Computational Molecular Dynamics: Challenges, Methods, Ideas: Proceedings of the 2nd International Symposium on Algorithms for Macromolecular Modelling, Berlin, May 21–24, 1997 PDF
Preview Computational Molecular Dynamics: Challenges, Methods, Ideas: Proceedings of the 2nd International Symposium on Algorithms for Macromolecular Modelling, Berlin, May 21–24, 1997
Lecture Notes in Computational Science 4 and Engineering Editors M.Griebel,Bonn D.E.Keyes,Norfolk R.M.Nieminen,Espoo D.Roose,Leuven T.Schlick,NewYork Texts and Monographs in Physics Series Editors: R. Balian W. Beiglbock H. Grosse E. H. Lieb N. Reshetikhin H. Spohn W Thirring Springer-Verlag Berlin Heidelberg GmbH Peter Deuflhard Ja n Hermans Benedict Leimkuhler Alan E. Mark Sebastian Reich Robert D. Skeel (Eds.) Computational Molecular Dynamics: Challenges, Methods, Ideas Proceedings of the 2nd International Symposium on Algorithms for Macromolecular Modelling, Berlin, May 21-24, 1997 With 117 Figures and 22 Tables Springer Editors Peter Deuflhard AlanE.Mark Konrad-Zuse-Zentrum Berlin (ZIB) Laboratorium fiir Physikalische Chemie Takustrasse 7 ETHZentrum D-14195 Berlin-Dahlem, Germany CH-8092 Ziirich, Switzerland [email protected] [email protected] JanHermans Sebastian Reich Department of Biochemistry Department of Mathematics and Biophysics and Statistics University of North Carolina University of Surrey Chapel Hill, NC 27599-7260, USA Guildford, Surrey GU2 5XH, [email protected] United Kingdom [email protected] Benedict Leimkuhler Department of Mathematics Robert D. Skeel University of Kansas Department of Computer Science 405 Snow HalI University of Illinois Lawrence, KS 66045, USA 1304 West Springfield Avenue [email protected] Urbana, IL 61801-6631, USA [email protected] Front cover figure created by Thomas Steinke and Olaf Paetsch, Konrad-Zuse-Zentrum (ZIB), Berlin-Dahlem, Germany Mathematics Subject Classification (1991): 34C35, 65-06, 68-06, 70-06, 81-06, 82-06, 92-06 Cataloging-in-Publication Data applied for Die Deutsche Bibliothek - CIP-Einheitsaufnahme Computational molecular dynamics: challenges, methods, ideas; proceedings of the 2nd International Symposium on Algorithms for Macromolecular Modelling, Berlin,May 21-24, 1997; with 22 tables I Peter Deuflhard ..• (ed.). Berlin; Heidelberg; New York; Barcelona; Hong Kong; London; Milan; Paris; Singapore; Tokyo: Springer, 1999 (Lecture notes in computational science and engineering; 4) ISBN 978-3-540-63242-9 ISBN 978-3-642-58360-5 (eBook) DOI 10.1007/978-3-642-58360-5 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer-Verlag. Violations are liable for prosecution under the German Copyright Law. © Springer-Verlag Berlin Heide1berg 1999 Originally published by Springer-Verlag Berlin Heidelberg New York in 1999 The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Cover Design: Friedhelm Steinen-Broo, Estudio Calamar, Spain Cover production: design &-production GmbH, Heidelberg 1)rpesetting: Camera-ready by Sebastian Reich SPIN 10568830 46/3143 - 5 4 3 2 1 o - Printed on acid-free paper Preface In May 21 - 24, 1997the Second International Symposiumon Algorithms for Macromolecular Modelling was held in the new building ofthe Konrad Zuse Zentrum on the attractive Science Campus ofthe Free University of Berlin. Organizersofthesymposiumweretheeditorsofthis book, plusBernieBrooks and Wilfred van Gunsteren. The event brought together computational sci entists infields like biochemistry, biophysics, physicalchemistry, orstatistical physics and numerical analysts as well as computer scientists working on the advancement ofalgorithms, for a totalofover 120 participants from 19coun tries. In the course ofthe symposium, it was agreed not to write traditional proceedings, but rather to produce a representative volume that combines survey articles and original papers (all refereed) that would give an account ofthe current state ofthe art