials Science Research
journal, July 2016

Blaiszik, B.; Chard, K.; Pruyne, J.
JOM, Vol. 68, Issue 8
DOI: 10.1007/s11837-016-2001-3

Molecular graph convolutions: moving beyond fingerprints
journal, August 2016

Kearnes, Steven; McCloskey, Kevin; Berndl, Marc
Journal of Computer-Aided Molecular Design, Vol. 30, Issue 8
DOI: 10.1007/s10822-016-9938-8

MoleculeNet: a benchmark for molecular machine learning
journal, January 2018

Wu, Zhenqin; Ramsundar, Bharath; Feinberg, Evan N.
Chemical Science, Vol. 9, Issue 2
DOI: 10.1039/C7SC02664A

A new mixing of Hartree–Fock and local density‐functional theories
journal, January 1993

Becke, Axel D.
The Journal of Chemical Physics, Vol. 98, Issue 2
DOI: 10.1063/1.464304

XSEDE: Accelerating Scientific Discovery
journal, September 2014

Towns, John; Cockerill, Timothy; Dahan, Maytal
Computing in Science & Engineering, Vol. 16, Issue 5
DOI: 10.1109/MCSE.2014.80

SchNet – A deep learning architecture for molecules and materials
journal, June 2018

Schütt, K. T.; Sauceda, H. E.; Kindermans, P. -J.
The Journal of Chemical Physics, Vol. 148, Issue 24
DOI: 10.1063/1.5019779

The TensorMol-0.1 model chemistry: a neural network augmented with long-range physics
journal, January 2018

Yao, Kun; Herr, John E.; Toth, David W.
Chemical Science, Vol. 9, Issue 8
DOI: 10.1039/C7SC04934J

Big Data Meets Quantum Chemistry Approximations: The Δ-Machine Learning Approach
journal, April 2015

Ramakrishnan, Raghunathan; Dral, Pavlo O.; Rupp, Matthias
Journal of Chemical Theory and Computation, Vol. 11, Issue 5
DOI: 10.1021/acs.jctc.5b00099

Gaussian-4 theory using reduced order perturbation theory
journal, September 2007

Curtiss, Larry A.; Redfern, Paul C.; Raghavachari, Krishnan
The Journal of Chemical Physics, Vol. 127, Issue 12
DOI: 10.1063/1.2770701

Thirty years of density functional theory in computational chemistry: an overview and extensive assessment of 200 density functionals
journal, April 2017

Mardirossian, Narbe; Head-Gordon, Martin
Molecular Physics, Vol. 115, Issue 19
DOI: 10.1080/00268976.2017.1333644

Merck molecular force field. I. Basis, form, scope, parameterization, and performance of MMFF94
journal, April 1996

Halgren, Thomas A.
Journal of Computational Chemistry, Vol. 17, Issue 5-6
DOI: 10.1002/(SICI)1096-987X(199604)17:5/6<490::AID-JCC1>3.0.CO;2-P

Quantum chemistry structures and properties of 134 kilo molecules
journal, August 2014

Ramakrishnan, Raghunathan; Dral, Pavlo O.; Rupp, Matthias
Scientific Data, Vol. 1, Issue 1
DOI: 10.1038/sdata.2014.22

G n theory : G
journal, June 2011

Curtiss, Larry A.; Redfern, Paul C.; Raghavachari, Krishnan
Wiley Interdisciplinary Reviews: Computational Molecular Science, Vol. 1, Issue 5
DOI: 10.1002/wcms.59

Next generation interatomic potentials for condensed systems
journal, July 2014

Handley, Christopher Michael; Behler, Jörg
The European Physical Journal B, Vol. 87, Issue 7
DOI: 10.1140/epjb/e2014-50070-0

Multi-fidelity machine learning models for accurate bandgap predictions of solids
journal, March 2017

Pilania, G.; Gubernatis, J. E.; Lookman, T.
Computational Materials Science, Vol. 129
DOI: 10.1016/j.commatsci.2016.12.004

Open Babel: An open chemical toolbox
journal, October 2011

O'Boyle, Noel M.; Banck, Michael; James, Craig A.
Journal of Cheminformatics, Vol. 3, Issue 1
DOI: 10.1186/1758-2946-3-33

Atomistic calculations and materials informatics: A review
journal, June 2017

Ward, Logan; Wolverton, Chris
Current Opinion in Solid State and Materials Science, Vol. 21, Issue 3
DOI: 10.1016/j.cossms.2016.07.002

Machine learning for quantum mechanics in a nutshell
journal, July 2015

Rupp, Matthias
International Journal of Quantum Chemistry, Vol. 115, Issue 16
DOI: 10.1002/qua.24954

