Project

Back to overview

From atomistic exploration of chemical compound space towards bio-molecular design: Quantum mechanical rational compound design (QM-RCD)

English title From atomistic exploration of chemical compound space towards bio-molecular design: Quantum mechanical rational compound design (QM-RCD)
Applicant von Lilienfeld-Toal Otto Anatole
Number 138932
Funding scheme SNSF Professorships
Research institution
Institution of higher education University of Basel - BS
Main discipline Physical Chemistry
Start/End 01.07.2013 - 30.06.2017
Approved amount 1'601'647.00
Show all

All Disciplines (9)

Discipline
Physical Chemistry
Biophysics
Chemical Engineering
Theoretical Physics
Molecular Biology
Mathematics
Biochemistry
Information Technology
Condensed Matter Physics

Keywords (8)

computational ligand design; gradient based optimization; machine learning; chemical compound space; computational drug design; density functional theory; van der Waals interactions; rational compound design

Lay Summary (German)

Lead
Computergestutztes Design gewinnt in allen Gebieten der Technik kontinuierlich an Bedeutung - nur nicht fuer das Optimieren von Chemikalien. Dies liegt zum einem am enormen Rechenaufwand, der noetig ist um das Verhalten neuer Chemikalien mit ausreichender Praezision vorherzusagen. Zum anderen ist der chemische Raum, d.h. alle theoretisch darstellbaren Verbindungen, derart gigantisch, dass ein systematisches Ausprobieren, selbst virtuell auf dem Computer, so ineffizient wie illusorisch ist.
Lay summary
Diesem Projekt liegt das Entwickeln von effizienten Optimierungsmethoden zugrunde, die es ermoeglichen sollen, mit dem Computer die vielversprechendsten chemischen Verbindungen zu identifizieren, denen mit hoher Wahrscheinlichkeit, die erwuenschten physikalischen, chemischen, und sogar biologischen Eigenschaften inne wohnt. Diese koennen sodann von experimentellen Chemikern synthetisiert und getestet werden.
Hierzu sollen zwei Ansaetze verfolgt und an biochemischen Systemen veranschaulicht werden: 
1. Ableitungen von Eigenschaften, wie z.B. potentielle Bindungsenergie oder optisches Spektrum, nach chemischer Zusammensetzung werden auf ihre Eignung fuer Optimierungen untersucht. 
2. Sollen ausreichend praezise Vorhersagen von diesen Eigenschaften durch statistische Methoden der kuenstlichen Intelligenz um mehrere Groessenordnungen schneller werden. 
Beides soll im Rahmen der Quantenmechanik stattfinden, da nur diese es ermoeglicht, sowohl alle chemischen Zusammensetztungen zu beruecksichtigen, als auch Eigenschaften mit ausreichender Praezision vorherzusagen.

Diese Arbeit hat das Potential, die Grundlagen fuer dramatische Veraenderungen in fast allen chemischen Anwendungungsbereichen zu legen. So wie heutzutage Autos und Gebaeude auf Computern entwickelt und optimiert werden, so hoffen wir auch einmal neue chemische Verbindungen verlaesslich und routinemaessig zu ``entdecken''. Hiervon wuerden insbesondere pharmazeutische Forschung und Industrie profitieren, ebenso wie die Materialwissenschaften. Grosse Relevanz koennte diese Forschung entwickeln, wenn durch sie neue Medikamente (Antibiotika) oder Materialien (Katalysatoren, Batterien) entstehen.
Direct link to Lay Summary Last update: 04.03.2013

