The Generation and Use of Large 3D Databases in Drug Discovery

Paul S. Charifson, Andrew R. Leach, and Andrew Rusinko III

Glaxo Wellcome Inc.
Five Moore Drive
Research Triangle Park, NC 27709 USA

and

Gunnels Wood Road
Stevenage, Hertfordshire
SG1 2NY, UK.

http://www.netsci.org/Science/Cheminform/feature03.html

Introduction

Pick up any recent issue of the Journal of Chemical Information and Computer Sciences or the Journal of Computer-Aided Molecular Design and you will find at least one paper which either directly or indirectly addresses the topic of 3D databases. The reason for this is that searching 3D databases has proven to be a viable method for identifying novel molecular entities based upon an initial query. Queries can range from simple ones, such as the distance between a pair of atoms or a pharmacophoric model, to the more complex steric and electrostatic complementarity defined by a protein binding site. Reviews covering accomplishments in this field through 1992 are included in references [1-3].

The current state of 3D databases and their uses is still an evolving research area. Recent discussions in the literature address optimization of such details as: query definitions [4], partial query solutions [5],conformational flexibility [6-13], and conformational diversity [14-16]. Searching 3D databases has led to some clear successes in both the pharmacophore-based searching arena [17-20], as well as in the molecular docking arena [21-24], and we expect this trend to increase dramatically in the immediate future. It is not the intent of this paper to address De Novo Design approaches which utilize searching 3D databases to identify molecular fragments, scaffolds, bridging functionalities [25-29] or substructural frequency occurrences [30]. It should be noted, however, that some of these De Novo methods have also been used for more pharmacophore-based 3D database searches [31] and molecular docking, as well [32].

Generation of 3D Databases

The generation of any three dimensional database should be determined by the intended use and search strategy. One can categorize 3D database search methods into three groups: those that represent and search each molecule as a single conformer [5,33-37], those that store multiple conformers of each molecule [12,15,16,31], and those that perform "on-the-fly" conformer generation [8,9] as the search is being performed. There also exist some methods which store only a single conformer but either perform a limited conformational search at run time [7,10], or access stored distance-based search keys describing the conformational space available to a given molecule based upon a prior conformational search [6]. The advantages and disadvantages of each of these approaches have been discussed previously [1-3,35,36]. For 3D database searches involving either stored single or multiple conformers, a variety of methods exist for the generation of these molecules. Reviews on this topic include references 3 and 38-41. In the case of multiple conformer generation, one could attempt to generate all of the possible conformers corresponding to energy minima and then cluster them into conformational families, but this is inefficient. Moreover, the active conformation may not be at an energy minimum for the isolated molecule. One potential drawback with storing multiple conformations is that many of the conformations may be very similar. This can be problematic, particularly for pharmacophore-based queries where much of the conformational variation may reside in parts of the molecule removed from the areas of interest. One approach that can overcome this limitation is to directly generate inherently "dissimilar" conformers during the conformational search process [14-16].

A major limitation in most 3D search systems is the choice of the representation of the structure stored in the database. Obviously, the molecular geometry will be affected by inclusion of stereochemical information wherever possible. Many structures, however, were entered into electronic databases long before chiral synthesis or separation techniques were routinely available. Unavailable stereochemical information can lead to the generation of irrelevant 3D models. For example, very few of the synthetic amino acids registered should be modeled in the D-form. Yet it is quite possible with current software tools to build equal numbers of models in either the D- or the L- form for a given peptide. A second potential pitfall is the choice of ionization state and/or tautomeric form of a molecule when stored electronically. Potential hydrogen bond donor or acceptor capabilities as well as conformations selected can be affected by starting from different "official" stored representations. Since most structural databases were designed for 2D substructure searching convenience, their 2D representations may not always reflect the best place to start for a 3D search. In fact, irrelevant and occasionally higher energy tautomeric forms used as starting points for 3D searches may contribute to the large number of false positives ("hits" that are not active) and false negatives (active compounds that didn't match the query) found in a typical 3D substructure search. One solution to this problem is to generate all possible tautomeric forms for several different environments (i.e. change the ionization state of the structure to reflect varying pH) and determine if these models are significantly different from each other. After analysis of this data, subsequent 3D substructure searches would be performed only on the more relevant models. Hopefully, this will produce fewer false positives and negatives.

