| PHARMACOPHORE
PERCEPTION, DEVELOPMENT, AND USE IN DRUG DESIGN edited by OSMAN F. GÜNER |
|
| Contents Preface xv List of Contributors xxi Part I. THE ORIGINS OF PHARMACOPHORE RESEARCH 1 1. Evolution of the Pharmacophore Concept in Pharmaceutical Research 3 Peter Gund Part II. ANALOG-BASED PHARMACOPHORES 13 2. Manual Pharmacophore Generation: Visual Pattern Recognition 17 Osman F. Güner 3. Pharmacophore Definition Using the Active Analog Approach 21 Denise D. Beusen and Garland R. Marshall 3.1. Introduction 23 3.2. The Active Analog Approach 24 3.3. Systematic Search and the Rigid Geometry Approximation 26 3.4. Combinatorial Nature of Systematic Search 26 3.5. Strategies for Defeating the Combinatorial Explosion 27 3.5.1. Rigid Body Rotations and Building Molecules from Aggregates 27 3.5.2. Look-Ahead 31 3.5.3. Energy Filtering 31 3.5.4. Ping Closures 31 3.6. Systematic Search Parameters Which Impact Sampling Completeness 33 3.7. Analysis of Datasets 38 3.8. Considerations in Using the Active Analog Approach 39 3.8.1. Pharmacophore vs. Active-Site Models 39 3.8.2. Selection of Molecules for Analysis 40 3.8.3. Interpreting the Results 40 3.9. Examples of the Application of the Active Analog Approach 41 3.9.1. Morphiceptin Analogs 41 3.9.2. ACE Inhibitors 42 3.9.3. Substance P Antagonists 43 Automated Pharmacophore Development Systems 47 4. DISCO: What We Did Right and What We Missed 49 Yvonne Connolly Martin 4.1. Overview of DISCO 51 4.2. The Problem of Searching for Correspondences 54 4.3. The Importance of Conformational Searching 55 4.4. Selecting the Conformational Ensemble 56 4.5. Selecting Compounds for the DISCO Analysis 58 4.6. What DISCO Uses as Points to Match 60 4.7. Tolerance: The Trade-Off Between a Close Match and Including More Points 61 4.8. Typical Outcomes and Strategies for Follow-Up 62 4.9. Shortcomings in DISCO 63 4.10. DISCO in the Age of HTS and Molecular Diversity 64 5. HipHop: Pharmacophores Based on Multiple Common-Feature Alignments 69 Omoshile O. Clement and Adrea Trope Mehl 5.1. Background 71 5.2. Methodology 72 5.2.1. General Considerations 72 5.2.2. Algorithm 73 5.2.3. Choosing Relevant Conformations 73 5.2.4. Feature Definition 74 5.2.5. Generating Common Feature Hypotheses Using HipHop 74 5.2.6. Input Conformers 75 5.2.7. Parameter Setting 75 5.3. Applications 77 5.3.1. 5HTP-3 Antagonists 77 5.3.2. ETA Endothelin Antagonists 79 5.3.3. HIV-1 Protease Inhibitors 79 5.4. Conclusion 82 6. GASP: Genetic Algorithm Superimposition Program 85 Gareth Jones and Peter Willett 6.1. Introduction 87 6.2. The Chromosome Representation and the Genetic Operators 89 6.2.1. Input of 3D Structures 89 6.2.2. Chromosome Representation 90 6.2.3. Genetic Operators 91 6.3. The Fitness Function 92 6.3.1. Generation of Conformations and Least-Squares Fitting 93 6.3.2. Calculation of the van der Waals Energy 94 6.3.3. Calculation of the Volume Integral 94 6.3.4. Calculation of the Similarity Score 95 6.4. Identification of Pharmacophores Using GASP 101 6.4.1. Leu-Enkephalin and a Hybrid Morphine 101 6.4.2. Overlay of Four 5-HT3 Antagonists 102 6.4.3. Overlay of Six Angiotensin II Receptor Antagonists 103 6.5. Conclusion 104 7. Exploring Pharmacophores with Chem-X 107 Stephen J. Cato 7.1. Introduction 109 7.2. Centers in Chem-X 110 7.3. Centers in 3D Searching 111 7.4. Pharmacophore Keys Defined 114 7.5. Generating Pharmacophore Keys 117 7.6. Working with Pharmacophore Keys 119 7.7. Pharmacophore Diversity 121 7.8. Virtual Screening 123 7.9. Diamond Pharmacophores 124 Predictive Model Development—3D QSAR 127 8. Apex-3D: Activity Prediction Expert System with 3D QSAR 129 Erich R. Vorpagel and Valery E. Golender 8.1. Introduction 131 8.2. General Description of Apex-3D 133 8.3. Steroid Binding Data 136 8.4. Conformer Generation Strategy 137 8.5. Activity Classification Analysis 138 8.5.1. Pharmacophores Identified 140 8.5.2. Prediction (Classification) Results 141 8.6. 