Responsible applicant and co-applicants

Employees

Publications

Publication
Boosting Quantum Machine Learning Models with a Multilevel Combination Technique: Pople Diagrams Revisited
Zaspel Peter, Huang Bing, Harbrecht Helmut, von Lilienfeld O. Anatole (2019), Boosting Quantum Machine Learning Models with a Multilevel Combination Technique: Pople Diagrams Revisited, in Journal of Chemical Theory and Computation, 15(3), 1546-1559.
Torsional Potentials of Glyoxal, Oxalyl Halides, and Their Thiocarbonyl Derivatives: Challenges for Popular Density Functional Approximations
Tahchieva Diana N., Bakowies Dirk, Ramakrishnan Raghunathan, von Lilienfeld O. Anatole (2018), Torsional Potentials of Glyoxal, Oxalyl Halides, and Their Thiocarbonyl Derivatives: Challenges for Popular Density Functional Approximations, in Journal of Chemical Theory and Computation, 14(9), 4806-4817.
AlxGa1−xAs crystals with direct 2 eV band gaps from computational alchemy
Chang K. Y. Samuel, von Lilienfeld O. Anatole (2018), AlxGa1−xAs crystals with direct 2 eV band gaps from computational alchemy, in Physical Review Materials, 2(7), 073802-073802.
Alchemical and structural distribution based representation for universal quantum machine learning
Faber Felix A., Christensen Anders S., Huang Bing, von Lilienfeld O. Anatole (2018), Alchemical and structural distribution based representation for universal quantum machine learning, in The Journal of Chemical Physics, 148(24), 241717-241717.
Guest Editorial: Special Topic on Data-Enabled Theoretical Chemistry
Rupp Matthias, von Lilienfeld O. Anatole, Burke Kieron (2018), Guest Editorial: Special Topic on Data-Enabled Theoretical Chemistry, in The Journal of Chemical Physics, 148(24), 241401-241401.
Non-covalent interactions across organic and biological subsets of chemical space: Physics-based potentials parametrized from machine learning
Bereau Tristan, DiStasio Robert A., Tkatchenko Alexandre, von Lilienfeld O. Anatole (2018), Non-covalent interactions across organic and biological subsets of chemical space: Physics-based potentials parametrized from machine learning, in The Journal of Chemical Physics, 148(24), 241706-241706.
Generalized Density-Functional Tight-Binding Repulsive Potentials from Unsupervised Machine Learning
Kranz Julian J., Kubillus Maximilian, Ramakrishnan Raghunathan, von Lilienfeld O. Anatole, Elstner Marcus (2018), Generalized Density-Functional Tight-Binding Repulsive Potentials from Unsupervised Machine Learning, in Journal of Chemical Theory and Computation, 14(5), 2341-2352.
Prediction Errors of Molecular Machine Learning Models Lower than Hybrid DFT Error
Faber Felix A., Hutchison Luke, Huang Bing, Gilmer Justin, Schoenholz Samuel S., Dahl George E., Vinyals Oriol, Kearnes Steven, Riley Patrick F., von Lilienfeld O. Anatole (2017), Prediction Errors of Molecular Machine Learning Models Lower than Hybrid DFT Error, in Journal of Chemical Theory and Computation, 13(11), 5255-5264.
Genetic optimization of training sets for improved machine learning models of molecular properties
Browning Nicholas, Ramakrishnan Raghunathan, von Lilienfeld O. Anatole, Rothlisberger Ursula (2017), Genetic optimization of training sets for improved machine learning models of molecular properties, in J Phys Chem Lett, 1351.
Machine Learning, Quantum Chemistry, and Chemical Space
Ramakrishnan Raghunathan, von Lilienfeld O. Anatole (2017), Machine Learning, Quantum Chemistry, and Chemical Space, in Reviews in Computational Chemistry, 30, 225.
Machine Learning Energies of 2 Million Elpasolite (ABC2D6) Crystals
Faber Felix A., Lindmaa Alexander, von Lilienfeld O. Anatole, Armiento Rickard (2016), Machine Learning Energies of 2 Million Elpasolite (ABC2D6) Crystals, in Physical Review Letters, 117(13), 135502-135502.
Alchemical screening of ionic crystals
Solovyeva Alisa, von Lilienfeld O. Anatole (2016), Alchemical screening of ionic crystals, in Phys Chem Chem Phys, 18, 31078.
Fast and accurate predictions of covalent bonds in chemical space