ofMolecular Dynamics (MD). At present, the main challenge of computational molecular dynamics stems from the huge discrepancy oftimescales: phenomena of interest, such as proteinfolding oractivesitedocking, occurona micro-ormillisecondtime scale while we are routinely able to do computations on a scale ofonly oneor a few nanoseconds. Inorderto bridge this gap, a drasticspeedup ofour algo rithmsandsoftwareappearsnecessary- besidesanyspeeduporiginatingfrom advancesincomputertechnology. However, this will not beenoughto achieve our goal. In addition, there is the need to explore further the potential for improved physical modelling and to develop both new theoretical concepts and new algorithmic ideas. That is why this volume deliberately allocates considerable space to new concepts and ideas from physics and mathematics. Withthe mainchallengeand the generalintentionsoftheeditorsin mind, the volume begins with an Introductory Survey by a longtime leader in the field, HERMAN BERENDSEN, drawingon his longexperience and deep insight into the current status of molecular simulations and their future as an in creasingly important method in structural biology and chemistry. With his unique personalinsight, this articlewill bethe beginningofmanydiscussions as the book as a whole will serve as a forum for alternative views and further perspectives. The remaining 28 articles have been grouped in five chapters that reflect the main topics of the Berlin meeting. As in any interdisciplinary volume, there is a degree ofarbitrariness in the allocation ofsome ofthe articles. The first chapter, on Conformational Dynamics, includes discussion ·of several rather recent computational approaches to treat the dominant slow modes of molecular dynamical systems. In the first paper, SCHULTEN and his group review the new field of "steered molecular dynamics" (SMD), in which "large" external forces are applied in order to be able to study un bindingofligands and conformationchanges ontime scales accessible to MD VI Preface simulations. The second paper, byHELMS & MCCAMMON, surveys a wide range ofdifferent computational techniques for the exploration ofconforma tional transitions of proteins, including the use ofstochastic dynamics with the Poisson-Boltzmann approximation as a simple solvent model. The ar ticle by EICHINGER ET AL. combines several speedup techniques: multiple time stepping algorithms adapted to fit fast multipole methods (see also the last chapter of this book), the previously mentioned SMD technique, and GRUBMULLER'S method of "computational flooding", which uses local po tential modifications in order to successively drive the system to different low-energy basins. The novel approach taken by DEUFLHARD ET AL. em ploys ideas from the mathematics ofdynamical systems to construct certain almost invariant sets in phase space, which can be interpreted as chemical conformations; their algorithm also supplies patterns and rates ofconforma tional changes. In the last paper ofthis chapter, TOLSTUROKOV & VIRNIK describe another use ofdynamical systems tools and propose a simplified set ofdifferentialequationsfor the descriptionofan observedhysteresis behavior in water adsorption-desorption of nucleic acids. The second chapter, on Thermodynamic Modelling, is devoted largely to methods for computing free energies and potentials ofmean force. The paper by HERMANS ET AL. reviews experimental and theoretical techniques for studying the stability of protein-ligand complexes, including a new method for computing absolute free energies of binding with MD simulations, and summarizesrecent applications from their laboratory. MARK ET AL. describe a new method to estimate relative binding free energies ofa series ofrelated ligands on the basis of a single simulated trajectory of a reference state in which a specially constructed, artificial ligand is modelled with a special "soft" potential function. KUCZERA describes a multiple-dimension approach bywhich conformation space is explored, while the potential ofmean force is simultaneouslycomputed. Thejointpaper from the groupsofLESYNG and of MCCAMMON reviews an algorithm for the prediction ofionization constants in proteins; calculations of the relevant protein-solvent system are based on thealreadymentionedPoisson-Boltzmannequation. ThepaperbySTRAUB & ANDRICIOAEIemploysthe Tsallisstatisticstospeedupphasespacesampling. In the final article of this chapter, NEUMAIER ET AL. construct empirical potentials for possible use in off-lattice protein studies. The third chapter, on Enhanced Time-Stepping Algorithms, opens with a personalaccount onlong-timestep integrationby SCHLICK. She assesses both the successes and the limitations ofvarious algorithmic approaches including implicit discretization, harmonic/anharmonicseparation ofmodes, and force splitting techniques combined with Langevin dynamics. The second paper, by ELBER ET AL., describes a large step-size approximation of stochastic path integrals arising from Langevin dynamics - requiring, however, knowl edge about both initial and final states. Onthe basis ofa detailed case study ASCHER & REICH argue that implicit discretizations should not be used with timesteps significantly larger than typical periods of the fast oscilla- Preface VII tions. In the paper by BERNE, the r-RESPA multiple timestepping (MTS) method is described and applied in the context ofHybrid Monte Carlometh ods for sampling techniques such as J-Walking and S-Walking with the aim of a more rapid exploration of rugged energy landscapes. In the next pa per, SKEEL & IZAGUIRRE advocate the use of MTS in a mollified impulse method to overcome resonance instabilities that are inherent in the standard impulsemethod. YetanotherMTS-likeapproachcanbefound inthepaperby JANEZIC & MERZEL, who suggest to split offa harmonic high frequency part ofthe motion and integrate that analytically. Finally, LEIMKUHLER demon strates the stability of the recently proposed explicit symplectic integrators (with fixed timestep) in the numerical integration ofrigid body motion over long time spans. The fourth chapter, on Quantum-Classical Simulations, deals with the integration of molecular systems, parts of which are modelled in terms of quantum mechanics, where a full quantum mechanical treatment would be impossible. In the first paper, JUNGWIRTH & GERBER treat clusters ofinert gases by calculating effective single-mode potentials from classical molecular dynamics which are then used in quantum calculations. An extension be yond the separability approximation is also suggested. The quality of the quantum-classical molecular dynamics (QCMD) model compared with full quantum mechanics (QM) and the Born-Oppenheimer approximation (BO) isconsideredby SCHUTTE & BORNEMANN in termsofapproximationtheory. TheyalsosuggestanextendedQCMD modelthat mayopennewperspectives in the case ofenergy level crossings, where BO is known to break down. Re cently developed structure-preserving numerical integrators for this QCMD modelare given by NETTESHEIM & SCHUTTE. Symplecticmultiple timestep pingvariantsofthese integrators arederived in the paper by NETTESHEIM & REICH. An alternative scheme is presented by HOCHBRUCK & LUBICH, who suggest that a type of mollified exponential integrators are especially well suited for highly oscillatory systems such as QCMD and the Car-Parrinello approximation. The latter approximationis also used in the paper by MEIER ET AL. on ab-initio MD simulations ofcatalysis in a polymerization process. In the last paper of this chapter, IZVEKOV describes an algorithm for the calculation ofabsorption spectra based on exciton-phonon interactions. The fifth and final chapter, on Parallel Force Field Evaluation, takes account of the fact that the bulk of CPU time spent in MD simulations