2011-07-13 Scoring Wissensbasierte Optimierung der Geometrien von Protein-Ligand Komplexen 2011-08-08 Natural sciences + mathematics Naturwissenschaften Knowledge-based Optimization of Protein-Ligand-Complex Geometries Pharmazeutische Chemie https://doi.org/10.17192/z2011.0458 Docking Optimization 2011 doctoralThesis Fachbereich Pharmazie opus:3793 application/pdf Publikationsserver der Universitätsbibliothek Marburg Universitätsbibliothek Marburg The aim of this work was to develop a tool to optimize insilico generated protein-ligand complexes according to DrugScore (DS) potentials. DS is typically used to rescore ligand geometries that were generated by docking. Thus, these poses are optimized according to the scoring function used by the selected docking algorithm. Applying DS to such a geometry does not necessarily guarantee reliable and relevant scoring. Considering the steepness of the DS potentials, even small variations in the atomic positions can lead to large differences in the resulting scores. Thus, a local optimization with respect to DS is strongly recommended in this case. In 2009, O’Boyle et al. stated, that a local optimization is always constrained to the energy well on the potential surface in which the original pose resides. So there may be even deeper wells nearby which are not considered in the local optimization but would be equally valid. The new tool MiniMuDS, developed in this thesis, should account for this problem. On the other hand, MiniMuDS is not intended to perform a global optimization since this would, at the end, result in a new docking algorithm. Instead, the new algorithm is supposed to stay close to the pose generated by the original docking engine and simply adapts it to the DS function, a task typically addressed by local search methods. MiniMuDS was to combine these two tasks by avoiding a strictly local optimization without extending to a fully global search at the same time. Therefore, a strategy was implemented, that contains elements of a global optimization, but is still restricted to a local part of the search space. Simply speaking, the applied algorithm can overcome small hills on the potential surface, but only if the following valley is deeper than the current one. Thus, major energetic barriers between basically different conformations will not be passed. In the validation of MiniMuDS several important properties were shown: 1. The optima of the applied energy model correspond impressively well to the experimentally determined native states of the evaluated complexes. This was shown by the optimization of the original crystal structures, which resulted in an average rmsd of about 0.5Å, a value much smaller than the one observed in case of in-silico generated geometries. This deviation has to be seen in light of the positional accuracy estimated for experimental structure determination. The observed deviations virtually fall into the same range. 2. The aim of conserving the given binding modes was achieved. The presented method allows for modifications up to 2Å rmsd compared to the input geometry. Remarkably, not even 5% of the optimized docking poses fully exploited this available space. On average a modified geometry shows an rmsd of about 1Å to the input structure. 3. MiniMuDS improves a given docking solution by about 0.1Å on average. The best performance was observed for well docked poses between 1 and 2Å rmsd which could be improved by up to 0.3Å on average. 4. It was shown that an optimization exceeding the restrictions of a strictly local search can improve the resulting ranking. Up to 4.7% better success rates at a 2Å cutoff and an improvement of up to 9.3% at the 1Å level were received when comparing MiniMuDS to a local optimization. 