Our Experiences

We are interested both in traditional pharmacophore-based 3D searches as well as searching 3D databases in the context of the binding site of a given structural target (i.e. molecular docking). Initially, these were considered to be separate problems but the increasing availability of structural information of biological targets (via X-ray crystallography or NMR) has encouraged the development of methods that can perform a pharmacophore-based search in the context of a binding site. In such programs, the binding site is often represented as a collection of "exclusion spheres", which are regions of space where no atom of the ligand may be placed. The incorporation of exclusion spheres into a query fits naturally into the minimization protocols that many of the commercial and proprietary 3D database search systems employ [4,7,10,34,37]. The use of exclusion spheres does mean that a significant level of detail about the binding site is lost. However, even this crude representation may be useful in eliminating molecules that would interact unfavorably with the binding site. Given the current speed of such algorithms, it is not generally feasible to consider all possible pharmacophores that could be derived from the macromolecular target structure. The user must base the search upon prior knowledge of the binding mode of an existing inhibitor or a careful analysis of the binding site.

It might be considered that performing a conformational search of all database ligands (or those that pass some type of initial screen) in the confines of a putative binding site would tend toward the direction of increased rigor; ultimately both ligand and macromolecule flexibility will hopefully be addressed. Although a few researchers are actively pursuing this type of approach with respect to ligand flexibility in a static binding site [42-47], and with partial receptor flexibility [48,49], it is still not routinely possible to accomplish this task given the size of most pharmaceutical compound collections; this will likely change within the next few years.

There have been several discussions presented in the past which have given reasons for the opposition to the generation and storage of multiple conformers.[1-3]. Although some of these arguments are correct, (especially within the context of pharmacophore-based searches) it is clear for the case of docking ligands to a structural target, that maximally-dissimilar multiple conformers of "good" quality should have a higher probability of producing a "hit" over a single conformer [12]. Multiple conformer databases have many uses in a drug discovery paradigm. Properties such as molecular surface area and volume, ionization potential, etc. [50] can only be computed from a fixed geometry. A representative set of conformers would yield a better, or at least a more comprehensive set of results from querying molecular property databases. Shape and property similarity [51-56] can be used to determine potential new leads for biological screening. Certainly, the strongest arguments against storing multiple conformers based upon storage requirements are no longer valid; gigabytes of local disk storage are now reasonably priced.

 

Our philosophy, which is similar to that of the Merck group [12], is to produce a multiple conformer database of maximally dissimilar representations of our compound collection for use with the programs, DOCK [56] and CAVEAT [26]. We use a combined rule-based distance geometry and molecular mechanics refinement approach to this end. Our generation process begins by conversion of MACCS SD files to Daylight isomeric SMILES strings which retain local stereochemical information. This is accomplished via a utility program provided by DAYLIGHT CIS, Inc. We then employ the program, Rubicon [57] (Rule-Based Invention of Conformers) to convert our single 2D molecular representation into multiple 3D conformers per molecule. We limit the maximum number of conformers generated to be 20 (although we have generated other databases with up to 100 conformers per molecule). Rubicon is fundamentally a distance geometry program which allows the user to define the upper and lower bounds for each molecule's distance matrix via a set of "rules" which can be either experimentally or empirically-derived. Additionally, the user can define a variable which controls the degree of similarity (in terms of the distance matrix) between conformers which have been generated and those in the process of being considered; this feature can be used to ensure a conformationally diverse generation process. After conformer generation, we perform some checks to ensure that all bond lengths are reasonable and that groups which are likely ionized at physiological pH are identified. We then perform 50 steps of conjugate gradient minimization using the OPLS [58] forcefield within Batchmin [59] to which several types of missing organic functionality parameters have been added. During the energy refinement process, we employ the GB/SA continuum solvation model [60] to obtain solvation free energy estimates for each conformer which can later be used in certain scoring functions [61], if desired. We also keep track of the conformational strain energy of each conformer relative to the lowest energy conformer identified. This can also be a useful measure in the ultimate selection of "hits" after the docking process. Finally, our database is converted into DOCK2.3 or DOCK3.5 database formats, as well as CAVEAT vector database format. The DOCK databases are generally split into several pieces for running across multiple cpu's. In one attempt at such a 3D database construction, we successfully converted 103,400 SMILES strings to 1,132,000 3D structures.This required 1.75 Gb of disk storage for the ultimate DOCK2.3 databases and 1.5 Gb for the CAVEAT vector database.