is required for evaluation of the force field. In the first paper, BOARD and his coworkers present a comparison of the performance of various parallel implementations of Ewald and multipole summations together with recommendations for their application. The second paper, by PHILLIPS ET AL., addresses the special problems associated with the design of parallel MD programs. Conflicting issues that shape the design of such codes are identified and the use offeatures suchas multiple threadsand message-driven execution is described. The final paper, by OKUNBOR & MURTY, compares three force decomposition techniques (the checkerboardpartitioningmethod, VIII Preface the force-row interleaving method, and the force-stripped row method) in the context ofa benchmark test problem. August 31, 1998 Peter Deufihard Jan Hermans Benedict Leimkuhler Alan E. Mark Sebastian Reich Robert D. Skeel Table of Contents Introductory Survey Molecular Dynamics Simulations: The Limits and Beyond. ........... 3 Herman J.C. Berendsen I Conformational Dynamics Steered Molecular Dynamics. 39 0 0 0 0 •• 0 0 •• 0 0 •• 0 • 0 0 0 • 0 •• 0 0 •• 0 0 •• 0 0 0 0 0 0 Sergei Izrailev, Sergey Stepaniants, Barry Isralewitz, Dorina Kosztin, Hui Lu, Ferenc Molnar, Willy Wriggers, Klaus Schulten Conformational Transitions ofProteins from Atomistic Simulations o. 66 0 • Volkhard Helms, Jo Andrew McCammon Conformational Dynamics Simulations ofProteins 78 0 0 0 •• 0 0 •• 0 0 •• 0 0 0 0 0 0 Markus Eichinger, Berthold Heymann, Helmut Heller, Helmut Grubmuller, Paul Tavan ComputationofEssentialMolecularDynamicsbySubdivisionTechniques 98 Peter Deuflhard, Michael Dellnitz, Oliver Junge, ChristofSchutte Mathematical Model ofthe Nucleic Acids Conformational Transitions with Hysteresis over Hydration-Dehydration Cycle. 116 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Michael Yeo Tolstorukov, Konstantin Mo Virnik II Thermodynamic Modelling Simulation Studies ofProtein-Ligand Interactions ... 129 0 0 0 • 0 0 0 0 • 0 ••• 0 •• Jan Hermans, Geoffrey Mann, Lu Wang, Li Zhang Estimating Relative Free Energies from a Single Simulation ofthe Ini- tial State o. 149 0 0 ••• 0 ••• 0 •••• 0 0 0 •••• 0 0 0 0 0 ••• 0 0 0 0 0 0 0 0 • 0 0 0 0 0 • 0 0 0 0 • 0 •• Alan E. Mark, Heiko Schiifer, Haiyan Liu, Wilfred van Gunsteren Exploration ofPeptide Free Energy Surfaces 163 0 •• 0 •• 0 ••••••••••••••• 0 • KrzysztofKuczera Prediction of pKas ofTitratable Residues in Proteins Using a Poisson Boltzmann Model ofthe Solute-Solvent System 176 0 0 0 ••••••••••• 0 0 0 •••• X Table ofContents Jan Antosiewicz, Elibieta Blachut-Okrasinska, Tomasz Grycuk, James M. Briggs, Stanislaw T. Wlodek, Bogdan Lesyng, 1. Andrew McCammon Exploiting Tsallis Statistics 197 John E. Straub, loan Andricioaei New Techniques for the Construction ofResidue Potentials for Protein Folding 212 Arnold Neumaier, Stefan Dallwig, Waltraud Huyer, Hermann Schichl III Enhanced Time-Stepping Algorithms Some Failures and Successes of Long-Timestep Approaches to Biomolecular Simulations .. ...................................... 227 Tamar Schlick Application of a Stochastic Path Integral Approach to the Computa- tions ofan Optimal Path and Ensembles ofTrajectories 263 Ron Elber, Benoit Roux, Roberto Olender OnSomeDifficultiesinIntegratingHighlyOscillatoryHamiltonianSys- terns 281 Uri M. Ascher, Sebastian Reich Molecular Dynamics in Systems with Multiple Time Scales: Reference System Propagator Algorithms 297 Bruce J. Berne The Five Femtosecond Time Step Barrier 318 Robert D. Skeel, Jesus A. Izaguirre Long Time Step MD Simulations Using Split Integration Symplectic Method 332 Dusanka Janeiic, Franci Merzel Comparison ofGeometric Integrators for Rigid Body Simulation 349 Benedict J. Leimkuhler IV Quantum-Classical Simulations New Methods in Quantum Molecular Dynamics of Large Polyatomic Systems 365 Pavel Jungwirth, R. Benny Gerber
The list of books you might like