5. Taking into account not only the top ranked solution but the whole ranking, it was shown that MiniMuDS strongly improves the discrimination between near-native and misplaced poses. Geometries with lower rmsd values to the crystal structure are more likely to be placed within the first positions of the ranking. 6. The inclusion of additional flexible components into the optimization is easy to manage while results can strongly benefit. This was shown using the example of protein side chain flexibility and binding relevant water molecules. 7. Considering computational efforts, it was shown that it is sufficient to only subject the 10 top-ranked docking solutions to an optimization. This consistently yielded slightly better ranking results for all applied protocols compared to an optimization of all generated docking solutions. At 80% less computational effort, up to 4.7% higher success rates at 2Å and 2.1% higher once at a 1Å cutoff were recorded. Especially the last aspect confirms that it is advisable to focus only on those docking poses that were already ranked high by another scoring function. This way, only poses that score well with respect to two different scoring functions are considered, taking thereby advantage of some kind of consensus effect. In light of these findings, the usage of at least a local optimization has to be strongly recommend before applying DS for rescoring purposes. Beyond that, the application of a more sophisticated search strategy like the one implemented in MiniMuDS is suggested. In particular, when dealing with small, lead-like structures, the presented method showed to substantially improve the results. Protein-Ligand Docking urn:nbn:de:hebis:04-z2011-04581 Optimierung Arzneimitteldesign Wissensbasiertes System Scoring ths Prof. Dr. Klebe Gerhard Klebe, Gerhard (Prof. Dr.) Knowledge-based System 2011-08-09 https://archiv.ub.uni-marburg.de/diss/z2011/0458/cover.png Spitzmüller, Andreas Spitzmüller Andreas Philipps-Universität Marburg Protein-Ligand Komplex Ziel dieser Arbeit war die Entwicklung eines Programms zur Optimierung in-silico generierter Protein-Ligand Komplexe auf Grund von DrugScore (DS) Potentialen. Die Funktion DS wird typischerweise für die Nachbewertung von Ligandgeometrien genutzt, die durch Docking erzeugt wurden. Daher sind diese Geometrien zunächst für die intern verwendete Scoringfunktion des gewählten Docking Algorithmus optimiert. Wird DS auf eine solche Geometrie angewandt, ist nicht automatisch eine aussagekräftige Bewertung garantiert. Bedenkt man die Steilheit der DS Potentiale, so können bereits kleine Abweichungen der Atompositionen zu großen Unterschieden in der Bewertung führen. Daher wird eine lokale Optimierung in diesem Fall ausdrücklich empfohlen. 2009 führten O’Boyle et al. aus, dass eine lokale Optimierung grundsätzlich auf das Tal der Potentialoberfläche beschränkt ist, in dem sich die Ausgangspose befindet. Es könnte aber tiefere Täler in der Nähe geben, die bei einer lokalen Optimierung nicht berücksichtigt werden, obwohl sie ebenso zulässige Lösungen darstellen. Das in dieser Arbeit entwickelte Programm MiniMuDS soll diesem Problem gerecht werden. Andererseits soll keine globale Optimierung durchgeführt werden, da dies zu einem neuen Docking Algorithmus führen würde. Stattdessen soll MiniMuDS nahe an der ursprünglich erzeugten Pose bleiben und diese nur an die DS Funktion anpassen. Um beide Anforderungen zu erfüllen, wurde eine Suchstrategie implementiert, die Elemente einer globalen Suche enthält, sich aber dennoch auf einen abgegrenzten Teil des vollständigen Suchraums beschränkt. Einfach ausgedrückt kann der Algorithmus kleine Hürden auf der Potentialoberfläche überwinden, jedoch nur wenn sich direkt dahinter auch ein tieferes Tal befindet. Größere energetische Barrieren zwischen grundsätzlich unterschiedlichen Konformationen können so nicht passiert werden. Durch die Validierung von MiniMuDS konnten verschiedene wichtige Eigenschaften gezeigt werden: 1. Die Optima der angewandten Zielfunktion stimmen beeindruckend genau mit experimentell bestimmten Komplexstrukturen überein. Dies wurde durch die Optimierung von original Kristallstrukturen gezeigt, die in einer mittleren Abweichung von etwa 0,5Å resultierten. Dies sind deutlich kleinere Abweichungen als im Fall von in-silico generierten Geometrien. Darüber hinaus fallen diese Abweichungen etwa in den Bereich der geschätzten Genauigkeit experimenteller Strukturaufklärung. 