Examples

Although it is typical to use the program DOCK to identify ligands which possess some degree of geometric, steric, and electrostatic complementarity to a putative binding site, it is also possible to use DOCK as a general purpose 3D database similarity searching tool [62]. This is referred to as the "positive docking" mode or "similarity docking" mode. In this approach to using DOCK, one attempts to match "like" ligand atom types with "like" sphere types. This is different from the traditional chemical matching in DOCK 3.5 in which one attempts to match complementary sphere and ligand atom types. Initial force field-based scoring is accomplished via the use of a grid-based energy evaluation in which the repulsive part of the VDW potential is set to zero and the sign of the electrostatic interaction term is reversed. Additionally, a limit is set on the maximum absolute magnitude of the electrostatic interaction energy that a ligand atom can receive as well as a minimum VDW interaction energy that a ligand atom can receive. Because these scores have no physical meaning, they are used only as an initial filter to determine the top N molecules which get written to disk. We then typically rescore this list via a simple fitness function in which the goodness of fit for like atom-types superimposed upon like-sphere types is evaluated.

A successful example of the use of this approach combined with our multiconformer database was in the identification of ligands that might bind to the matrix metalloproteinase, collagenase. In attempting to identify nonpeptide inhibitors based upon a portion of one of our known carboxyalkylamino inhibitors [63] (Figure 1), we performed a similarity docking run to this pharmacophore. A portion of our multiconformer database was considered which consisted of 516,485 conformers that represented 64,833 discrete molecules. For the chemical matching definitions, we employed 6 sphere and ligand atom types (positive, negative, donor, acceptor, neutral and polar hydrogen). The top 600 "ligands" based upon the grid-based energy scoring scheme described above were identified and rescored. 47 compounds were tested in a competitive binding assay and one compound was found to inhibit collagenase with an IC50 of 2.8 micromolar. This compound is shown superimposed with the initial pharmacophore query (Figure 2). This anthraquinone, alizarin complexone, compound was only ranked 478 out of the top 600 via the force field-based scoring scheme. However, upon rescoring, this compound moved up to 33 out of 600. While it is difficult to generalize from a single instance, it is interesting to note that the single CONCORD [64] generated conformer did not make it into the initial top 600 list based upon the force field scores. This result tends to agree with references 12 and 13 in the sense that even a modest multiconformer database will improve the hit rate over single conformer 3D databases.

Another example involving the use of the multiconformer database in conjunction with DOCK 3.5 in the more traditional molecular docking mode centers around the threonine kinase, cyclin-dependent kinase 2 (CDK2).We were interested in identifying ligands which were capable of binding to the ATP site in CDK2 [65]. In this case we docked 324,800 conformers representing 44,285 discrete molecules to the ATP binding site in CDK2. For the chemical matching definitions, we employed 5 sphere and ligand atom types (positive, negative, donor, acceptor, and neutral). Prior to scoring, a grid-based simplex minimization was employed for some minor rigid-body refinement. The top 400 "ligands" based upon the grid-based energy scoring scheme were identified and subjected to a further quasi-Newton rigid-body minimization. One compound from this list was found to possess an IC50 of 10 micromolar in a competitive assay with ATP. This flavonoid compound, chrysin (Figure 3), was surprisingly ranked 400 out of the top 400 after the initial simplex minimization, and 396 out of the top 400 after the quasi-Newton minimization. It has been reported that flavonoids are non-selective inhibitors of protein kinases [66,67] and are competitive with respect to ATP. Although DOCK was able to identify a known class of CDK2 inhibitor, we have reason to believe that the flavone ring is oriented approximately 90 degrees off from what one would expect in the context of maximizing hydrogen bonds with the N-terminal beta sheet, B5 [68].

Future Prospects

The future of 3D database searching appears to be quite promising. Representation of the molecules in databases have shifted from explicit definition of atoms [34] to a more fuzzy atom type description [4]. In theory, these more general atom types should have the effect of increasing the number of new classes of hits discovered via a 3D search. Techniques for generating a pharmacophoric model have improved considerably in the last few years[69-71].

Thus, pharmacophoric models can be developed faster and more reliably. Software tools for analyzing the results of 3D searches, however, have yet to make a dramatic impact. Since many 3D pharmacophore-based queries yield hundreds to thousands of "hits", scoring functions such as the one embedded in DOCK need to be further improved to rank these hits in a useful fashion. Techniques capable of performing a pharmacophore-based search in the context of a binding site followed by energy minimization should increase the likelihood of finding new leads. Furthermore, a new generation of techniques must be developed to analyze the results of 3D searches when no target structural information is available. Tools, such as the one developed by Martin, et al. [72] to determine a unique core of a set of hits, aid in the discovery of the fundamental structural components that are required for activity.