Chang K. Y. Samuel, Fias Stijn, Ramakrishnan Raghunathan, von Lilienfeld O. Anatole (2016), Fast and accurate predictions of covalent bonds in chemical space, in JOURNAL OF CHEMICAL PHYSICS, 144(17), 174110.
Guiding ab initio calculations by alchemical derivatives
Baben M. To, Achenbach J. O., von Lilienfeld O. A. (2016), Guiding ab initio calculations by alchemical derivatives, in JOURNAL OF CHEMICAL PHYSICS, 144(10), 104103.
Properties and reactivity of nucleic acids relevant to epigenomics, transcriptomics, and therapeutics
Gillingham Dennis, Geigle S, von Lilienfeld O. Anatole (2016), Properties and reactivity of nucleic acids relevant to epigenomics, transcriptomics, and therapeutics, in Chem. Soc. Rev., 2637.
Tuning dissociation using isoelectronically doped graphene and hexagonal boron nitride: Water and other small molecules
Al-Hamdani Yasmine S., Alfe Dario, von Lilienfeld O. Anatole, Michaelides Angelos (2016), Tuning dissociation using isoelectronically doped graphene and hexagonal boron nitride: Water and other small molecules, in JOURNAL OF CHEMICAL PHYSICS, 144(15), 154706.
Understanding molecular representations in machine learning: The role of uniqueness and target similarity
Huang Bing, von Lilienfeld O. Anatole (2016), Understanding molecular representations in machine learning: The role of uniqueness and target similarity, in J Chem Phys, 161102.
Electronic spectra from TDDFT and machine learning in chemical space
Ramakrishnan Raghunathan, Hartmann Mia, Tapavicza Enrico, von Lilienfeld O. Anatole (2015), Electronic spectra from TDDFT and machine learning in chemical space, in The Journal of Chemical Physics, 143(8), 084111-084111.
Understanding kernel ridge regression: Common behaviors from simple functions to density functionals
Vu Kevin, Snyder John C., Li Li, Rupp Matthias, Chen Brandon F., Khelif Tarek, Müller Klaus-Robert, Burke Kieron (2015), Understanding kernel ridge regression: Common behaviors from simple functions to density functionals, in International Journal of Quantum Chemistry, 115(16), 1115-1128.
Communication: Water on hexagonal boron nitride from diffusion Monte Carlo
Al-Hamdani Yasmine S., Ma Ming, Alfè Dario, von Lilienfeld O. Anatole, Michaelides Angelos (2015), Communication: Water on hexagonal boron nitride from diffusion Monte Carlo, in The Journal of Chemical Physics, 142(18), 181101-181101.
Big Data Meets Quantum Chemistry Approximations: The Δ-Machine Learning Approach
Ramakrishnan Raghunathan, Dral Pavlo O., Rupp Matthias, von Lilienfeld O. Anatole (2015), Big Data Meets Quantum Chemistry Approximations: The Δ-Machine Learning Approach, in Journal of Chemical Theory and Computation, 11(5), 2087-2096.
Semi-quartic force fields retrieved from multi-mode expansions: Accuracy, scaling behavior, and approximations
Ramakrishnan Raghunathan, Rauhut Guntram (2015), Semi-quartic force fields retrieved from multi-mode expansions: Accuracy, scaling behavior, and approximations, in The Journal of Chemical Physics, 142(15), 154118-154118.
Big Data Meets Quantum Chemistry Approximations: The Δ-Machine Learning Approach
Raghunathan Ramakrishnan Pavlo O. Dral Matthias Rupp O. Anatole von Lilienfeld (2015), Big Data Meets Quantum Chemistry Approximations: The Δ-Machine Learning Approach, in Journal of Chemical Theory and Computation, 2087.
Crystal structure representations for machine learning models of formation energies
Felix Faber Alexander Lindmaa O. Anatole von Lilienfeld Rickard Armiento (2015), Crystal structure representations for machine learning models of formation energies, in International Journal of Quantum Chemistry, 1094.
Fourier series of atomic radial distribution functions: A molecular fingerprint for machine learning models of quantum chemical properties
O. Anatole von Lilienfeld Raghunathan Ramakrishnan Matthias Rupp Aaron Knoll (2015), Fourier series of atomic radial distribution functions: A molecular fingerprint for machine learning models of quantum chemical properties, in International Journal of Quantum Chemistry, 1083.
