Current Topics in Medicinal Chemistry, Volume 5, No. 8, 2005
Contents
Theoretical
Approaches in Medicinal Chemistry and Drug Discovery
Guest
Editor: Jurgen Bajorath
Editorial Pp. 737-738
Jurgen Bajorath
Hierarchical
Docking of Databases of Multiple Ligand Conformations Pp. 739-749
David M. Lorber and Brian K. Shoichet
Library Design
for Fragment Based Screening
Pp. 751-762
Ansgar Schuffenhauer, Simon Ruedisser, Andreas L.
Marzinzik, Wolfgang Jahnke, Marcel Blommers, Paul Selzer and Edgar Jacoby
Chemical and
Biological Profiling of an Annotated Compound Library Directed to the Nuclear Receptor
Family Pp. 763-772
Montserrat Cases, Ricard Garcia-Serna, Kristina Hettne,
Marc Weeber, Johan van der Lei, Scott Boyer and Jordi Mestres
Web Enabling
Technology for the Design, Enumeration, Optimization and Tracking of Compound
Libraries Pp. 773-783
Bradley P. Feuston, Subhas J. Chakravorty, John F.
Conway, J. Christopher Culberson, Joseph Forbes, Bryan Kraker, Patricia A.
Lennon, Craig Lindsley, Georgia B. McGaughey, Ralph Mosley, Robert P. Sheridan,
Mario Valenciano and Simon K. Kearsley
Information
Extraction in the Life Sciences: Perspectives for Medicinal Chemistry,
Pharmacology and Toxicology
Pp. 785-796
Marc Zimmermann, Juliane Fluck, Le Thuy Bui Thi,
Corinna Kolarik, Kai Kumpf, and Martin Hofmann
Potency-Scaled
Partitioning in Descriptor Spaces with Increasing Dimensionality Pp. 797-803
Jurgen Bajorath
Progress in the
Development of Selective Inhibitors of Aurora Kinases Pp. 807-821
Andrew A. Mortlock, Nicholas J. Keen, Frederic H. Jung,
Nicola M. Heron, Kevin M. Foote, Robert W. Wilkinson and Stephen Green
Abstracts
[Back to top] Editorial
Jurgen Bajorath
Medicinal chemistry programs have long benefited from computational contributions, in particular, applications of molecular modeling techniques, QSAR methods, or docking simulations. Ultimately, candidate molecules need to be synthesized and tested, but hit identification and -perhaps even more so- lead optimization efforts tend to be heavily supported by computational means. Moreover, in current drug discovery research, novel computational concepts continue to be introduced, ranging from those that play a role during the early stages of the discovery process, including data analysis and mining or compound library design, to others that attempt to predict downstream or in vivo effects of compounds and thereby help to reduce the high fall-out rates of clinical candidates. Such developments significantly extend the spectrum of more classical computational chemistry techniques in drug discovery.
This special issue of Current Topics in Medicinal Chemistry brings together contributions from research groups in major pharmaceutical companies and academic institutions. These contributions were selected in order to highlight different types of computational approaches, although most of them ultimately target various aspects of advanced chemical library design. What they all have in common, however, is a strong focus on novel technologies with high relevance for pharmaceutical research. As such, they are thought to provide interesting examples of ongoing research and development activities in both pharmaceutical and academic settings.
On a first glance, the paper by Lorber and Shoichet, with which this issue begins, provides an extensive comparative analysis of flexible and rigid ligand docking calculations on multiple target proteins. However, its major novel aspect lies in a hierarchical design and organization scheme for multiple ligand conformations to facilitate pseudo-flexible docking. This hierarchical design and docking strategy permits efficient pruning of arrays of pre-computed ligand conformations. The degree of flexibility achieved here is shown to significantly improve ligand rankings for five of seven enzyme targets when compared to rigid docking, without dramatically increasing calculation time.
The design of specialized compound libraries, albeit very different ones, is also the topic of the next two contributions. Schuffenhauer and colleagues focus on computational design of molecular fragment libraries that are highly relevant for fragment screening and subsequent structure-based design of potent leads from multiple weakly bound fragments. In this case, the library design rationale and framework are provided by a previously reported model estimating the role of small molecule complexity in protein-ligand binding, and the design approach pursued here is also hierarchical in nature. Synthetic feasibility criteria for fragment modification are rigorously taken into account (and so is target information). The authors elegantly derive and organize pairs of corresponding screening and synthesis fragments.
Cases et al. introduce yet another hierarchical library design approach; here, however, the focus completely changes, as these researchers report on the generation of an annotated compound library targeting nuclear receptors. The library contains compounds that have been tested against different groups of nuclear receptors, and its key feature is the annotation of these molecules with hierarchically organized biological and chemical information. Systematic annotation permits profiling of either receptors or chemical scaffolds and has led to the identification of chemical promiscuity patterns of nuclear receptors. This type of library integrates information about ligand and protein families and their interactions and thus represents a prototype of a molecular system designed for chemical genetics or genomics applications.