3D QSAR Models 141 8.6.1. Testosterone-Binding 3D QSAR Model 142 8.6.2. Corticosteroid 3D QSAR Model 144 8.7. Comparison with other 3D QSAR Methods 145 8.8. Conclusions 147 9. Pharmacophore Models and Comparative Molecular Field Analysis (CoMFA) 151 Robert D. Clark, Joseph M. Leonard, and Alexander Strizhev 9.1. What Is CoMFA? 153 9.2. Alignment Rules 155 9.3. The s Receptor Dataset 156 9.4. Charges and Energy Minimization 159 9.5. Identifying an Initial Query 160 9.6. Refining the Query 162 9.7. Evaluating the Model 163 9.8. Applicability 166 10. HypoGen: An Automated System for Generating 3D Predictive Pharmacophore Models 171 Hong Li, Jon Sutter, and Rémy Hoffmann 10.1. Introduction 173 10.2. General Strategy 174 10.3. Methodology 175 10.3.1. Preparing to Run HypoGen 176 10.3.2. Running HypoGen 179 10.4. Case Study 179 10.5. Conclusion 187 Applications in Drug Design 191 11. Metric for Analyzing Hit Lists and Pharmacophores 193 Osman F. Güner and Douglas R. Henry 11.1. Introduction 195 11.2. Results and Discussion 196 11.2.1. Application Examples 205 11.3. Conclusions 210 12. Strategies for Database Mining and Pharmacophore Development 213 Osman F. Güner, Marvin Waldman, Rémy Hoffmann, and Jong-Hoon Kim 12.1. Introduction 215 12.2. Methods 217 12.3. Results and Discussion 217 12.3.1. Use of Query Clustering and Merging 217 12.3.2. Receptor-Based versus Ligand-Based Pharmacophore Models 221 12.3.3. Use of Shape versus Pharmacophore versus Merged Pharmacophore/Shape Queries 223 12.3.4. The Significance of Training Set Selection: Using Similar versus Diverse Compounds 227 12.3.5. Manual versus Automated Pharmacophore Model Generation 230 12.3.6. Rigid versus Flexible Searching 231 13. Pharmacophore Modeling by Automated Methods: Possibilities and Limitations 237 Morten Langgård, Berith Bjørnholm, and Klaus Gundertofte 13.1. Introduction 239 13.1.1. Pharmacophore Models 240 13.2. The Methods 241 13.2.1. Flo96 241 13.2.2. Catalyst 243 13.3. Results and Discussion 244 13.3.1. Flo96 244 13.3.2. Catalyst 246 13.4. Conclusion 249 14. Database Mining Using Pharmacophore Models to Discover Novel Structural Prototypes 251 James J. Kaminski, Dinanath. F. Rane, and Marnie L. Rothofsky 15. Predicting Drug-Drug Interactions in Silico Using Pharmacophores: Paradigm for the Next Millennium 269 Sean Ekins, Barbara J. Ring, Gianpaolo Bravi, James H. Wikel, and Steven A. Wrighton 15.1. Introduction 271 15.2. Shifting the Drug Metabolism Paradigm 273 15.3. Cytochrome P450 Pharmacophore Modeling Methodology 278 15.4. Utilizing and Interpreting Data Generated by Computational Models for Enzymes and Transporters Involved in Drug Metabolism 282 15.5. The Challenge Ahead 286 15.6. Concluding Remarks 288 16. Feature-Based Pharmacophores: Application to Some Biological Systems 301 Rémy. Hoffmann, Hong Li, and Thierry Langer 16.1. Introduction 303 16.2. First Case: TXSI-TXRA 304 16.2.1. Introduction 304 16.2.2. Methods 305 16.2.3. Results and Discussion 306 16.3. Second Case: Retinoic Acid Receptor Ligands 307 16.3.1. Introduction 307 16.3.2. Training Set and Methods 308 16.3.3. Results and Discussion 310 16.4. Third Case: Feature-Based