Machine Learning of Parameters for Accurate Semiempirical Quantum Chemical Calculations
Pavlo O. Dral O. Anatole von Lilienfeld Walter Thiel (2015), Machine Learning of Parameters for Accurate Semiempirical Quantum Chemical Calculations, in Journal of Chemical Theory and Computation, 2120.
Machine Learning Predictions of Molecular Properties: Accurate Many-Body Potentials and Nonlocality in Chemical Space
Katja Hansen Franziska Biegler Raghunathan Ramakrishnan Wiktor Pronobis O. Anatole von Lilienfel (2015), Machine Learning Predictions of Molecular Properties: Accurate Many-Body Potentials and Nonlocality in Chemical Space, in Journal of Chemical Physics Letters, 2326.
Many Molecular Properties from One Kernel in Chemical Space
Ramakrishnan Raghunathan von Lilienfeld O. Anatole (2015), Many Molecular Properties from One Kernel in Chemical Space, in CHIMIA, 182.
Transferable Atomic Multipole Machine Learning Models for Small Organic Molecules
Tristan Bereau Denis Andrienko O. Anatole von Lilienfeld (2015), Transferable Atomic Multipole Machine Learning Models for Small Organic Molecules, in Journal of Chemical Theory and Computation, 3225.
Water on hexagonal boron nitride from diffusion Monte Carlo
Yasmine S. Al-Hamdani Ming Ma Dario Alfè O. Anatole von Lilienfeld Angelos Michaelides (2015), Water on hexagonal boron nitride from diffusion Monte Carlo, in Journal of Chemical Physics, 181101.
Machine learning for many-body physics: The case of the Anderson impurity model
Louis-François Arsenault Alejandro Lopez-Bezanilla O. Anatole von Lilienfeld and Andrew J. Millis (2014), Machine learning for many-body physics: The case of the Anderson impurity model, in Physical Review B, 90, 155136.
Modeling electronic quantum transport with machine learning
Lopez-Bezanilla Alejandro, von Lilienfeld O. Anatole (2014), Modeling electronic quantum transport with machine learning, in PHYSICAL REVIEW B, 89(23), 235411.
Quantum chemistry structures and properties of 134 kilo molecules
Ramakrishnan R., Dral P., Rupp M., von Liliefneld OA (2014), Quantum chemistry structures and properties of 134 kilo molecules, in Scientific Data, 1, 140022.
Toward transferable interatomic van der Waals interactions without electrons: The role of multipole electrostatics and many-body dispersion
Bereau Tristan, Von Lilienfeld O. Anatole (2014), Toward transferable interatomic van der Waals interactions without electrons: The role of multipole electrostatics and many-body dispersion, in Journal of Chemical Physics, 141(3), 034101.
Water on BN doped benzene: A hard test for exchange-correlation functionals and the impact of exact exchange on weak binding
Yasmine S. Al-Hamdani Dario Alfè O. Anatole von Lilienfeld and Angelos Michaeledis (2014), Water on BN doped benzene: A hard test for exchange-correlation functionals and the impact of exact exchange on weak binding, in Journal of Chemical Physics, 141, 18C530.
Application of Diffusion Monte Carlo to Materials Dominated by van der Waals Interactions
Benali Anouar, Shulenburger Luke, Romero Nichols A., Kim Jeongnim, von Lilienfeld O. Anatole, Application of Diffusion Monte Carlo to Materials Dominated by van der Waals Interactions, in Journal of Chemical Theory and Computation, 0(0), 0.
Electronic Spectra from TDDFT and Machine Learning in Chemical Space
R. Ramakrishnan M. Hartmann E. Tapavicza O. A. von Lilienfeld, Electronic Spectra from TDDFT and Machine Learning in Chemical Space, in Journal of Chemical Physics.
Machine Learning for Quantum Mechanical Properties of Atoms in Molecules
M. Rupp R. Ramakrishnan O. A. von Lilienfeld, Machine Learning for Quantum Mechanical Properties of Atoms in Molecules, in Journal of Physical Chemistry Letters.
Quantum Mechanical Treatment of Variable Molecular Composition: From ``Alchemical'' Changes of State Functions to Rational Compound Design
Chang KYS, von Lilienfeld OA, Quantum Mechanical Treatment of Variable Molecular Composition: From ``Alchemical'' Changes of State Functions to Rational Compound Design, in CHIMIA, 0(0), 0-0.