The continued improvement in certain experimental methods on the one hand (e.g. combinatorial libraries) combined with an increasing number of macromolecular structures and computational developments on the other (improved hardware and new algorithms) will ultimately allow for 3D-database search techniques to quickly identify new classes of feasible lead compounds.Two key challenges which remain are concerned with conformational flexibility (of BOTH the ligand and the macromolecular target) and the development of scoring functions that can rapidly and accurately assess the free energy of binding of the candidate structures.

 

[Figure 1]

Figure 1
Carboxyalkylamino inhibitor of collagenase: atoms used to define the pharmacophore query are colored red.

[BTN]Return to the paper

 

[Figure 2]

Figure 2
Result of positive DOCK similarity search: anthraquinone compound shown inset and superimposed on original carboxyalkylamino inhibitor query. Conformation of carboxyalkylamino inhibitor taken from reference 63. Carbon atoms of the carboxyalkylamino inhibitor are green and carbon atoms of the anthraquinone, alizarin complexone, are yellow.


[BTN]Return to the paper

 

[Figure 3]

Figure 3
Docked orientation of the flavone, chrysin in ATP binding site of the cyclin-dependent kinase 2. The chemical structure of chrysin is also depicted in the inset.


[BTN]Return to the paper

Footnotes



 

1. a. Martin, Y.C., Bures, M.G., Willett, P. Searching Databases of Three- Dimensional Structures. In Rev. in Computational Chemistry; Lipkowitz, K., Boyd, D., Eds.; VCH Publishers, Inc.: New York, 1990; Vol. I, pp.213-263. b. Willett, P., in Three-Dimensional Chemical Structure Handling: Research Studies Press Ltd.: Taunton, Somerset, England, 1991.

2. Martin, Y.C., 3D Database Searching in Drug Design. J. Med. Chem., 1992, 35, 2145-2154.

3. Pearlman, R.S., 3D Molecular Structures: Generation and Use in 3D Searching, In 3D QSAR in Drug Design: Theory, Methods, and Applications; Kubinyi, H., Ed.; ESCOM Science Publishers: Leiden, 1993; pp. 41-79.

4. Greene, J., Kahn, S., Savoj, H., Sprague, P., Teig, S. Chemical Function Queries for 3D Database Search. J. Chem. Inf. Comput. Sci., 1994, 34, 1297-1308.

5. Ho, C.M.W., Marshall, G.R., FOUNDATION: A Program to Retrieve All Possible Structures Containing a User-Defined Minimum Number of Matching Query Elements From Three-Dimensional Databases, J.Comput.-Aided Mol. Des., 1993, 7, 3-22.


[BTN]Return to the paper

 

6. Murrall, N.W., Davies, E.K., Conformational Freedom in 3-D Databases. 1. Techniques, J. Chem. Inf. Comput. Sci., 1990, 30, 312-316.

7. Guner, O.F., Henry, D.R., Pearlman, R.S., Use of Flexible Queries for Searching Conformationally Flexible Molecules in Databases of Three Dimensional Structures, J. Chem. Inf. Comput. Sci., 1992, 32, 101-109.

8. Clark, D.E., Willett, P., Kenny, P.W., Pharmacophoric Pattern Matching in Files of Three-Dimensional Chemical Structures: Use of Bounded Distance Matrices for the Representation and Searching of Conformationally Flexible Molecules, 1992, 10, 194-204.

9. Clark, D.E., Willett, P., Kenny, P.W., Pharmacophoric Pattern Matching in Files of Three-Dimensional Chemical Structures: Implementation of Flexible Searching. J. Mol. Graphics, 1993, 11, 146-156.

10. a. Hurst, T., Flexible 3D Searching: The Directed Tweak Technique. J. Chem. Inf. Comput. Sci., 1994, 34, 190-196. b. Moock, T.E., Henry, D.R., Ozkabak, A.G., Alamgir, M. Conformational Searching in ISIS/3D Databases. J. Chem. Inf. Comput. Sci., 1994, 34, 184-189.

11. Clark, D.E., Jones, G., Willett, P., Pharmacophoric Pattern Matching in Files of Three-Dimensional Chemical Structures: Comparison of Conformational Searching Algorithms for Flexible Searching. J. Chem. Inf. Comput. Sci., 1994, 34, 197-206.

