 |
|
 |
|
|||||||||||||||
|
|
|
|
 |
|
 |
|
|
|
|
|
|
|
|
||
|
||||
|
|
||||||||||
|
PR-1 : Pharmacophore Elucidation in MOE (Molecular Operating Environment) Steve Maginn; Chemical Computing Group The latest, 2005 version of the Molecular Operating Environment (MOE) system contains a new application for the derivation of pharmacophores from a database of known active compounds. This new capability builds on the Flexible Alignment, Conformation Import and Consensus Pharmacophore capabilities already available in MOE, and is receiving one of its first public showings at Noordwijkerhout.
PR-2 : Pipeline Pilot Product Update Robert Brown; SciTegic Pipeline Pilot is SciTegic’s data pipelining product that allows the construction and rapid execution of workflows known as protocols. Individual operations within a protocol are provided by components, each of which performs a single task on the records flowing through the pipeline. The capabilities of Pipeline Pilot have been greatly expanded in recent releases through the addition of new component collections and through the availability of new components integrating 3rd party software. This product review will introduce new component collections from SciTegic for
In addition, both SciTegic and our ISV (Independent Software Vendor) partners are developing component collections that integrate 3rd party applications. We will highlight the latest developments and discuss the availability of these components.
PR-3 : Infotherm: Thermophysical Data of Mixtures and Pure Compounds Joerg Homann; FIZ CHEMIE Berlin The database Infotherm, currently available via Internet, comprises 167,000 tables of PVT-properties, phase equilibria, transport and surface properties, caloric properties, acoustic and optical properties of 25,000 mixtures and 6,700 pure compounds taken from journals, data collections, manuals and measurement reports some of which exclusive to Infotherm. Infotherm was relaunched in November 2004 with improved search functions in order to combine 98 properties, 25 conditions, 22 types of equilibria, 10 chemical systems, definable value ranges, substance names, formulas and CAS registry numbers by boolean operators. Besides the improved search procedure the data sources and quality assurance will be presented.
PR-4 : COSMO-RS Based Methods in Drug Design Karin Wichmann, COSMOlogic GmbH & Co. KG COSMOtherm is our approved program of the Conductor-like Screening Model for Realistic Solvation method (COSMO-RS), i.e. for the quantitative calculation of fluid phase thermodynamics based on quantum chemical COSMO calculations. Besides many features primarily developed for the very demanding chemical engineering mixture thermodynamics, COSMOtherm allows for predictive calculations of many properties relevant for life-science research, as water solubility and solubility in other solvents, partition coefficients between any solvents, and pKA. It has predefined models for logPOW, blood-brain partitioning, intestinal absorption, albumin binding, and allows for the definition of other models using the s-moments approach. New features of the current distribution include a graphical user interface (GUI), and improved COSMOtherm functionality and applicability, e.g. new pKA computation for bases. COSMOfrag is a tool which performs the rapid automated fragment-wise construction of approximate s-profiles and the calculation of the resulting physicochemical properties of larger molecules, bringing the COSMO-RS technology to high-throughput performance. It is based on a database of presently approximately 40,000 COSMO files of highly diverse, small basic and larger drug-like compounds. After a highly efficient database search for the most appropriate locally similar fragments of a new compound, COSMOfrag generates COSMO meta files comprising the fragmentation description of the new compound. These meta files and the resulting fragment-based s-profiles can then be used as starting point for many COSMO-RS calculations, i.e. physicochemical, physiological or environmental property calculations, drug similarity searching or even receptor binding approaches. COSMOsim provides a novel approach for the quantification of drug similarity, which makes use of the surface polarities s as defined in COSMO-RS. The histogram of these surface polarities, the s-profiles, have been proven to be the key for the calculation of all kinds of partition and adsorption coefficients, and thus of relevant ADME parameters like solubility, logBB and many others. It also carries a large part of the information required for the estimation of desolvation and binding processes responsible for the inhibition of enzyme receptors by drug molecules. Consequently, a large degree of similarity with respect to the s-profiles appears to be a necessary condition for drugs of similar physiological action. Driven by this insight, we have developed a s-profile based drug similarity measure SMS for the detection of new bioisosteric drug candidates. In several examples and in a number of real drug design projects COSMOsim already has demonstrated its statistical and pharmaceutical plausibility, its practicability for real drug research projects, and its unique independence from the chemical structure which enables scaffold hopping in a natural way. COSMObase is a database of high quality DFT/COSMO files for about 2800 common compounds with a focus on common solvents. COSMObase can be very useful in the context of solvent and co-solvent selection and other screening applications.