Collaboration

Group / person Country
Types of collaboration
Jean-Louis Reymond/Chemistry Department/University of Berne Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Research Infrastructure
Klaus-Robert Mueller/Machine Learning Group/TU Berlin Germany (Europe)
- in-depth/constructive exchanges on approaches, methods or results
Alexandre Tkatchenko/Theory Department/FHI-Berlin Germany (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
Mark Tuckerman/Chemistry Department/New York University United States of America (North America)
- in-depth/constructive exchanges on approaches, methods or results
Markus Meuwly/Department of Chemistry/University of Basel Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results

Scientific events

Active participation

Title Type of contribution Title of article or contribution Date Place Persons involved
American Chemical Society meeting Talk given at a conference Quantum Machine Learning 02.04.2017 San Francisco, United States of America von Lilienfeld-Toal Otto Anatole;
American Physical Society Meeting Talk given at a conference Quantum Machine Learning 13.03.2017 New Orleans, United States of America von Lilienfeld-Toal Otto Anatole;
Symposium on ''Integrated Data Assimilation'' Talk given at a conference Quantum Machine Learning 19.01.2017 Stuttgart, Germany von Lilienfeld-Toal Otto Anatole;
Understanding Many-Particle Systems with Machine Learning: Machine Learning Meets Many-Particle Problems Talk given at a conference Quantum Machine Learning 26.09.2016 Los Angeles, UCLA, United States of America von Lilienfeld-Toal Otto Anatole;
99th Canadian Society for Chemistry Meeting Talk given at a conference Big Data meets Quantum Chemistry: The Δ-Machine Learning Approach 06.06.2016 Halifax, Canada von Lilienfeld-Toal Otto Anatole;
Materials Research Society Meeting Talk given at a conference Quantum Properties from Machine Learning in Chemical Space 28.03.2016 Phoenix, Arizona, United States of America von Lilienfeld-Toal Otto Anatole;
American Chemical Society Meeting Talk given at a conference Quantum Properties from Machine Learning in Chemical Space 13.03.2016 San Diego, California, United States of America von Lilienfeld-Toal Otto Anatole;
Seminar ueber Big Data in Firma Merck Individual talk Big Data meets Chemistry: Exploring Quantum Properties in Chemical Space with Machine Learning 01.03.2016 Darmstadt, Germany von Lilienfeld-Toal Otto Anatole;
TMS meeting Talk given at a conference Big data in quantum chemistry 15.02.2016 Nashville, Tennessee, United States of America von Lilienfeld-Toal Otto Anatole;
Kolloquium am Institut fuer Physikalische Chemie, KIT Individual talk First principles based exploration of chemical compound space with artificial intelligence and alchemy 01.02.2016 Karlsruhe, Germany von Lilienfeld-Toal Otto Anatole;
GdCH Vortrag, Universitaet Kiel Individual talk Exploring Chemical Space with Machine Learning and Alchemy … and Quantum Mechanics 21.01.2016 Kiel, Germany von Lilienfeld-Toal Otto Anatole;
Big Data of Materials Science from First Principles -- Critical Next Steps (CECAM workshop) Talk given at a conference Quantum Properties from Machine Learning in Chemical Space 30.11.2015 Lausanne, EPFL, Switzerland von Lilienfeld-Toal Otto Anatole;
Psi-k conference Talk given at a conference Machine Learning Methods for the Rapid Yet Accurate Sampling of Chemical Compound Space 06.09.2015 San Sebastian, Spain von Lilienfeld-Toal Otto Anatole;
User Meeting Center for Nanophase Materials Sciences at Oak Ridge National Lab Talk given at a conference Machine Learning Methods for the Rapid Yet Accurate Sampling of Chemical Compound Space 02.09.2015 Oakridge Tennessee, United States of America von Lilienfeld-Toal Otto Anatole;
Summer School of Max-Planck-EPFL Center for Molecular Nanoscience & Technology, Schloss Ringberg Talk given at a conference Machine Learning in Chemical Space 27.07.2015 Tegernsee, Germany von Lilienfeld-Toal Otto Anatole;