Machine learning with systematic density-functional theory calculations: Application to melting temperatures of single- and binary-component solids
journal, February 2014

Seko, Atsuto; Maekawa, Tomoya; Tsuda, Koji
Physical Review B, Vol. 89, Issue 5
DOI: 10.1103/PhysRevB.89.054303

Alchemical and structural distribution based representation for universal quantum machine learning
journal, June 2018

Faber, Felix A.; Christensen, Anders S.; Huang, Bing
The Journal of Chemical Physics, Vol. 148, Issue 24
DOI: 10.1063/1.5020710

Accurate quantum chemical energies for 133 000 organic molecules
journal, January 2019

Narayanan, Badri; Redfern, Paul C.; Assary, Rajeev S.
Chemical Science, Vol. 10, Issue 31
DOI: 10.1039/C9SC02834J

SchNetPack: A Deep Learning Toolbox For Atomistic Systems
journal, November 2018

Schütt, K. T.; Kessel, P.; Gastegger, M.
Journal of Chemical Theory and Computation, Vol. 15, Issue 1
DOI: 10.1021/acs.jctc.8b00908

Genetic Optimization of Training Sets for Improved Machine Learning Models of Molecular Properties
journal, March 2017

Browning, Nicholas J.; Ramakrishnan, Raghunathan; von Lilienfeld, O. Anatole
The Journal of Physical Chemistry Letters, Vol. 8, Issue 7
DOI: 10.1021/acs.jpclett.7b00038

Predicting molecular properties with covariant compositional networks
journal, June 2018

Hy, Truong Son; Trivedi, Shubhendu; Pan, Horace
The Journal of Chemical Physics, Vol. 148, Issue 24
DOI: 10.1063/1.5024797

A graph-convolutional neural network model for the prediction of chemical reactivity
journal, January 2019

Coley, Connor W.; Jin, Wengong; Rogers, Luke
Chemical Science, Vol. 10, Issue 2
DOI: 10.1039/C8SC04228D

Genetic Optimization of Training Sets for Improved Machine Learning Models of Molecular Properties
text, January 2017

Browning, Nicholas J.; Ramakrishnan, Raghunathan; Von Lilienfeld, O. Anatole
American Chemical Society
DOI: 10.5451/unibas-ep63254

Big Data Meets Quantum Chemistry Approximations: The Δ-Machine Learning Approach
text, January 2015

Ramakrishnan, Raghunathan; Dral, Pavlo O.; Rupp, Matthias
American Chemical Society
DOI: 10.5451/unibas-ep43348

Machine learning with systematic density-functional theory calculations: Application to melting temperatures of single and binary component solids
text, January 2013

Seko, Atsuto; Maekawa, Tomoya; Tsuda, Koji
arXiv
DOI: 10.48550/arxiv.1310.1546

MoleculeNet: A Benchmark for Molecular Machine Learning
preprint, January 2017

Wu, Zhenqin; Ramsundar, Bharath; Feinberg, Evan N.
arXiv
DOI: 10.48550/arxiv.1703.00564

The TensorMol-0.1 Model Chemistry: a Neural Network Augmented with Long-Range Physics
preprint, January 2017

Yao, Kun; Herr, John E.; Toth, David W.
arXiv
DOI: 10.48550/arxiv.1711.06385

SchNet - a deep learning architecture for molecules and materials
text, January 2017

Schütt, Kristof T.; Sauceda, Huziel E.; Kindermans, Pieter-Jan
arXiv
DOI: 10.48550/arxiv.1712.06113

Works referencing / citing this record:

A data ecosystem to support machine learning in materials science
journal, October 2019

Blaiszik, Ben; Ward, Logan; Schwarting, Marcus
MRS Communications, Vol. 9, Issue 4
DOI: 10.1557/mrc.2019.118

Similar Records in DOE PAGES and OSTI.GOV collections:

Quantum-Chemically Informed Machine Learning: Prediction of Energies of Organic Molecules with 10 to 14 Non-hydrogen Atoms

Journal Article Dandu, Naveen ; Ward, Logan ; Assary, Rajeev S. ; ... - Journal of Physical Chemistry. A, Molecules, Spectroscopy, Kinetics, Environment, and General Theory

High-fidelity quantum-chemical calculations can provide accurate predictions of molecular energies, but their high computational costs limit their utility, especially for larger molecules. We have shown in previous work that machine learning models trained on high-level quantum-chemical calculations (G4MP2) for organic molecules with one to nine non-hydrogen atoms can provide accurate predictions for other molecules of comparable size at much lower costs. Here we demonstrate that such models can also be used to effectively predict energies of molecules larger than those in the training set. To implement this strategy, we first established a set of 191 molecules with 10–14 non-hydrogen atomsmore »« less
https://doi.org/10.1021/acs.jpca.0c01777