Pharmacophores Derived from Structural Information 312 16.4.1. Introduction 312 16.4.2. Methods 313 16.4.3. Results and Discussion 315 16.4.4. Mapping the Chemical Features on the Active Conformation of Methotrexate 316 16.5. Conclusion 316 17. Receptor-Based Pharmacophore Perception and Modeling 339 C. M. Venkatachalam, Paul Kirchhoff, and Marvin Waldman 17.1. Introduction 321 17.2. Pharmacophore Definition 322 17.3. Retinoid Series 1 327 17.4. Retinoid Series 2 328 17.5. Retinoid Series 3 329 17.6. Pharmacophore Validation 331 17.7. Conclusions 333 Part III. RECEPTOR-BASED PHARMACOPHORES 337 18. Receptor-Based Pharmacophore Perception and Modeling 339 C. M. Venkatachalam, Paul Kirchhoff, and Marvin Waldman 18.1. Introduction 341 18.2. Cerius2 Structure-Based Focusing Method 342 18.2.1. Active Site 343 18.2.2. Ludi Interaction Map 344 18.2.3. Cluster Analysis of the Interaction Map 344 18.2.4. Multiples Queries 345 18.2.5. Volume Exclusions 345 18.2.6. Catalyst Database Search 346 18.3. Results and Discussion for the Estrogen Binding 346 19. Pharmacophore-Based Molecular Docking 351 Bert E. Thomas IV, Diane Joseph-McCarthy, and Juan C. Alvarez 19.1. Introduction 353 19.2. The Dock Algorithm 355 19.2.1. Conformational Ensemble Docking 356 19.3. Pharmacophore-Based Docking 357 19.3.1. Database Preparation 360 19.3.2. Site Point Generation 361 19.3.3. Validation 361 19.4. Discussion 362 Applications in Drug Design 369 20. The Use of Multiple Excluded Volumes Derived from X-Ray Crystallographic Structures in 3D Database Searching and 3D QSAR 371 Mikael Gillner and Paulette Greenidge 20.1. Introduction 373 20.2. Background 374 20.3. Methods 376 20.3.1. Catalyst Pharmacophore Construction and 3D QSAR 376 20.4. Results and Discussion 378 20.5. Conclusion 382 21. Docking-Derived Pharmacophores from Models of Receptor-Ligand Complexes 385 Renate Griffith, John B. Bremner, and Burak Coban 21.1. Introduction 387 21.2. Agonist and Antagonist Binding Site(s) on the Adrenergic Receptors 389 21.3. 3D Model Building and Docking 391 21.3.1. Revision of the Models 394 21.3.2. Docking of Adrenaline 395 21.3.3. Docking of a Rigid Antagonist (IQC) 397 21.4. Construction of Docking-Derived Pharmacophores 398 21.4.1. The a1A Pharmacophore 399 21.4.2. The a1B pharmacophore 400 21.4.3. Comparing the Docking-Derived Pharmacophores 400 21.4.4. Design of Potentially a1B-Selective Ligands 401 22. Technique for Developing a Pharmacophore Model that Accommodates Inherent Protein Flexibility: An Application to HIV-1 Integrase 409 Kevin M. Masukawa, Heather A. Carlson, and J. Andrew McCammon 22.1. Introduction 411 22.2. Computational Details 413 22.2.1. The MD Simulation and Preparation of the Available Crystal Structures 413 22.2.2. MUSIC 414 22.2.3. The Dynamic Pharmacophore Model for HIV-1 Integrase 415 22.2.4. Static Pharmacophore Models 418 22.2.5. Using the Catalyst Programs 418 22.3. Results and Discussion 420 22.4. Conclusion 423 Part IV. NEW ALGORITHMS IN PHARMACOPHORE DEVELOPMENT 429 23. Pharmacophores Derived from the 3D Substructure Perception 431 Sandra Handschuh and Johann Gasteiger 23.1. Introduction 433 23.2. General Principles of the Genetic Algorithm 434 23.2.1. Encoding of the Individuals 436 23.2.2. Optimization Criteria 437 23.2.3. The Genetic and Nongenetic Operators—A Short Description 438 23.2.4. The Pareto Fitness of Individuals 442 23.3. Matching the Conformations—Directed Tweak 444 23.4. Special Features of the Program 447 23.4.1. Close Contact Check of van der Waals Radii 447 23.4.2. Matching Criteria 447 