These three examples, describing hierarchical multi-conformer, fragment screening, and annotated target class libraries, nicely illustrate how multi-facetted and specialized the library design field has become since the early days of diversity design about 10 years ago. Going beyond specialized libraries and moving on to computational infrastructures, the fourth paper by Feuston and colleagues describes the organization and implementation of intranet tools to support the planning, synthesis, analysis, and optimization of chemical libraries. This contribution nicely shows how such efforts are carried out, coordinated, and documented inside a pharmaceutical giant using web technologies. A library targeting G protein-coupled receptors serves as an example to illustrate the application web-based tools, for example, for synthon analysis and reagent selection.
Departing from the library design theme, the following paper, a contribution from our group, introduces partitioning algorithms that utilize binary descriptor transformation and operate in chemical spaces with variable dimensionality. One of these algorithms (DMC) has been extended to include potency scaling (POT-DMC) in order to tune virtual screening calculations toward the recognition of increasingly potent hits. In a comparison of DMC and POT-DMC calculations on different activity classes, a systematic and significant enrichment of potent compounds among correctly identified hits is achieved under scaling conditions.
The final paper in this issue by Zimmermann et al. provides insights into emerging technologies for extraction and retrieval of biological and chemical information, not only from documents but also from images. No doubt, methods for extracting structural and functional information concerning synthetic molecules from abundant, complex, and diverse chemical information sources will become increasingly important in the future; even if one would only think about issues such as alleviating the need to re-draw the steadily growing number of structures collected from publications or presentations. Among other things, the authors describe a prototypic tool kit for structure recognition and reconstruction.
In summary, it is of course not possible to cover the selected research topics in full depth in a single issue of the Journal. However, it is hoped that the exemplary contributions put together here provide instructive examples of ongoing research in computational medicinal chemistry and drug discovery that might be of interest to many readers.
[Back to top] Hierarchical
Docking of Databases of Multiple Ligand Conformations
David M. Lorber and Brian K. Shoichet
Ligand flexibility is an important problem in molecular docking and virtual screening. To address this challenge, we investigate a hierarchical pre-organization of multiple conformations of small molecules. Such organization of pre-calculated conformations removes the exploration of ligand conformational space from the docking calculation and allows for concise representation of what can be thousands of conformations. The hierarchy also recognizes and prunes incompatible conformations early in the calculation, eliminating redundant calculations of fit. We investigate the method by docking the MDL Drug Data Report (MDDR), an annotated database of 100,000 molecules, into apo and holo forms of seven unrelated targets. This annotated database allows us to track the ranking of tens to hundreds of annotated ligands in each of the docking systems. The binding sites and database are prepared in an automated fashion in an attempt to remove some human bias from the calculations. Many thousands of explicit and implicit ligand conformations may be docked in calculations not much longer than required for single conformer docking. As long as internal energies are not considered, recombination with the hierarchy is additive as the number of degrees of freedom is increased. Molecules with even millions of conformations can be docked in a few minutes on a single desktop computer.
[Back to top] Library
Design for Fragment Based Screening
Ansgar Schuffenhauer, Simon Ruedisser, Andreas L.
Marzinzik, Wolfgang Jahnke, Marcel Blommers, Paul Selzer and Edgar Jacoby
According to Hann’s model of molecular complexity an increased probability of detection binding to a target protein can be expected when small, low complex molecular fragments are screened with high sensitivity instead of full-sized ligands with lower sensitivity. Analysis of the HTS summary data of Novartis and comparison with NMR screening results obtained on generic fragment libraries indicate this expectation to be true with hitrates of 0.001% - 0.151% observed in the identification of ligands with an IC50 threshold in the micromolar range in an HTS setup and hitrates above or equal to 3% observed in NMR screening of fragments with an affinity threshold in the millimolar range. It is however necessary to keep in mind that the sets of target studied were not identical for both method and the experience in NMR screening is too limited for a final conclusion. The term hitrate as used here reflects only the success rate in the observation of ligand binding event. It must not be confused with the overall success rate of fragment and high throughput screening in the lead finding process, which can be entirely different, since the steps required to follow-up a ligand binding event to a lead are different for both methods.
A survey of fragment-based lead discovery case studies given in the literature shows that in approximately half of the cases the initial hit fragment was discovered by screening a generic library, whereas in the other cases some knowledge about an initial ligands or the protein binding site has been used, whereas systematic virtual screening of fragment databases has been only rarely reported.
As comparatively high hitrates were obtained, further consideration to optimize the generic fragment screening library were directed to the chemical tractability of the fragment. As several functional groups preferred by chemists for modification and linking of the fragments are also preferentially involved in interactions between the fragments and the target protein, a set of screening fragments was derived from chemical building blocks by masking its linker group by a chemical transformation which can be later on used in the chemical follow-up of the fragment hit. For example primary amines can be masked as acetamides. If the screening fragment is active the related building block can then be used for synthesis of a follow-up library.