CECAM Workshop on ``Next generation quantum based molecular dynamics: challenges and perspectives'' Talk given at a conference Fast potentials from navigating chemical space with machine learning or alchemical coupling 13.07.2015 Bremen, Germany von Lilienfeld-Toal Otto Anatole;
CECAM workshop on ``From Many-Body Hamiltonians to Machine Learning and Back' Talk given at a conference Fourier Series of Atomic Radial Distribution Functions As Molecular Descriptor 11.05.2015 Berlin, Germany Rupp Matthias; Faber Felix Andreas; von Lilienfeld-Toal Otto Anatole; Li Zhenwei; Ramakrishnan Raghunathan;
Workshop on ``Machine Learning for Many-Particle Systems'', Institute of Pure and Applied Mathematics, UCLA, USA Talk given at a conference Machine Learning Models in Chemical Space 23.02.2015 Los Angeles, California, United States of America von Lilienfeld-Toal Otto Anatole;
Workshop on ``Opportunities in Materials Informatics'', University of Wisconsin-Madison, USA Talk given at a conference Efficient Methods for Sampling Chemical Space from First Principles 09.02.2015 Madison, Wisconsin, United States of America von Lilienfeld-Toal Otto Anatole;
Workshop on ``Machine Learning in materials'', Aalto University, Helsinki, Finland Talk given at a conference Machine Learning for the Sampling of Chemical Space from First Principles 03.12.2014 Aalto, Finland von Lilienfeld-Toal Otto Anatole;
Soiree on ``Machine Learning for Atomistic Simulation'', Thomas-Young-Centre, London, UK Talk given at a conference Machine Learning Methods for the Sampling of Chemical Space from First Principles 27.11.2014 London, Great Britain and Northern Ireland von Lilienfeld-Toal Otto Anatole;
Advancing Research through High-Performance Computing Talk given at a conference Efficient methods for sampling materials space from first principles 15.10.2014 University of Pittsburgh, United States of America von Lilienfeld-Toal Otto Anatole;
WATOC 2014 Talk given at a conference Efficient methods for sampling chemical space 05.10.2014 Santiago, Chile von Lilienfeld-Toal Otto Anatole;
Seminar Individual talk COMPUTATIONAL ALCHEMY AND MACHINE LEARNING METHODS FOR THE SAMPLING OF CHEMICAL SPACE FROM FIRST PRINCIPLES 20.06.2014 IBM, Zuerich, Switzerland von Lilienfeld-Toal Otto Anatole;
Seminar Individual talk COMPUTATIONAL ALCHEMY AND MACHINE LEARNING METHODS FOR THE SAMPLING OF CHEMICAL SPACE FROM FIRST PRINCIPLES 17.06.2014 Max-Planck-Institute for Polymer Research, Germany von Lilienfeld-Toal Otto Anatole;
Enhancing DoD Collaborations in Computational Materials Science and Engineering Talk given at a conference COMPUTATIONAL ALCHEMY AND MACHINE LEARNING METHODS FOR THE SAMPLING OF CHEMICAL SPACE FROM FIRST PRINCIPLES 02.06.2014 Dayton, OH, United States of America von Lilienfeld-Toal Otto Anatole;
PASC 2014 Talk given at a conference QM/ML for accurate high-throughput screening of thermochemistry of organic molecules 02.06.2014 Zuerich, Switzerland Ramakrishnan Raghunathan;
Woodward lectures Individual talk Quantum Machine: Supervised learning of Schrödinger's equation in chemical compound space 30.01.2014 Harvard University, Boston, United States of America von Lilienfeld-Toal Otto Anatole;
Seminar Individual talk Quantum Machine: Supervised learning of Schrödinger's equation in chemical compound space 20.01.2014 Max-Planck Institute for Carbon research, Germany von Lilienfeld-Toal Otto Anatole;
Group Seminar in U. Roethlisberger group Individual talk Quantum Machine: Supervised learning of Schrödinger's equation in chemical compound space 14.01.2014 EPFL, Switzerland von Lilienfeld-Toal Otto Anatole;
Quantum Monte Carlo in the Apuan Alps VIII Talk given at a conference Preaching on first principles views on chemical compound space: Atom centered potentials and statistical learning 27.07.2013 Apuan Alps, Italy von Lilienfeld-Toal Otto Anatole;