Full Text Available
Accurate Quantum Chemical Energies for 133,000 Organic Molecules

Journal Article Narayanan, Badri ; Redfern, Paul C. ; Assary, Rajeev S. ; ... - Chemical Science

The energies of the 133 000 molecules in the GDB-9 database have been calculated at the G4MP2 level of theory and then were used to calculate their enthalpies of formation. This database contains organic molecules having nine or less atoms of carbon, nitrogen, oxygen, and fluorine, as well as hydrogen atoms. The accuracy of the G4MP2 energies was investigated on a subset of 459 of the molecules having experimental enthalpies of formation with small uncertainties. On this subset the G4MP2 enthalpies of formation have an accuracy of 0.79 kcal mol(-1), which is similar to its accuracy previously reported for themore »« less
https://doi.org/10.1039/c9sc02834j
Multifidelity Information Fusion with Machine Learning: A Case Study of Dopant Formation Energies in Hafnia

Journal Article Batra, Rohit ; Pilania, Ghanshyam ; Uberuaga, Blas P. ; ... - ACS Applied Materials and Interfaces

Cost versus accuracy trade-offs are frequently encountered in materials science and engineering, where a particular property of interest can be measured/computed at different levels of accuracy or fidelity. Naturally, the most accurate measurement is also the most resource and time intensive, while the inexpensive quicker alternatives tend to be noisy. In such situations, a number of machine learning (ML) based multifidelity information fusion (MFIF) strategies can be employed to fuse information accessible from varying sources of fidelity and make predictions at the highest level of accuracy. In this work, we perform a comparative study on traditionally employed single-fidelity and threemore »« less
Cited by 8
https://doi.org/10.1021/acsami.9b02174

Full Text Available
General Protocol for the Accurate Prediction of Molecular ¹³C/¹H NMR Chemical Shifts via Machine Learning Augmented DFT

Journal Article Gao, Peng ; Zhang, Jun ; Peng, Qian ; ... - Journal of Chemical Information and Modeling

An accurate prediction of NMR chemical shifts at affordable computational cost is very important for different types of structural assignments in experimental studies. Density functional theory (DFT) and gauge-including atomic orbital (GIAO) are two of the most popular computational methods for NMR calculation, yet they often fail to resolve ambiguities in structural assignments. In this work, we present a new method that uses machine learning (ML) techniques (DFT + ML) that significantly increases the accuracy of ¹³C/¹H NMR chemical shift prediction for a variety of organic molecules. The input of the generalizable DFT + ML model contains two critical parts:more »« less
https://doi.org/10.1021/acs.jcim.0c00388

Full Text Available
Charge-Dependent Cavity Radii for an Accurate Dielectric Continuum Model of Solvation with Emphasis on Ions: Aqueous Solutes with Oxo, Hydroxo, Amino, Methyl, Chloro, Bromo and Fluoro Functionalities

Journal Article Ginovska, Bojana ; Camaioni, Donald M ; Dupuis, Michel ; ... - Journal of Physical Chemistry A, 112(42):10604-10613

Dielectric continuum solvation models are widely used because they are a computationally efficacious way to simulate equilibrium properties of solutes. With advances that allow for molecular-shaped cavities, they have reached a high level of accuracy, in particular for neutral solutes. However, benchmark tests show that existing schemes for defining cavities are unable to consistently predict accurately the effects of solvation on ions, especially anions. This work involves the further development of a protocol put forth earlier for defining the cavities of aqueous solutes, with resulting advances that are most striking for anions. Molecular cavities are defined as interlocked spheres aroundmore »« less
https://doi.org/10.1021/jp804092v

Similar Records

Title: Machine learning prediction of accurate atomization energies of organic molecules from low-fidelity quantum chemical calculations

Abstract

Citation Formats

Boosting Quantum Machine Learning Models with a Multilevel Combination Technique: Pople Diagrams Revisited journal, December 2018

The Materials Data Facility: Data Services to Advance Materials Science Research journal, July 2016

Molecular graph convolutions: moving beyond fingerprints journal, August 2016

MoleculeNet: a benchmark for molecular machine learning journal, January 2018

A new mixing of Hartree–Fock and local density‐functional theories journal, January 1993

XSEDE: Accelerating Scientific Discovery journal, September 2014

SchNet – A deep learning architecture for molecules and materials journal, June 2018

The TensorMol-0.1 model chemistry: a neural network augmented with long-range physics journal, January 2018

Big Data Meets Quantum Chemistry Approximations: The Δ-Machine Learning Approach journal, April 2015

Gaussian-4 theory using reduced order perturbation theory journal, September 2007

Thirty years of density functional theory in computational chemistry: an overview and extensive assessment of 200 density functionals journal, April 2017