[Back to top] Chemical and
Biological Profiling of an Annotated Compound Library Directed to the Nuclear
Receptor Family
Montserrat Cases, Ricard Garcia-Serna, Kristina Hettne,
Marc Weeber, Johan van der Lei, Scott Boyer and Jordi Mestres
Nuclear receptors form a family of ligand-activated transcription factors that regulate a wide variety of biological processes and are thus generally considered relevant targets in drug discovery. We have constructed an annotated compound library directed to nuclear receptors (NRacl) as a means for integrating the chemical and biological data being generated within this family. Special care has been put in the appropriate storage of annotations by using hierarchical classification schemes for both molecules and nuclear receptors, which takes the ability to extract knowledge from annotated compound libraries to another level. Analysis of NRacl has ultimately led to the identification of scaffolds with highly promiscuous nuclear receptor profiles and to the classification of nuclear receptor groups with similar scaffold promiscuity patterns. This information can be exploited in the design of probing libraries for deorphanization activities as well as for devising screening batteries to address selectivity issues.
[Back to top] Web Enabling
Technology for the Design, Enumeration, Optimization and Tracking of Compound
Libraries
Bradley P. Feuston, Subhas J. Chakravorty, John F.
Conway, J. Christopher Culberson, Joseph Forbes, Bryan Kraker, Patricia A.
Lennon, Craig Lindsley, Georgia B. McGaughey, Ralph Mosley, Robert P. Sheridan,
Mario Valenciano and Simon K. Kearsley
Motivated by the need to augment Merck’s in-house small molecule collection, web-based tools for designing, enumerating, optimizing and tracking compound libraries have been developed. The path leading to the current version of this Virtual Library Tool Kit (VLTK) is discussed in context of the (then) available commercial offerings and the constraints and requirements imposed by the end users. Though the effort was initiated to simplify the tasks of designing novel, drug-like and diverse compound libraries containing between 2K-10K unique entities, it has also evolved into a powerful tool for outsourcing syntheses as well as lead identification and optimization. The web tool includes components that select reagents, analyze synthons, identify backup reagents, enumerate libraries, calculate properties, optimize libraries and finally track the synthesized compounds through biological assays. In addition to accommodating project specific designs and virtual 3D library scanning, the application includes tools for parallel synthesis, laboratory automation and compound registration.
[Back to top] Information
Extraction in the Life Sciences: Perspectives for Medicinal Chemistry,
Pharmacology and Toxicology
Marc Zimmermann, Juliane Fluck, Le Thuy Bui Thi,
Corinna Kolarik, Kai Kumpf, and Martin Hofmann
Information extraction approaches have been successfully applied to mine the scientific literature in biology and medicine. So far, the main focus of research and development in this domain was on the recognition and extraction of gene and protein names in the context of molecular biology and genome research and on disease names and other medical terms in the context of clinical research. Similar to biology and medical sciences, medicinal chemistry, pharmacology and toxicology are descriptive sciences. However, information extraction approaches in these disciplines encounter a number of problems that are specific to the fact that these scientific areas are essentially centred at chemical compounds and their structures. In this review, we will give a short overview on general information extraction strategies in the life sciences and we will introduce new approaches to apply information extraction to the domain of pharmacology, medicinal chemistry and toxicology. Finally, we will emphasize on how information extraction approaches will support public and commercial research in medicinal chemistry, pharmacology and toxicology by linking information on chemical structures to biological information.
[Back to top] Potency-Scaled
Partitioning in Descriptor Spaces with Increasing Dimensionality
Jurgen Bajorath
Partitioning algorithms are described that operate in chemical reference spaces formed by combinations of binary-transformed molecular descriptors and aim at the identification of potent hits in ligand-based virtual screening. One of these approaches depends on mapping of consensus positions of compound activity sets in descriptor spaces followed by step-wise extension of the dimensionality of these spaces and re-mapping of activity-dependent consensus positions. Dimension extension is carried out to increase the discriminatory power of descriptor combinations and distinguish database compounds from potential hits. This method was originally named Dynamic Mapping of Consensus positions (DMC) and subsequently extended in order to take different potency levels of known active molecules into account and increase the probability of recognizing potent database hits. The extension was accomplished by adding potency scaling to DMC calculations, and the resulting approach was termed POT-DMC. Results of comparisons of DMC and POT-DMC calculations on different classes of active compounds with substantially varying potency levels support the validity of the POT-DMC approach.
[Back to top] Progress in the
Development of Selective Inhibitors of Aurora Kinases
Andrew A. Mortlock, Nicholas J. Keen, Frederic H. Jung, Nicola M. Heron, Kevin M. Foote, Robert W. Wilkinson and Stephen Green
Errors in the mitotic process are thought to be one of the principal sources of the genetic instability that hallmarks cancer. Unsurprisingly, many of the proteins that regulate mitosis are aberrantly expressed in tumour cells when compared to their normal counterparts. These may represent a good source of targets for the development of novel anti-cancer agents. The Aurora kinases represent one such family of mitotic regulators.
In recent years there has been intense interest in both understanding the role of the Aurora kinases in cell cycle regulation and also in developing small molecule inhibitors as potential novel anti-cancer drugs. With several companies now starting to take Aurora kinase inhibitors into clinical development, the time is right to review the medicinal chemistry contribution to developing the field, in particular to review the increasingly broad range of small molecule inhibitors with activity against this kinase family.