12. a. Kearsley, S.K., Underwood, D.J., Sheridan, R.P., Miller, M.D., Flexibases: A Way to Enhance the Use of Molecular Docking Methods. J.Comput.- Aided Mol. Des., 1994, 8, 565-582. b. Miller, M., Kearsley, S.K., Underwood, D.J., Sheridan, R.P., FLOG: A System to Select 'Quasi-Flexible' Ligands Complementary to a Receptor of Known Three-Dimensional Structure, J.Comput.- Aided Mol. Des., 1994, 8, 153-174.

13. Haraki, K.S., Sheridan, R.P., Venkataraghavan, R., Dunn, D.A., McCulloch, R., Looking for Pharmacophores in 3-D Databases: Does Conformational Searching Improve the Yield of Actives ?, Tetrahedron Comput. Methodol., 1990, 3(6c), 565-573.

14. Leach, A.R., An Algorithm To Directly Identify a Molecule's "Most Different" Conformations, J. Chem. Inf. Comput. Sci., 1994, 34, 661- 670.

15. Smellie, A., Kahn, S.D., Teig, S.L., Analysis of Conformational Coverage. 1. Validation and Estimation of Coverage, J. Chem. Inf. Comput. Sci., 1995, 35, 285-294.

16. Smellie, A., Kahn, S.D., Teig, S.L., Analysis of Conformational Coverage. 1. Application of Conformational Models, J. Chem. Inf. Comput. Sci., 1995, 35, 295-304.

17. Bures, M.G., Hutchins, C.W., Maus, M., Kohlbrenner, W., Kadam, S., Erickson, J.W., Using Three-Dimensional Substructure Searching to Identify Novel NonPeptidic Inhibitors of HIV-1 Protease, Tetrahedron Comput. Methodol., 1990, 3(6c), 673-680.

18. Bures, M.G., Black-Schaefer, C., Gardner, G., The Discovery of Novel Auxin Transport Inhibitors by Molecular Modeling and Three- Dimensional Pattern Analysis, J.Comput.-Aided Mol. Des., 1991, 5, 323-334.

19. Milne, G.W.A., Nicklaus, M.C., Driscoll, J.S., Wang, S., Zaharevitz, D., National Cancer Institute Drug Information System 3D Database, J. Chem. Inf. Comput. Sci., 1994, 34, 1219-1224.

20. Lam, P.Y.S., et al., Rational Design of Potent, Bioavailable, Nonpeptide Cyclic Ureas as HIV Protease Inhibitors, Science, 1994, 263, 380-384.

21. DesJarlais, R.L., Seibel, G.L., Kuntz. I.D., Furth, P.S., Alvarez, J.C., Ortiz de Montellano, P.R., DeCamp, D.L., Babe, L.M., Craik, C.S., Structure-Based Design of Nonpeptide Inhibitors Specific for the Human Immunodeficiency Virus 1 Protease, Proc. Natl. Acad. Sci. USA, 1990, 87, 6644-6648.

22. Shoichet, B.K., Stroud, R.M., Santi, D.V., Kuntz, I.D., Perry, K.M., Structure-Based Discovery of Inhibitors of Thymidylate Synthase, Science, 1993, 259, 1445-1450.

23. Bodian, D.L., Yamasaki, R.B., Buswell, R.L., Stearns, J.F., White, J.M., Kuntz, I.D., Inhibition of the Fusion-Inducing Conformational Change of Influenza Hemagglutinin by Benzoquinones and Hydroquinones, Biochemistry, 1993, 32, 2967-2978.

24. Ring, C.S., Sun, E., Mckerrow, J.H., Lee, G.K., Rosenthal, P.J., Kuntz, I.D., Cohen, F.E., Structure-Based Inhibitor Design by Using Protein Models for the Development of Antiparasitic Agents, Proc. Natl. Acad. Sci. USA, 1990, 90, 3583-3587.

25. Bohm, H.-J., LUDI: Rule-Based Automatic Design of New Substituents for Enzyme Inhibitor Leads, J.Comput.-Aided Mol. Des., 1992, 6, 593- 606.

26. a. Bartlett, P.A., Shea, G.T., Telfer, S.J., Waterman, S., In Molecular Recognition in Chemical and Biological Problems, S.M.Roberts, Ed. (Royal Chemical Society, London, 1989) Vol. 78, pp.182-196. b. Bartlett, P.A., et al., in Chemistry at The Frontiers of Medicine (Proceedings of the Robert A. Welch Foundation Conference on Chemical Research XXXV, 1992), pp.45-68.

