How machine learning can assist the interpretation of ab initio molecular dynamics simulations and conceptual understanding of chemistry

被引:81
作者
Hase, Florian [1 ]
Galvan, Ignacio Fdez. [2 ]
Aspuru-Guzik, Alan [3 ,4 ,5 ,6 ]
Lindh, Roland [2 ]
Vacher, Morgane [2 ]
机构
[1] Harvard Univ, Dept Chem & Chem Biol, Cambridge, MA 02138 USA
[2] Uppsala Univ, Theoret Chem Programme, Dept Chem Angstrom, Box 538, S-75121 Uppsala, Sweden
[3] Univ Toronto, Dept Chem, Toronto, ON M5S 3H6, Canada
[4] Univ Toronto, Dept Comp Sci, Toronto, ON M5S 3H6, Canada
[5] Vector Inst Artifcial Intelligence, Toronto, ON M5S 1M1, Canada
[6] Canadian Inst Adv Res CIFAR, Toronto, ON M5S 1M1, Canada
基金
瑞典研究理事会;
关键词
CHEMILUMINESCENCE; ATOMS; 1,2-DIOXETANE; BONDS; FORCE;
D O I
10.1039/c8sc04516j
中图分类号
O6 [化学];
学科分类号
0703 ;
摘要
Molecular dynamics simulations are often key to the understanding of the mechanism, rate and yield of chemical reactions. One current challenge is the in-depth analysis of the large amount of data produced by the simulations, in order to produce valuable insight and general trends. In the present study, we propose to employ recent machine learning analysis tools to extract relevant information from simulation data without a priori knowledge on chemical reactions. This is demonstrated by training machine learning models to predict directly a specific outcome quantity of ab initio molecular dynamics simulations - the timescale of the decomposition of 1,2-dioxetane. The machine learning models accurately reproduce the dissociation time of the compound. Keeping the aim of gaining physical insight, it is demonstrated that, in order to make accurate predictions, the models evidence empirical rules that are, today, part of the common chemical knowledge. This opens the way for conceptual breakthroughs in chemistry where machine analysis would provide a source of inspiration to humans.
引用
收藏
页码:2298 / 2307
页数:10
相关论文
共 57 条
[1]   The valency and the periodical system - Attempt on a theory of molecular compound [J].
Abegg, R .
ZEITSCHRIFT FUR ANORGANISCHE CHEMIE, 1904, 39 (03) :330-380
[2]   EFFECTS OF METHYLATION ON THE THERMAL-STABILITY AND CHEMI-LUMINESCENCE PROPERTIES OF 1,2-DIOXETANES [J].
ADAM, W ;
BAADER, WJ .
JOURNAL OF THE AMERICAN CHEMICAL SOCIETY, 1985, 107 (02) :410-416
[3]  
[Anonymous], 2017, OPENMOLCAS
[4]   Molcas 8: New capabilities for multiconfigurational quantum chemical calculations across the periodic table [J].
Aquilante, Francesco ;
Autschbach, Jochen ;
Carlson, Rebecca K. ;
Chibotaru, Liviu F. ;
Delcey, Mickael G. ;
De Vico, Luca ;
Galvan, Ignacio Fdez ;
Ferre, Nicolas ;
Frutos, Luis Manuel ;
Gagliardi, Laura ;
Garavelli, Marco ;
Giussani, Angelo ;
Hoyer, Chad E. ;
Li Manni, Giovanni ;
Lischka, Hans ;
Ma, Dongxia ;
Malmqvist, Per Ake ;
Mueller, Thomas ;
Nenov, Artur ;
Olivucci, Massimo ;
Pedersen, Thomas Bondo ;
Peng, Daoling ;
Plasser, Felix ;
Pritchard, Ben ;
Reiher, Markus ;
Rivalta, Ivan ;
Schapiro, Igor ;
Segarra-Marti, Javier ;
Stenrup, Michael ;
Truhlar, Donald G. ;
Ungur, Liviu ;
Valentini, Alessio ;
Vancoillie, Steven ;
Veryazov, Valera ;
Vysotskiy, Victor P. ;
Weingart, Oliver ;
Zapata, Felipe ;
Lindh, Roland .
JOURNAL OF COMPUTATIONAL CHEMISTRY, 2016, 37 (05) :506-541
[5]   Atomic Cholesky decompositions: A route to unbiased auxiliary basis sets for density fitting approximation with tunable accuracy and efficiency [J].
Aquilante, Francesco ;
Gagliardi, Laura ;
Pedersen, Thomas Bondo ;
Lindh, Roland .
JOURNAL OF CHEMICAL PHYSICS, 2009, 130 (15)
[6]   The Matter Simulation (R)evolution [J].
Aspuru-Guzik, Alan ;
Lindh, Roland ;
Reiher, Markus .
ACS CENTRAL SCIENCE, 2018, 4 (02) :144-152
[7]   Neural network approach to quantum-chemistry data: Accurate prediction of density functional theory energies [J].
Balabin, Roman M. ;
Lomakina, Ekaterina I. .
JOURNAL OF CHEMICAL PHYSICS, 2009, 131 (07)
[8]  
Barbatti M, 2006, NEWTON X PACKAGE NEW
[9]   Generalized neural-network representation of high-dimensional potential-energy surfaces [J].
Behler, Joerg ;
Parrinello, Michele .
PHYSICAL REVIEW LETTERS, 2007, 98 (14)
[10]   Adaptive machine learning framework to accelerate ab initio molecular dynamics [J].
Botu, Venkatesh ;
Ramprasad, Rampi .
INTERNATIONAL JOURNAL OF QUANTUM CHEMISTRY, 2015, 115 (16) :1074-1083