Article (Scientific journals)
Machine learning predictions of molecular properties: Accurate many-body potentials and nonlocality in chemical space
Hansen, K.; Biegler, F.; Ramakrishnan, R. et al.
2015In Journal of Physical Chemistry Letters, 6 (12), p. 2326-2331
Peer Reviewed verified by ORBi
 

Files


Full Text
Interaction Potentials in Molecules and Non-Local Information in Chemical Space.pdf
Author postprint (421.87 kB)
Download

All documents in ORBilu are protected by a user license.

Send to



Details



Keywords :
Artificial intelligence; Atomization; Chemical compounds; Density functional theory; Electronic properties; Forecasting; Learning systems; Molecules; Atomization energies; Frontier orbital energies; Many-body interactions; Many-body potentials; Molecular properties; Pharmaceutical industry; Rational compound designs; Sophisticated machines; Chemical bonds
Abstract :
[en] Simultaneously accurate and efficient prediction of molecular properties throughout chemical compound space is a critical ingredient toward rational compound design in chemical and pharmaceutical industries. Aiming toward this goal, we develop and apply a systematic hierarchy of efficient empirical methods to estimate atomization and total energies of molecules. These methods range from a simple sum over atoms, to addition of bond energies, to pairwise interatomic force fields, reaching to the more sophisticated machine learning approaches that are capable of describing collective interactions between many atoms or bonds. In the case of equilibrium molecular geometries, even simple pairwise force fields demonstrate prediction accuracy comparable to benchmark energies calculated using density functional theory with hybrid exchange-correlation functionals; however, accounting for the collective many-body interactions proves to be essential for approaching the "holy grail" of chemical accuracy of 1 kcal/mol for both equilibrium and out-of-equilibrium geometries. This remarkable accuracy is achieved by a vectorized representation of molecules (so-called Bag of Bonds model) that exhibits strong nonlocality in chemical space. In addition, the same representation allows us to predict accurate electronic properties of molecules, such as their polarizability and molecular frontier orbital energies. © 2015 American Chemical Society.
Disciplines :
Physics
Identifiers :
eid=2-s2.0-84935014439
Author, co-author :
Hansen, K.;  Fritz-Haber-Institut der Max-Planck-Gesellschaft, Faradayweg 4-6, Berlin, Germany
Biegler, F.;  Machine Learning Group, Technical University of Berlin, Marchstr. 23, Berlin, Germany
Ramakrishnan, R.;  Department of Chemistry, National Center for Computational Design and Discovery of Novel Materials, University of Basel, Klingelbergstrasse 80, Basel, Switzerland
Pronobis, W.;  Machine Learning Group, Technical University of Berlin, Marchstr. 23, Berlin, Germany
Von Lilienfeld, O. A.;  Department of Chemistry, National Center for Computational Design and Discovery of Novel Materials, University of Basel, Klingelbergstrasse 80, Basel, Switzerland, Argonne Leadership Computing Facility, Argonne National Laboratory, 9700 South Cass Avenue, Argonne, IL, United States
Müller, K.-R.;  Machine Learning Group, Technical University of Berlin, Marchstr. 23, Berlin, Germany, Department of Brain and Cognitive Engineering, Korea University, Anam-dong, Seongbuk-ku, Seoul, South Korea
Tkatchenko, Alexandre ;  Fritz-Haber-Institut der Max-Planck-Gesellschaft, Faradayweg 4-6, Berlin, Germany
External co-authors :
yes
Language :
English
Title :
Machine learning predictions of molecular properties: Accurate many-body potentials and nonlocality in chemical space
Publication date :
2015
Journal title :
Journal of Physical Chemistry Letters
ISSN :
1948-7185
Publisher :
American Chemical Society
Volume :
6
Issue :
12
Pages :
2326-2331
Peer reviewed :
Peer Reviewed verified by ORBi
Funders :
MU 987/20, DFG, Natural Sciences and Engineering Research Council of Canada; ERC, Natural Sciences and Engineering Research Council of Canada; NSERC, Natural Sciences and Engineering Research Council of Canada; PP00P2-138932, SNSF, Natural Sciences and Engineering Research Council of Canada
Available on ORBilu :
since 15 March 2016

Statistics


Number of views
239 (12 by Unilu)
Number of downloads
173 (7 by Unilu)

Scopus citations®
 
542
Scopus citations®
without self-citations
484
OpenCitations
 
493
WoS citations
 
529

Bibliography


Similar publications



Contact ORBilu