23.4.3. Superimposition Restrictions 449 23.5. Summary 450 24. The Electron-Conformational Method of Identification of Pharmacophore and Anti Pharmacophore Shielding 455 Isaac B. Bersuker, Süleyman Bahçeci, and James E. Boggs 24.1. Introduction: Improved Definition of Pharmacophore 457 24.2. Description by EC Matrices and Pharmacophore Identification 459 24.3. Anti-Pharmacophore Shielding and Other Auxiliary Groups. Formula of Activity 462 24.4. Parameterization and Results 464 24.4.1. Angiotensin Converting Enzyme (ACE) Inhibitors 466 24.4.2. Rice Blast Activity (RBA) 470 24.5. Conclusions 472 25. Development and Optimization of Property-Based Pharmacophores 477 Ali G. Özkabak, Mitchell A. Miller, and Douglas R. Henry 25.1. Introduction 479 25.2. Flexible Substructure- or Field-Based Superposition of Structures 482 25.3. Identification of Relevant Functional and Physicochemical Property Groups 484 25.4. Generating an Initial Pharmacophore 487 25.5. Optimizing the Pharmacophore 490 25.6. Summary 493 26. Effect of Variable Weights and Tolerances on Predictive Model Generation 499 Jon Sutter, Osman Güner, Rémy Hoffmann, Hong Li, and Marvin Waldman 26.1. Background 501 26.2. Methodology 502 26.3. Cost Function 504 26.4. Case Study 506 26.5. Conclusions 510 PART V. THE FUTURE OF PHARMACOPHORE RESEARCH 513 27. Future Directions In Pharmacophore Discovery 515 John H. Van Drie 27.1. Introduction 517 27.2. Goals for a Pharmacophore Discovery Method 520 27.2.1. Objectivity 520 27.2.2. Completeness 521 27.2.3. Robustness 522 27.2.4. Computational Controls 523 27.2.5. Statistical Measures of Quality 524 27.2.6. Prospective Applications of Pharmacophores 524 27.3. Not All Datasets Are Created Equal 525 27.4. Conformational Analysis 526 27.5. Frontiers 527 Index 531 |
|
| Preface
Perceiving a pharmacophore is the most important first
step towards understanding the interaction between a receptor and a
ligand. In the early 1900s, Paul Ehrlich offered the first definition
for a pharmacophore: "a molecular framework that carries (phoros) the
essential features responsible for a drug’s (pharmacon) biological
activity" (Ehrlich P: Dtsch Chem Ges
1909, 42:17). That definition
of a pharmacophore remained unperturbed for over 90 years. The current
widely used definition was presented by Peter Gund in 1977: “a set of
structural features in a molecule that is recognized at a receptor site
and is responsible for that molecule's biological activity” (Gund
P: Prog Mol Subcell Biol 1977, 5:117-143).
This modern definition is remarkably loyal to the earliest definitions.
It is only appropriate, then,
that this books starts with the “Evolution of the Pharmacophore Concept
in Pharmaceutical Research” in Chapter 1, written by Peter Gund who
also developed the first 3D searching software, Molpad, and the first
ideas for computational pharmacophores.2 It took 15 years for the
commercial 3D searching software to become available following its
original publication by Gund, Langridge, and Wipke
(Gund P, Wipke WT, Langridge R: Proc Intl Conf on Computers in Chem Res
and Educa, Ljubljana, 1973:5/33). John Van Drie, on the other
hand, was involved with the development of two commercial 3D searching
software applications: one of the earliest ones, Aladdin, and the
latest one, Catalyst®. He has also been active and forward-looking in
his work with pharmacophores as, for example, “shrink-wrap
pharmacophores,” (Van Drie J: J Chem
Inf Comput Sci 1997, 37:38-42)
and in his program Dante (Van Drie J: J
Comput-Aided Mol Des 1997, 11:39-52).