2. Das Ziel den vorgegebenen Bindemodus beizubehalten wurde erreicht. MiniMuDS erlaubt Modifikationen bis zu 2Å rmsd gegenüber der Ausgangspose. Bemerkenswerterweise nutzten nicht einmal 5% der optimierten Docking Lösungen diesen Raum aus. Sie zeigten durchschnittliche Abweichungen von etwa 1Å auf. 3. Bezogen auf den rmsd zur Kristallstruktur verbessert MiniMuDS eine gegebene Konformation um etwa 0,1 Å. Die besten Ergebnisse wurden für bereits gut gedockte Posen mit einem ursprünglichen rmsd zwischen 1 und 2Å beobachtet, die im Mittel um bis zu 0,3Å verbessert wurden. 4. Es wurde gezeigt, dass durch die Überwindung der Einschränkungen einer rein lokalen Suche die erzielte Rangliste verbessert wurde. Im Vergleich zu einer lokalen Optimierung wurden bis zu 4,7% bessere Erfolgsraten bei der Erkennung nativ-ähnlicher Posen unter 2Å rmsd auf Rang 1 erzielt. Für Posen unter 1Å lag die Verbesserung bei 9,3 %. 5. Betrachtet man nicht nur die bestbewertete Lösung, sonder die gesamte Rangliste, so wurde gezeigt, dass MiniMuDS die Trennung zwischen nativ-nahen und falsch platzierten Posen deutlich verbessert. Geometrien mit niedrigen rmsd Werten tauchen häufiger auf den vorderen Positionen der Rangliste auf. 6. Die Berücksichtigung zusätzlicher flexibler Komponenten in der Optimierung ist mit MiniMuDS leicht zu bewältigen, wodurch die erzielten Ergebnisse deutlich verbessert werden können. Dies wurde am Beispiel von flexiblen Protein Seitenketten und an der Bindung beteiligter Wasser Moleküle gezeigt. 7. Es wurde gezeigt, dass es ausreichend ist, die zehn besten Lösungen eines Docking Experiments zu Optimieren. Dadurch wurden durchgängig etwas bessere Ergebnisse erzielt als bei der Optimierung aller fünfzig erzeugten Lösungen. Bei 80% geringerem Rechenaufwand wurden so bis zu 4,7% bessere Erfolgsraten erzielt. Besonders der letzte Punkt bestätigt, dass es empfehlenswert ist, sich auf solche Posen zu konzentrieren, die bereits von einer anderen Scoringfunktion gut bewertet wurden. So kann zusätzlich von einem Konsensus Effekt profitiert werden. Mit Blick auf die erzielten Ergebnisse muss die Anwendung zumindest einer lokalen Optimierung dringend Empfohlen werden, bevor DS für Nachbewertungen herangezogen wird. Allerdings wird die Verwendung einer darüber hinausgehenden Suchstrategie wie sie in MiniMuDS implementiert wurde nahe gelegt. Insbesondere bei kleineren Leitstrukturen verbesserte die vorgestellte Methode die Nachbewertungsergebnisse deutlich. monograph Hans F. G. Velec. Neue Methoden in der Computerchemie zur Bewertung und Optimierung von Protein-Ligand-Komplexen. PhD thesis, Philipps- Universität Marburg, 2008. Bernardetta Addis, Marco Locatelli, and Fabio Schoen. Local optima smoothing for global optimization. Optim Method Softw, 20(4):417–437, 2005. Murray, and Richard D. Taylor. Improved protein-ligand docking using GOLD. Proteins, 52(4):609–623, Sep 2003. Caterina Bissantz, Gerd Folkers, and Didier Rognan. Protein-Based Vir- tual Screening of Chemical Databases. 