Atomistic structure learning

One endeavour of modern physical chemistry is to use bottom-up approaches to design materials and drugs with desired properties. Here we introduce an atomistic structure learning algorithm (ASLA) that utilizes a convolutional neural network to build 2D compounds and layered structures atom by atom. The algorithm takes no prior data or knowledge on atomic interactions but inquires a first-principles quantum mechanical program for physical properties. Using reinforcement learning, the algorithm accumulates knowledge of chemical compound space for a given number and type of atoms and stores this in the neural network, ultimately learning the blueprint for the optimal structural arrangement of the atoms for a given target property. ASLA is demonstrated to work on diverse problems, including grain boundaries in graphene sheets, organic compound formation and a surface oxide structure. This approach to structure prediction is a first step toward direct manipulation of atoms with artificially intelligent first principles computer codes.

[1]  M Schmid,et al.  Structure of Ag(111)-p(4 x 4)-O: no silver oxide. , 2006, Physical review letters.

[2]  Yu Xie,et al.  Global minimization of gold clusters by combining neural network potentials and the basin-hopping method. , 2015, Nanoscale.

[3]  Thomas Blaschke,et al.  Molecular de-novo design through deep reinforcement learning , 2017, Journal of Cheminformatics.

[4]  Demis Hassabis,et al.  Mastering the game of Go with deep neural networks and tree search , 2016, Nature.

[5]  Richard S. Sutton,et al.  Reinforcement Learning: An Introduction , 1998, IEEE Trans. Neural Networks.

[6]  J S Smith,et al.  ANI-1: an extensible neural network potential with DFT accuracy at force field computational cost , 2016, Chemical science.

[7]  Shane Legg,et al.  Human-level control through deep reinforcement learning , 2015, Nature.

[8]  Muratahan Aykol,et al.  Materials Design and Discovery with High-Throughput Density Functional Theory: The Open Quantum Materials Database (OQMD) , 2013 .

[9]  R. Kondor,et al.  Gaussian approximation potentials: the accuracy of quantum mechanics, without the electrons. , 2009, Physical review letters.

[10]  Ye Mei,et al.  Folding of a helix at room temperature is critically aided by electrostatic polarization of intraprotein hydrogen bonds. , 2010, Journal of the American Chemical Society.

[11]  E. Ferroni,et al.  Chemisorption of oxygen on the silver (111) surface , 1974 .

[12]  Michele Parrinello,et al.  Generalized neural-network representation of high-dimensional potential-energy surfaces. , 2007, Physical review letters.

[13]  George E. Dahl,et al.  Prediction Errors of Molecular Machine Learning Models Lower than Hybrid DFT Error. , 2017, Journal of chemical theory and computation.

[14]  Gisbert Schneider,et al.  Computer-based de novo design of drug-like molecules , 2005, Nature Reviews Drug Discovery.

[15]  J. Xin,et al.  The favourable large misorientation angle grain boundaries in graphene. , 2015, Nanoscale.

[16]  Alexander Binder,et al.  On Pixel-Wise Explanations for Non-Linear Classifier Decisions by Layer-Wise Relevance Propagation , 2015, PloS one.

[17]  K. Müller,et al.  Fast and accurate modeling of molecular atomization energies with machine learning. , 2011, Physical review letters.

[18]  Lei Cao,et al.  A study of count-based exploration and bonus for reinforcement learning , 2017, 2017 IEEE 2nd International Conference on Cloud Computing and Big Data Analysis (ICCCBDA).

[19]  Demis Hassabis,et al.  Mastering the game of Go without human knowledge , 2017, Nature.

[20]  Qiang Yang,et al.  A Survey on Transfer Learning , 2010, IEEE Transactions on Knowledge and Data Engineering.

[21]  K. Müller,et al.  Machine Learning Predictions of Molecular Properties: Accurate Many-Body Potentials and Nonlocality in Chemical Space , 2015, The journal of physical chemistry letters.

[22]  Peter Stone,et al.  Transfer Learning for Reinforcement Learning Domains: A Survey , 2009, J. Mach. Learn. Res..

[23]  Klaus-Robert Müller,et al.  Machine learning of accurate energy-conserving molecular force fields , 2016, Science Advances.

[24]  Christopher K Prier,et al.  Discovery of an α-Amino C–H Arylation Reaction Using the Strategy of Accelerated Serendipity , 2011, Science.

[25]  Olexandr Isayev,et al.  Deep reinforcement learning for de novo drug design , 2017, Science Advances.

[26]  Zhenwei Li,et al.  Molecular dynamics with on-the-fly machine learning of quantum-mechanical forces. , 2015, Physical review letters.

[27]  Matthias Scheffler,et al.  When seeing is not believing: Oxygen on Ag(111), a simple adsorption system? , 2005 .

[28]  Ryan P. Adams,et al.  Design of efficient molecular organic light-emitting diodes by a high-throughput virtual screening and experimental approach. , 2016, Nature materials.

[29]  J. Nørskov,et al.  Computational high-throughput screening of electrocatalytic materials for hydrogen evolution , 2006, Nature materials.

[30]  Alán Aspuru-Guzik,et al.  Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules , 2016, ACS central science.

[31]  Sanguthevar Rajasekaran,et al.  Accelerating materials property predictions using machine learning , 2013, Scientific Reports.

[32]  O. A. von Lilienfeld,et al.  Communication: Understanding molecular representations in machine learning: The role of uniqueness and target similarity. , 2016, The Journal of chemical physics.

[33]  Martín Abadi,et al.  TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems , 2016, ArXiv.

[34]  Bjørk Hammer,et al.  Neural-network-enhanced evolutionary algorithm applied to supported metal nanoparticles , 2018 .

[35]  Burke,et al.  Generalized Gradient Approximation Made Simple. , 1996, Physical review letters.

[36]  P. Kirkpatrick,et al.  Chemical space , 2004, Nature.

[37]  Alexandre Tkatchenko,et al.  Quantum-chemical insights from deep tensor neural networks , 2016, Nature Communications.

[38]  Alán Aspuru-Guzik,et al.  Reinforced Adversarial Neural Computer for de Novo Molecular Design , 2018, J. Chem. Inf. Model..

[39]  Alán Aspuru-Guzik,et al.  Accelerating the discovery of materials for clean energy in the era of smart automation , 2018, Nature Reviews Materials.

[40]  Michael Walter,et al.  The atomic simulation environment-a Python library for working with atoms. , 2017, Journal of physics. Condensed matter : an Institute of Physics journal.

[41]  King,et al.  Imaging the surface and the interface atoms of an oxide film on Ag111 by scanning tunneling microscopy: experiment and theory , 2000, Physical review letters.

[42]  K. Jacobsen,et al.  Real-space grid implementation of the projector augmented wave method , 2004, cond-mat/0411218.

[43]  A Michaelides,et al.  Revisiting the structure of the p(4 x 4) surface oxide on Ag(111). , 2006, Physical review letters.

[44]  Volker L. Deringer,et al.  Data-Driven Learning of Total and Local Energies in Elemental Boron. , 2017, Physical review letters.