27. Tschinke, V., Cohen, N.C., The NEWLEAD Program: A New Method for the Design of Candidate Structures from Pharmacophoric Hypotheses, J. Med. Chem., 1993, 36, 3863-3870.

28. Ho, C.M.W., Marshall, G.R., SPLICE: A Program to Assemble Partial Query Solutions from Three-Dimensional Database Searches into Novel Ligands, J.Comput.-Aided Mol. Des., 1993, 7, 623-647.

29. a. Lewis, R.A., Roe, D.C., Huang, C., Ferrin, T.E., Langridge, R., Kuntz, I.D., Automated Site-Directed Drug Design Using Molecular Lattices, J. Mol. Graphics., 1992, 10, 66-78. b. Roe, D.C., Kuntz, I.D., BUILDER v.2: Improving the Chemistry of a De Novo Design Strategy, J.Comput.- Aided Mol. Des., 1995, 9, 269-282.


[BTN]Return to the paper

 

30. a. Chau, P.-L., Dean, P.M., Automated Site-Directed Drug Design: The Generation of a Basic Set of Fragments to be Used for Automated Structure Assembly, J.Comput.-Aided Mol. Des., 1992, 6, 385-396. b. Chau, P.-L., Dean, P.M., Automated Site-Directed Drug Design: Searches of the Cambridge Structural Database for Bond Lengths in Molecular Fragments to be Used for Automated Structure Assembly, J.Comput.- Aided Mol. Des., 1992, 6, 397-406. c. Chau, P.-L., Dean, P.M., Automated Site- Directed Drug Design: An Assessment of the Transferability of Atomic Charges (CNDO) for Molecular Fragments, J.Comput.-Aided Mol. Des., 1992, 6, 407- 426.

31. Charifson, P.S., The Generation and Use of Large 3D Databases of Molecular Conformations in Drug Discovery, presented at the 208th ACS Meeting, Washington, D.C., August 1994. The Use of CAVEAT as a general purpose 3D database searching tool was also discussed.

32. Bohm, H.-J., On the Use of LUDI to Search the Fine Chemicals Directory for Ligands of Proteins of Known Three-Dimensional Structure, J.Comput.-Aided Mol. Des., 1994, 8, 623-632.


[BTN]Return to the paper

 

33. a. Rusinko, A. III, Sheridan, R.P, Nilakantan, R., Haraki, K.S., Bauman, N., Venkataraghavan, R, Using CONCORD to Construct a Large Database of Three-Dimensional Coordinates from Connection Tables, J. Chem. Inf. Comput. Sci., 1989, 29, 251-255. b. W. Fisanick, K.P. Cross, J. Christopher Forman and A. Rusinko III, An Experimental System for Similarity and 3D Searching of CAS Registry Substances. I. 3D Substructure Searching, J. Chem. Inf. Comp. Sci., 1993, 33(4), 548-59.

34. Sheridan, R.P, Nilakantan, R., Rusinko, A. III, Bauman, N., Haraki, K.S., Venkataraghavan, R, 3DSEARCH: A System for Three- Dimensional Substructure Searching, J. Chem. Inf. Comput. Sci., 1989, 29, 255-260.

35. Sheridan, R.P, Rusinko, A. III, Nilakantan, R., Venkataraghavan, R, Searching for Pharmacophores in Large Coordinate Databases and its use in Drug Design, Proc. Natl. Acad. Sci. USA, 1989, 86, 8165-8169.

36. Martin, Y.C., Danaher, E.A., May, C.S., Weininger, D., MENTHOR, a Database System for the Storage and Retrieval of Three Dimensional Molecular Structures and Associated Data Searchable by Substructural, Biologic, Physical, or Geometric Properties, J.Comput.-Aided Mol. Des., 1988, 2, 15-29.

37. Van Drie, J.H., Weininger, D., Martin, Y.C., ALADDIN: An Integrated Tool for Computer-Assisted Molecular Design and Pharmacophore Recognition From Geometric, Steric, and Substructure Searching of Three- Dimensional Molecular Structures, J.Comput.-Aided Mol. Des., 1989, 3, 225-251.