1. Evaluation of Different Dock- ing/Scoring Combinations. Journal of Medicinal Chemistry, 43(25): 4759–4767, 2000. Hay Dvir, Hua Liang Jiang, Dawn M. Wong, Michal Harel, M. Chetrit, Xu Chang He, Gui Yu Jin, G. L. Yu, X. C. Tang, Israel Silman, Dina L. Bai, and Joel L. Sussman. X-ray Structures of Torpedo cal- ifornica Acetylcholinesterase Complexed with (+)-Huperzine A and (– )-Huperzine B: Structural Evidence for an Active Site Rearrangement. Biochemistry, 41(35):10810–10818, 2002a. Susan L. McGovern and Brian K. Shoichet. Information Decay in Molecu- lar Docking Screens against Holo, Apo, and Modeled Conformations of Enzymes. Journal of Medicinal Chemistry, 46(14):2895–2907, 2003. Tiejun Cheng, Xun Li, Yan Li, Zhihai Liu, and Renxiao Wang. Compar- ative assessment of scoring functions on a diverse test set. J Chem Inf Model, 49(4):1079–1093, Apr 2009. Michael J. D. Powell. An efficient method for finding the minimum of a function of several variables without calculating derivatives. The Com- puter Journal, 7(2):155–162, 1964. M. J. Box. A New Method of Constrained Optimization and a Comparison With Other Methods. The Computer Journal, 8(1):42–52, 1965. Stephanie B. de Beer, Nico P. Vermeulen, and Chris Oostenbrink. The role of water molecules in computational drug design. Current topics in medicinal chemistry, 10(1):55–66, 2010. Benjamin C. Roberts and Ricardo L. Mancera. Ligand-Protein Docking with Water Molecules. Journal of Chemical Information and Modeling, 48(2):397–408, 2008. Bernard R. Brooks, Robert E. Bruccoleri, Barry D. Olafson, David J. States, S. Swaminathan, and Martin Karplus. CHARMM: A program for macromolecular energy, minimization, and dynamics calculations. Journal of Computational Chemistry, 4(2):187–217, 1983. Paolini, and Roger P. Mee. Empirical scoring functions: I. The develop- ment of a fast empirical scoring function to estimate the binding affinity of ligands in receptor complexes. J Comput Aided Mol Des, 11(5):425– 445, Sep 1997. Rosenberry, Israel Silman, and Joel L. Sussman. 3D Structure of Tor- pedo californica Acetylcholinesterase Complexed with Huprine X at 2.1 Å Resolution: Kinetic and Molecular Dynamic Correlates. Biochemistry, 41(9):2970–2981, 2002b. Gerard Klebe and Thomas Mietzner. A fast and efficient method to gen- erate biologically relevant conformations. J Comput-Aided Mol Des, 8 (5):583–606, Oct 1994. Matthias Rarey, Bernd Kramer, Thomas Lengauer, and Gerard Klebe. A fast flexible docking method using an incremental construction algo- rithm. J Mol Biol, 261(3):470–489, Aug 1996. Ingo Muegge and Yvonne C. Martin. A general and fast scoring function for protein-ligand interactions: a simplified potential approach. J Med Chem, 42(5):791–804, Mar 1999. Yipin Lu, Renxiao Wang, Chao-Yie Yang, and Shaomeng Wang. Analysis of Ligand-Bound Water Molecules in High-Resolution Crystal Structures of Protein-Ligand Complexes. Journal of Chemical Information and Modeling, 47(2):668–675, 2007. Scott J. Weiner, Peter A. Kollman, David A. Case, U. Chandra Singh, Caterina Ghio, Guliano Alagona, Salvatore Profeta, and Paul Weiner. A new force field for molecular mechanical simulation of nucleic acids and proteins. Journal of the American Chemical Society, 106(3):765– 784, 1984. John A. Nelder and Roger Mead. A Simplex Method for Function Mini- mization. The Computer Journal, 7(4):308–313, 1965. Elaine C. Meng, Brian K. Shoichet, and Irwin D. Kuntz. Automated docking with grid-based energy evaluation. Journal of Computational Chemistry, 13(4):505–524, 1992. Renxiao Wang, Yipin Lu, and Shaomeng Wang. Comparative evaluation of 11 scoring functions for molecular docking. J Med Chem, 46(12): 2287–2303, Jun 2003. Esther Kellenberger, Jordi Rodrigo, Pascal Muller, and Didier Rognan. Comparative evaluation of eight docking tools for docking and virtual screening accuracy. Proteins, 57(2):225–242, Nov 2004. Taylor, and Robin Taylor. Comparing protein-ligand docking programs is difficult. Proteins, 60(3):325–332, Aug 2005. Bibliography Akifumi Oda, Keiichi Tsuchida, Tadakazu Takakura, Noriyuki Yamaotsu, and Shuichi Hirono. Comparison of Consensus Scoring Strategies for Evaluating Computational Models of Protein-Ligand Complexes. Jour- nal of Chemical Information and Modeling, 46(1):380–391, 2006. Jason