It is only appropriate, then, to end this book with Chapter 27 on the
“Future Directions in Pharmacophores Discovery,” written by John Van
Drie.
The information provided in between these two chapters is pretty much everything you would want to know about pharmacophores, and probably more. The Part II of the book is
dedicated to analog-based pharmacophores. Your editor, Osman Güner,
presents an elementary introduction to the concept of pharmacophores in
Chapter 2, entitled “Manual Pharmaco-phore Generation: Visual Pattern
Recognition.” The analog-based pharmacophore development is then
introduced with the description of the Active Analog Approach by Denise
Beusen and Garland Marshall in Chapter 3: “Pharmacophore Definition
Using the Active Analog Approach.”
This information is followed by a detailed description in Chapters 4 through 10 of various pharmacophore development and 3D-QSAR software by the pioneers in this area, their inventors, and current developers. Yvonne Martin describes and critically evaluates one of the earlier pharmacophore development software products, DISCO, in Chapter 4 entitled “DISCO: What We Did Right and What We Missed.” Chapter 5 describes a common-feature based alignment software, “HipHop: Phar-macophores Based on Multiple Common-Feature Alignments,” written by Omoshile Clement and Adrea Mehl. Gareth Jones, Peter Willett, and Robert Glen describe one of the recent alignment and pharmacophore software in Chapter 6: “GASP: Genetic Algorithm Superimposition Program.” Stephen Cato then describes how to perceive pharmacophores in Chapter 7 with “Exploring Pharmacophores with Chem-X.” Moving onto the predictive model generation software, Erich Vorpagel and Valery Golander describe the importance of negative activities in Chapter 8 entitled “Apex-3D: Activity Prediction Expert System with 3D-QSAR.” Robert Clark, Joseph Leonard, and Alexander Strizhev emphasize the importance of molecular alignment in Chapter 9: “Pharmacophore Models and Comparative Molecular Field Analysis (CoMFA).” Finally, Hong Li, Jon Sutter, and Rémy Hoffmann bring the description of the methodology under Catalyst/HypoGen in Chapter 10 entitled “HypoGen: An Automated System for Generating 3D Predictive Pharmacophore Models.” Chapters 11 through 17
involve various applications of analog-based pharmacophores and success
stories. Chapter 11 introduces some scoring techniques for
pharmacophores and hit lists including “Metric for Ana-lyzing Hit Lists
and Pharmacophores,” written by Osman Güner and Douglas Henry.
Different database querying strategies are introduced by Osman Güner,
Marvin Waldman, Rémy Hoffmann, and Jong-Hoon Kim in Chapter 12 entitled
“Strategies in Database Mining and Pharmacophore Development.”
Evaluation of automated methods is presented in Chapter 13,
“Pharmacophore Modeling by Automated Methods: Possibil-ities and
Limitations,” by Morten Langgård, Berith Bjørnholm, and Klaus
Gundertofte. A successful example of the identification of novel
structures is provided by James Kaminski, Dinanath Rane, and Marnie
Rothofsky in Chapter 14 entitled “Database Mining Using Pharmaco-phore
Models to Discover Novel Structural Prototypes.” Other successes with
pharmacophores are described in Chapters 15 through 17 with “Predicting
Drug-Drug Interactions in Silico using Pharmacophores: A Paradigm for
the Next Millennium” by Sean Ekins, Barbara Ring, Gianpaolo Bravi,
James Wikel, and Steven Wrighton. “Feature-Based Pharmacophores:
Applications to Some Biological Systems” by Rémy Hoffmann, Hong Li, and
Thierry Langer; and “Pharmacophore Defini-tion of Retinoid-X-Receptor
Specific Ligands” by Steven K. White.