38. Ricketts, E.M., Bradshaw, J., Hann, M., Hayes, F., Tanna, N., Comparison of Conformations of Small Molecule Structures From the Protein Data Bank with Those Generated by Concord, Cobra, ChemDBS-3D, and Converter and Those Extracted from the Cambridge Structural Database, 1993, 33, 905-925.

39. Sadowski, J., Gasteiger, J., From Atoms and Bonds to Three- Dimensional Atomic Coordinates: Automatic Model Builders, Chem. Rev., 1993, 93, 2567- 2581.

40. Sadowski, J., Gasteiger, J., Comparison of Automatic Three- Dimensional Model Builders Using 639 X-ray Structures, J. Chem. Inf. Comput. Sci., 1994, 34, 1000-1008.

41. Leach, A.R., A Survey of Methods for Searching the Conformational Space of Small and Medium-Sized Molecules, In Rev. in Computational Chemistry; Lipkowitz, K., Boyd, D., Eds.; VCH Publishers, Inc.: New York, 1991; Vol. II, pp. 1-55.


[BTN]Return to the paper

 

42. DesJarlais, R.L., Sheridan, R.P., Dixon, J.S., Kuntz, I.D., Venkataraghavan, R., Docking Flexible Ligands to Macromolecular Receptors by Molecular Shape, J. Med. Chem., 1986, 29, 2149-2153.

43. Goodsell, D.S., Olson, A.J., Automated Docking of Substrates to Proteins by Simulated Annealing, PROTEINS: Struct. Funct. Genet., 1990, 8, 195-202.

44. Ripka, W.C., Blaney, J.M., Computer Graphics and Molecular Modeling in the Analysis of Synthetic Targets, Top. Stereochem., 1991, 20, 1- 85.

45. Smellie, A.S., Crippen, G.M., Richards, W.G., Fast Drug-Receptor Mapping by Site-Directed Distances: A Novel Method of Predicting New Pharmacological Leads, J. Chem. Inf. Comput. Sci., 1991, 31, 386-392.

46. a. Dixon, J.S., In Trends in QSAR and Molecular Modeling, Wermuth, C.G., Ed.; ESCOM; Leiden, 1993, pp.412-413. b. Oshiro, C.M., Kuntz, I.D., Dixon, J.S., Flexible Ligand Docking Using a Genetic Algorithm, J.Comput.- Aided Mol. Des., 1995, 9, 113-130.

47. Leach, A.R., Kuntz, I.D., Conformational Analysis of Flexible Ligands in Macromolecular Receptor Sites, J. Comp. Chem., 1992, 13, 730-748.

48. Leach, A.R., Ligand Docking to Proteins with Discrete Side-chain Flexibility, J. Mol. Biol., 1994, 235, 345-356.

49. Jones, G., Willett, P., Glen, R.C., Molecular Recognition of Receptor Sites Using a Genetic Algorithm with a Descriptor of Desolvation, J. Mol. Biol., 1995, 245, 43-53.


[BTN]Return to the paper

 

50. W. Fisanick, K.P. Cross, and A. Rusinko III, Characteristics of Computer- Generated 3D and Related Molecular Property Data for CAS Registry Substances, Tetrahedron Computer Methodology, 1990, 3(6C), 635.

51. Kearsley, S.K. and Smith, G.M., An Alternative Method for the Alignment of Molecular Structures: Maximizing Electrostatic and Steric Overlap, Tetrahedron Computer Methodology, 1990, 3(6C), 615-633.

52. Nilakantan, R., Bauman, N. and Venkataraghavan, R., New Method for Rapid Characterization of Molecular Shapes: Application in Drug Design, J. Chem. Inf. Comp. Sci., 1993, 33, 79-85.

53. W. Fisanick, K.P. Cross, and A. Rusinko III, Similarity Searching on CAS Registry Substances. I. Global Molecular Property and Generic Atom Triangle Geometric Searching, J. Chem. Inf. Comp. Sci., 1992, 32(6), 664-74.

54. R. Carbo and B. Calabuig, Molecular Quantum Similarity Measures and N-Dimensional Representation of Quantum Objects. I. Theoretical Foundations, Int. J. Quant. Chem., 1992, 42, 1681-93.

55. Klebe, G., Mietzner, T., Weber, F., Different Approaches Toward an Automatic Structural Alignment of Drug Molecules: Applications to Sterol Mimics, Thrombin, and Thermolysin Inhibitors., J.Comput.- Aided Mol. Design, 1994, 8, 751-778.