B. Cross, David C. Thompson, Brajesh K. Rai, J. Christian Baber, Kristi Yi Fan, Yongbo Hu, and Christine Humblet. Comparison of sev- eral molecular docking programs: pose prediction and virtual screening accuracy. J Chem Inf Model, 49(6):1455–1474, Jun 2009. Paul S. Charifson, Joseph J. Corkery, Mark A. Murcko, and W. Patrick Walters. Consensus scoring: A method for obtaining improved hit rates from docking databases of three-dimensional structures into proteins. J Med Chem, 42(25):5100–5109, Dec 1999. Robert D. Clark, Alexander Strizhev, Joseph M. Leonard, James F. Blake, and James B. Matthew. Consensus scoring for ligand/protein inter- actions. Journal of Molecular Graphics and Modelling, 20(4):281–295, 2002. Relibase: design and development of a database for comprehensive anal- ysis of protein-ligand interactions. J Mol Biol, 326(2):607–620, Feb 2003. Bibliography Gareth Jones, Peter Willett, Robert C. Glen, Andrew R. Leach, and Robin Taylor. Development and validation of a genetic algorithm for flexible docking. J Mol Biol, 267(3):727–748, Apr 1997. Manfred Hendlich. Databases for protein-ligand complexes. Acta Crystal- logr, D54(Pt 6 Pt 1):1178–1182, Nov 1998. Michael J. Hartshorn, Marcel L. Verdonk, Gianni Chessari, Suzanne C. Brewerton, Wijnand T. M. Mooij, Paul N. Mortenson, and Christo- pher W. Murray. Diverse, high-quality test set for the validation of protein-ligand docking performance. J Med Chem, 50(4):726–741, Feb 2007. Anders Wallqvist and David G. Covell. Docking enzyme-inhibitor com- plexes using a preference-based free-energy surface. Proteins: Structure, Function, and Bioinformatics, 25(4):403–419, 1996. Hans F. G. Velec, Holger Gohlke, and Gerhard Klebe. DrugScore(CSD)– knowledge-based scoring function derived from small molecule crystal Bibliography data with superior recognition rate of near-native ligand poses and better affinity prediction. J Med Chem, 48(20):6296–6303, Oct 2005. James Arvo. Fast random rotation matrices. In David Kirk, editor, Graph- ics Gems III, pages 117–120, San Diego, CA, USA, 1992. Academic Press Professional, Inc. ISBN 0-12-409671-9. Gerd Neudert and Gerhard Klebe. fconv: format conversion, manipulation and feature computation of molecular data. Bioinformatics, 27(7):1021– 1022, 2011. Wijnand T. M. Mooij and Marcel L. Verdonk. General and targeted statis- tical potentials for protein-ligand interactions. Proteins, 61(2):272–287, Nov 2005. Bibliography Jorge J. Moré and Zhijun Wu. Smoothing Techniques for Macromolecu- lar Global Optimization. Technical Report MCS-P542-0995, Argonne National Laboratory, September 1995. David E. Goldberg. Genetic Algorithms in Search, Optimization, and Ma- chine Learning. Addison-Wesley Professional, Boston, MA, USA, 1st edition, January 1989. ISBN 0201157675. Holger Gohlke, Manfred Hendlich, and Gerhard Klebe. Knowledge-based scoring function to predict protein-ligand interactions. J Mol Biol, 295 (2):337–356, Jan 2000. Jon A. Erickson, Mehran Jalaie, Daniel H. Robertson, Richard A. Lewis, and Michal Vieth. Lessons in Molecular Recognition: The Effects of Ligand and Protein Flexibility on Molecular Docking Accuracy. Journal of Medicinal Chemistry, 47(1):45–55, 2004. Bibliography Miklos Feher. Consensus scoring for protein-ligand interactions. Drug Discovery Today, 11(9-10):421–428, May 2006. Chandrika B-Rao, Jyothi Subramanian, and Somesh D. Sharma. Managing protein flexibility in docking and its applications. Drug Discovery Today, 14(7-8):394–400, 2009. Martin Stahl. Modifications of the scoring function in FlexX for virtual screening applications. Perspectives in Drug Discovery and Design, 20: 83–98, 2000. Andrew R. Leach. Molecular Modelling: Principles and Applications. Bichitra K. Biswal, Meitian Wang, Maia M. Cherney, Laval Chan, Con- stantin G. Yannopoulos, Darius Bilimoria, Jean Bedard, and Michael N. G. James. Non-nucleoside