Self-organised

Title Date Place
Exploring Chemical Space with Machine Learning and Quantum Mechanics (CECAM workshop) 30.05.2016 ETHZ, Zuerich, Switzerland

Knowledge transfer events

Active participation

Title Type of contribution Date Place Persons involved
Science+Fiction Festival 2017 Workshop 06.05.2017 Basel, Switzerland von Lilienfeld-Toal Otto Anatole;
Research talk for Chemistry Highschool teachers from Basel Talk 19.11.2014 University of Basel, Switzerland von Lilienfeld-Toal Otto Anatole;
Zurich.Minds Talk 11.12.2013 Zurich, Switzerland von Lilienfeld-Toal Otto Anatole;


Communication with the public

Communication Title Media Place Year
Talks/events/exhibitions Cafe Scientifique Basel German-speaking Switzerland 2013

Awards

Title Year
Professorship at TATA institute, India 2016
Guest Editor for a special Machine Learning Issue in the International Journal of Quantum Chemistry 2014

Abstract

Atomistic details of matter determine physical and chemical properties. Much research has been carried out in order to understand and model this relationship using quantum mechanical (QM) first principles, statistical mechanics, and the ever increasing role of the computational sciences. Employed methods are based on a physico-chemical framework that nowadays permits to routinely compute relevant properties for any combinations of atomic configurations and composition, also called chemical compound space (CCS). The task of using QM for the discovery of novel compounds with improved properties is therefore equivalent to a combinatorial optimization problem in CCS. Tackling this task through QM based high-throughput screening, albeit straight-forward, is the least efficient way. The overall goal of this proposal is to make a comprehensive exploration of CCS possible by developing QM based rational compound design (QM-RCD) schemes that are drastically more efficient than screening. The proposed QM-RCD approach is based on two of the most fundamental and complementary strategies in numerical optimization theory, namely variational and correlational methods. For the former, differential, response-like, QM predictions of properties will be studied for compounds that are ``close'' in CCS. For the latter, intelligent data analysis methods (machine learning) will be applied to exploit complex correlations. Such models promise to dramatically accelerate the estimation of QM properties, enabling the modeling of massive numbers of compounds within discrete combinatorial optimization implementations, e.g. evolutionary or Monte Carlo based. While many QM properties are amenable to such efforts, the focus of this proposal lies on the archetypical problem of ligand design. The usefulness of this highly interdisciplinary endeavor, tightly linking theoretical physics with computational chemistry and biochemistry, and with computer sciences, will be exemplified for realistic and relevant ligand leads, such as DNA intercalating drugs (dominated by pi-pi-interactions) and ligand-protein (dominated by hydrogen bonding and polarization) interactions.
-