56 Kuntz, I.D., Blaney, J.M., Oatley, S.J., Langridge, R., Ferrin, T.E., A Geometric Approach to Macromolecule-Ligand Interactions, J. Mol. Biol., 1982, 161, 269-288.


[BTN]Return to the paper

 

57. Distributed by DAYLIGHT Chemical Information Systems, Inc., Irvine, CA.

58. Jorgensen, W.L., Tirado-Rives, J., The OPLS Potential Functions for Proteins. Energy Minimizations for Crystals of Cyclic Peptides and Crambin, J. Amer. Chem. Soc., 1988, 110, 1657-1666.

59. Mohamadi, F., Richards, N.G.J., Guida, W.C., Liskamp, R., Lipton, M., Caufield, C., Chang, G., Hendrickson, T., Still, W.C., MacroModel-An Integrated Software System for Modeling Organic and Bioorganic Molecules Using Molecular Mechanics, J. Comp. Chem., 1990, 11, 440-467.

60. Still, W.C., Tempczyk, A., Hawley, R.C., Hendrickson, T., Semianalytical Treatment of Solvation for Molecular Mechanics and Dynamics, J. Amer.Chem. Soc., 1990, 112, 6127-6129.

61. Shoichet, B.K., Kuntz, I.D., Matching Chemistry and Shape in Molecular Docking, Prot. Eng., 1993, 6, 723-732

62. personal communication, Elaine Meng.

63. Lovejoy, B., et al., Structure of the Catalytic Domain of Fibroblast Collagenase Complexed with an Inhibitor, Science, 1994, 263, 375- 377

64. a. Rusinko, A. III, Skell, J.M., Balducci, R., McGarity, C.M., Pearlman, R.S., Concord, A Program for the Rapid Generation of High Quality Approximate 3- Dimensional Molecular Structures; The University of Texas at Austin and Tripos Associates: St. Louis, MO. 1988. b. Pearlman, R.S., Rapid Generation of High Quality Approximate 3D Molecular Structures. Chem. Des. Aut. News., 1987, 2, 1-6.

65. DeBondt, H.L., Rosenblatt, J., Jancarik, J., Jones, H.D., Morgan, D.O., Kim, S.-H., Crystal Structure of Cyclin-Dependent Kinase 2, Nature, 363, 595-602.

66. Graziani, Y., Erikson, E., Erikson, R.L., The Effect of Quercitin on the Phosphorylation Activity of the Rous Sarcoma Virus Transforming Gene Product in vitro and in vivo, Eur. J. Biochem., 1983, 135, 583-589.

67. Hagiwara, M., Inoue, S., Tanaka, T., Nunoki, K., Ito, M., Hidaka, H., Differential Effects of Flavonoids As Inhibitors of Tyrosine Protein Kinases and Serine/Threonine Protein Kinases, Biochem. Pharmacol., 1988, 37, 2987-2992.

68. personal communication, Brett Lovejoy.

69. a. Martin, Y.C., Bures, M.G., Danaher, E.A., DeLazzer, J., Lico, I., Pavlik, P.A., A Fast New Approach to Pharmacophore Mapping And It's Application to Dopaminergic and Benzodiazepine Agonists., J. Comput.- Aided Mol. Design, 1993, 7, 83-102. b. DISCO, Tripos Associates: St. Louis, MO.

70. a. Golender, V.E., Vorpagel, E.R., Computer Assisted Pharmacophore Identification, In 3D QSAR in Drug Design: Theory, Methods, and Applications; Kubinyi, H., Ed.; ESCOM Science Publishers: Leiden, 1993; pp. 137-149. b. APEX-3S, Biosym Technologies, San Diego.

71. Catalyst, Molecular Simulations Inc., 545 Oakmead Parkway, Sunnyvale, CA 94086.

72. Martin, Y. C. and van Drie, J. H., Identifying Unique Core Molecules From the Output of a 3-D Database Search, Chemical Structures 2, 1993, 315-26, Wendy Warr (Ed.), Springer-Verlag, Berlin.


[BTN]Return to the paper



NetSci, ISSN 1092-7360, is published by Network Science Corporation. Except where expressly stated, content at this site is copyright (© 1995 - 2010) by Network Science Corporation and is for your personal use only. No redistribution is allowed without written permission from Network Science Corporation. This web site is managed by:

Network Science Corporation
4411 Connecticut Avenue NW, STE 514
Washington, DC 20008
Tel: (828) 817-9811
E-mail: TheEditors@netsci.org
Website Hosted by Total Choice