Inhibitors Binding to Hepatitis C Virus Bibliography NS5B Polymerase Reveal a Novel Mechanism of Inhibition. Journal of Molecular Biology, 361(1):33–45, 2006. William H. Press, Saul A. Teukolsky, William T. Vetterling, and Brian P. Flannery. Numerical Recipes: The Art of Scientific Computing. Cam- bridge University Press, 3 edition, September 2007. ISBN 0521880688. Scott Kirkpatrick, C. D. Gelatt, and Mario P. Vecchi. Optimization by simulated annealing. Science, 220(4598):671–680, May 1983. Ingo Muegge. PMF scoring revisited. J Med Chem, 49(20):5895–5902, Oct 2006. Tudor I. Oprea. Property distribution of drug-related chemical databases. J Comput-Aided Mol Des, 14(3):251–264, Mar 2000. Sérgio Filipe Sousa, Pedro Alexandrino Fernandes, and Maria João Ramos. Protein-ligand docking: Current status and future challenges. Proteins: Structure, Function, and Bioinformatics, 65(1):15–26, 2006. Johannes Schulze Wischeler, Dong Sun, Nicola U. Sandner, Uwe Linne, Andreas Heine, Ulrich Koert, and Gerhard Klebe. Stereo-and Regiose- lective Azide/Alkyne Cycloadditions in Carbonic Anhydrase II via Teth- ering, Monitored by Crystallography and Mass Spectrometry. Chemistry – A European Journal, 17(21):5842–5851, 2011. Nicolas Foloppe, Lisa M. Fisher, Rob Howes, Peter Kierstan, Andrew Pot- ter, Alan G. S. Robertson, and Allan E. Surgenor. Structure-Based Design of Novel Chk1 Inhibitors: Insights into Hydrogen Bonding and Protein–Ligand Affinity. J Med Chem, 48(13):4332–4345, 2005. Noel M. O'Boyle, John W. Liebeschuetz, and Jason C. Cole. Testing as- sumptions and hypotheses for rescoring success in protein-ligand dock- ing. J Chem Inf Model, 49(8):1871–1878, Aug 2009. Frank H. Allen. The Cambridge Structural Database: a quarter of a million crystal structures and rising. Acta Crystallogr, B58(Pt 3 Pt 1):380–388, Jun 2002. Hans-Joachim Böhm. The development of a simple empirical scoring func- tion to estimate the binding constant for a protein-ligand complex of known three-dimensional structure. J Comput Aided Mol Des, 8(3): 243–256, Jun 1994. Helen M. Berman, John Westbrook, Zukang Feng, Gary Gilliland, T. N. Bhat, Helge Weissig, Ilya N. Shindyalov, and Philip E. Bourne. The Protein Data Bank. Nucleic Acids Res, 28(1):235–242, Jan 2000. Gerard J. Kleywegt, Mark R. Harris, Jin-yu Zou, Thomas C. Taylor, Anders Wählby, and T. Alwyn Jones. The Uppsala Electron-Density Server. Acta Crystallographica, D60(12 Part 1):2240–2249, Dec 2004. Matthew Clark, Richard D. Cramer, and Nicole Van Opdenbosch. Vali- dation of the general purpose tripos 5.2 force field. J Comput Chem, 10 (8):982–1012, 1989. Gregory L. Warren, C. Webster Andrews, Anna-Maria Capelli, Brian Clarke, Judith LaLonde, Millard H. Lambert, Mika Lindvall, Neysa Nevins, Simon F. Semus, Stefan Senger, Giovanna Tedesco, Ian D. Wall, James M. Woolven, Catherine E. Peishoff, and Martha S. Head. A criti- cal assessment of docking programs and scoring functions. J Med Chem, 49(20):5912–5931, Oct 2006. Philippe Ferrara, Holger Gohlke, Daniel J. Price, Gerhard Klebe, and Charles L. Brooks. Assessing scoring functions for protein-ligand inter- actions. J Med Chem, 47(12):3032–3047, Jun 2004. FlexX User Guide. FlexX Protein-Ligand Docker – User + Technical Ref- erence as Part of LeadIT 1.2. BioSolveIT GmbH, An der Ziegelei 79, 53757 St. Augustin, Germany, 2010. URL http://www.biosolveit.de. Marco Dorigo and Thomas Stützle. Ant Colony Optimization. The MIT Press, Cambridge, MA, USA, 1st edition, July 2004. ISBN 0262042193. Chung-Jung Tsai, Sandeep Kumar, Buyong Ma, and Ruth Nussinov. Fold- ing funnels, binding funnels, and protein function. Protein Sci, 8(6): 1181–1190, Jun 1999. Rafael Najmanovich, Josef Kuttner, Vladimir Sobolev, and Marvin Edel- man. Side-chain flexibility in proteins upon ligand binding. Proteins: Structure, Function, and Bioinformatics, 39(3):261–268, 2000. English Drug Design