Merck molecular force field. I. Basis, form, scope, parameterization, and performance of MMFF94 journal, April 1996

Quantum chemistry structures and properties of 134 kilo molecules journal, August 2014

G n theory : G journal, June 2011

Next generation interatomic potentials for condensed systems journal, July 2014

Multi-fidelity machine learning models for accurate bandgap predictions of solids journal, March 2017

Open Babel: An open chemical toolbox journal, October 2011

Atomistic calculations and materials informatics: A review journal, June 2017

Machine learning for quantum mechanics in a nutshell journal, July 2015

Machine learning with systematic density-functional theory calculations: Application to melting temperatures of single- and binary-component solids journal, February 2014

Alchemical and structural distribution based representation for universal quantum machine learning journal, June 2018

Accurate quantum chemical energies for 133 000 organic molecules journal, January 2019

SchNetPack: A Deep Learning Toolbox For Atomistic Systems journal, November 2018

Genetic Optimization of Training Sets for Improved Machine Learning Models of Molecular Properties journal, March 2017

Predicting molecular properties with covariant compositional networks journal, June 2018

A graph-convolutional neural network model for the prediction of chemical reactivity journal, January 2019

Genetic Optimization of Training Sets for Improved Machine Learning Models of Molecular Properties text, January 2017

Big Data Meets Quantum Chemistry Approximations: The Δ-Machine Learning Approach text, January 2015

Machine learning with systematic density-functional theory calculations: Application to melting temperatures of single and binary component solids text, January 2013

MoleculeNet: A Benchmark for Molecular Machine Learning preprint, January 2017

The TensorMol-0.1 Model Chemistry: a Neural Network Augmented with Long-Range Physics preprint, January 2017

SchNet - a deep learning architecture for molecules and materials text, January 2017

A data ecosystem to support machine learning in materials science journal, October 2019

Boosting Quantum Machine Learning Models with a Multilevel Combination Technique: Pople Diagrams Revisited
journal, December 2018

The Materials Data Facility: Data Services to Advance Materials Science Research
journal, July 2016

Molecular graph convolutions: moving beyond fingerprints
journal, August 2016

MoleculeNet: a benchmark for molecular machine learning
journal, January 2018

A new mixing of Hartree–Fock and local density‐functional theories
journal, January 1993

XSEDE: Accelerating Scientific Discovery
journal, September 2014

SchNet – A deep learning architecture for molecules and materials
journal, June 2018

The TensorMol-0.1 model chemistry: a neural network augmented with long-range physics
journal, January 2018

Big Data Meets Quantum Chemistry Approximations: The Δ-Machine Learning Approach
journal, April 2015

Gaussian-4 theory using reduced order perturbation theory
journal, September 2007

Thirty years of density functional theory in computational chemistry: an overview and extensive assessment of 200 density functionals
journal, April 2017

Merck molecular force field. I. Basis, form, scope, parameterization, and performance of MMFF94
journal, April 1996

Quantum chemistry structures and properties of 134 kilo molecules
journal, August 2014

G n theory : G
journal, June 2011

Next generation interatomic potentials for condensed systems
journal, July 2014

Multi-fidelity machine learning models for accurate bandgap predictions of solids
journal, March 2017

Open Babel: An open chemical toolbox
journal, October 2011

Atomistic calculations and materials informatics: A review
journal, June 2017

Machine learning for quantum mechanics in a nutshell
journal, July 2015

Machine learning with systematic density-functional theory calculations: Application to melting temperatures of single- and binary-component solids
journal, February 2014

Alchemical and structural distribution based representation for universal quantum machine learning
journal, June 2018

Accurate quantum chemical energies for 133 000 organic molecules
journal, January 2019

SchNetPack: A Deep Learning Toolbox For Atomistic Systems
journal, November 2018

Genetic Optimization of Training Sets for Improved Machine Learning Models of Molecular Properties
journal, March 2017

Predicting molecular properties with covariant compositional networks
journal, June 2018

A graph-convolutional neural network model for the prediction of chemical reactivity
journal, January 2019

Genetic Optimization of Training Sets for Improved Machine Learning Models of Molecular Properties
text, January 2017

Big Data Meets Quantum Chemistry Approximations: The Δ-Machine Learning Approach
text, January 2015

Machine learning with systematic density-functional theory calculations: Application to melting temperatures of single and binary component solids
text, January 2013

MoleculeNet: A Benchmark for Molecular Machine Learning
preprint, January 2017

The TensorMol-0.1 Model Chemistry: a Neural Network Augmented with Long-Range Physics
preprint, January 2017

SchNet - a deep learning architecture for molecules and materials
text, January 2017

A data ecosystem to support machine learning in materials science
journal, October 2019