Part III deals with
more recent ideas on receptor-based pharmacophores. It starts with
“Receptor-Based Pharmacophore Perception and Modeling” in Chapter 18 by
Venkatachalam, Paul Kirchhoff, and Marvin Waldman, Followed by Chapter
19 where Bert Thomas IV, Diane Joseph-McCarthy, and Juan Alvarez
describe “Pharmacophore-Based Molecular Docking.” Successful
applications of receptor-based pharmacophores are presented in the next
two Chapters 20 and 21: “Pharmacophores Including Multiple Excluded
Volumes Derived from X-Ray Crystallographic Structures of Nuclear
Receptors: Their Application in 3D Database Searching and 3D-QSAR” by
Mikael Gillner and Paulette Greenidge, and “Docking-Derived
Pharmacophores from Models of Receptor-Ligand Complexes” by Renate
Griffith, John Bremner, and Burak Coban. The complications arising from
the flexibility of the receptor structure is covered by Kevin Masukawa,
Heather Carlson, and Andrew McCammon in Chapter 22 entitled “Technique
for Developing a Pharmacophore Model That Accommodates Inherent Protein
Flexibility: An Application to HIV-1 Integrase.”
Part IV provides new
ideas and algorithms in pharmacophore development. It starts with a
contribution from Sandra Handschuh and Johann Gasteiger in Chapter 23
entitled “Pharmacophores Derived from the 3D Substructure Perception.”
Isaac Bersuker, Süleyman Bahçeci, and James Boggs present “The
Electron-Conformational Method of Identification of Pharmacophore and
Anti-Pharmacophore Shielding” a novel perspective towards pharmacophore
identification. Ali Özkabak, Mitchell Miller, Douglas Henry, and Osman
Güner discuss the concept of pharmacophore optimization in Chapter 25,
“Development and Optimization of Property-Based Pharmacophores.” An
enhancement to predictive pharmacophore model generation, “Effect of
Variable Weights and Tolerances on Predictive Model Generation,” is
introduced in Chapter 26 by Jon Sutter, Osman Güner, Rémy Hoffmann,
Hong Li, and Marvin Waldman
Finally, Chapter 27
provides a closing with John Van Drie's perception of the future
directions in this area.
If you are new to this
area, you should start with Chapters 1, 2, 3, and 27 for an
introduction; then move on to the desired software discussions, DISCO
at 4, HipHop at 5, GASP at 6, Chem-X at 7, Apex-3D at 8, CoMFA at 9,
and HypoGen at 10. If you feel you still need to be persuaded that
these approaches are effective, you will want to read Chapters 14, 15,
16, and 17 for some real-world success stories.
Seasoned database
searchers who want to improve their skills should first understand the
limitations outlined in Chapter 26, and then enrich their portfolio of
different querying techniques with Chapter 25. Finally, they can learn
how to analyze their hit lists by reading Chapter 11.
If the receptor structure is
available and you want to use it to improve your pharmacophore models,
different techniques in this area are described in Chapters 18 and 19
and successful applications of receptor-based pharmacophores are
presented at sections 20 and 21. You should also read chapter 22 to
appreciate the conformational flexibility of the receptor structure and
its impact on pharmacophore models.
If you are involved in
development of software tools in this area, several new ideas and
algorithms are detailed in Chapters 23, 24, and 26. If you are
interested in automating the pharmacophore optimization process,
Chapter 25 provides good ideas and Chapter 11 provides some scoring
functions that can be used for this purpose.
To get the full picture of
pharmacophores, it is always good to go back and read Chapters 1 and 27
for a historical perspective and future directions.
In closing, consider
the history of aviation with the very first flight taking place in the
early 1900s and, the moon landing a mere 50 years later. Contrast this
to the first use of the term “pharmacophore.” It was first used in
early 1900s as Peter Gund explains in Chapter 1; however, the meaning
of the definition remained remarkably unperturbed during its close to
90 years of existence. Today, pharmacophores are considered one of the
most important types of “information” that can be obtained from
receptor-ligand interactions. Yet, quite surprisingly, this book is the
first book that has the word “pharmacophore” in its title. We therefore
wanted to be very comprehensive in this first volume, covering all
aspects of pharmacophore perception, development, and use in drug
design. We hope that you will find this book useful to bring your
computer-aided drug design endeavor to a higher level.
Happy discoveries… |
|