PR-5 : MDL Patent Chemistry Database & DiscoveryGate Eva Seip, Elsevier MDL Patents are an important and under-used source of information in chemistry and life sciences research. While many text-based systems already exist for accessing patent information, structure-based searching offers a more powerful and flexible way for scientists to mine this vast pool of important information. To provide this capability, Elsevier MDL offers the new structure-searchable MDL® Patent Chemistry Database, a factual database indexing reactions, substances and their properties from organic chemistry and life science patent publications (World and European since 1978, U.S. since 1976). The MDL Patent Chemistry Database has been specifically designed for R&D scientists—chemists and bio-/medicinal scientists—to provide structure-based access to key patent chemistry information. The database can help researchers more effectively plan syntheses, better profile bioactivity and quickly understand the scope and relevance of patents. The database can assist researchers in designing new synthetic methods, developing drug profiles, selecting and optimizing leads and monitoring competitor activities and industry trends. Updated every two weeks, the Patent Chemistry Database currently contains about 1.8 million reactions, about 2 million organic, inorganic, organometallic (and polymeric*) compounds and related information from approximately 360,000 patents. For more information, visit www.mdl.com. With the Patent Chemistry Database available on the DiscoveryGate® content platform, scientists can explore patent information over a variety of integrated and complementary information sources such as Derwent Chemistry Resource or the MDL® Drug Data Report database. The latest version of DiscoveryGate is faster, simpler and easier to use, displaying hits as soon as they are retrieved and requiring fewer clicks and windows to achieve results. See www.discoverygate.com. * For patent applications published from December 2003 onwards
PR-6 : ChemAxon: Platform for Cheminformatics Miklos Vargyas, ChemAxon Ltd. ChemAxon has been well-known in the pharmaceutical industry for its Java based web-enabled and platform independent cheminformatics components and tools, primarily for MarvinView which is a 2D structure visualizer and for MarvinSketch, a chemical drawing tool. Recently, however, ChemAxon has also been acknowledged as a cheminformatics platform provider. The talk will move through the core technology of the JChem cheminformatics and discovery platform highlighting recent advances - in the form of connecting with posters being presented at the meeting. Mention will be made of the results of development of JChem Base chemical database server and of our JChem Cartridge for Oracle and the integration of ChemAxon's Chemical Terms language for advanced searching. A new product line developed in the past two years targets rational drug design. Applications include the Synthesizer and Reactor software that mimics chemical reactions in a knowledge based fashion, the Screen package coupled with JKlustor for high throughput virtual screening and library analysis. Most recently ChemAxon made a move into the 3D world of cheminformatics. A 3D conformation generator was marketed first. The latest product to be released is MarvinSpace, an OpenGL based 3D structure visualizer component. In the future MarvinSpace will develop to a drug discovery platform. The presentation will highlight some interesting features in a live demonstration.
PR-7 : LigandScout Review: An Application to Human Factor Xa Inhibitors Monika Rella; Inte:Ligand GmbH Factor Xa inhibitors are innovative anticoagulant agents that provide a better safety and efficacy profile compared to other anticoagulative drugs. We used LigandScout, a novel software package that fully automatically derives pharmacophores from protein-ligand-complexes, to identify interaction patterns of inhibitors bound to Human Factor Xa (PDB entries 1fjs, 1kns and 1eqz). The complex structures were selected regarding the criteria of high inhibitory potency (i.e. all ligands show Ki values against factor Xa in the sub-nanomolar range) and good resolution (i.e. at least 2.2A) in order to generate selective and high quality pharmacophore models. Common feature pharmacophores were generated using automated overlaying of several models obtained from the complex structures. The resulting chemical-feature based pharmacophore models were used for virtual screening of commercial molecular databases like the WDI 2003 database using the screening platform Catalyst [2]. The results obtained indicate that the models created using LigandScout are highly selective filters suitable for successful virtual screening experiments. References
PR-8 : The Proper Handling of Shape and Electrostatics in Ligand-based Design George Vacek, OpenEye Scientific Software ROCS and EON screen databases of molecules for shape and electrostatic similarity to a lead compound and thereby facilitate lead generation and library design for drug discovery. ROCS and EON report rigorous shape and electrostatic Tanimoto and Tversky measures between molecules, so that a molecular database can be quickly sorted according to similarity. More importantly, they provide an intuitive measure of similarity; that is, when two molecules with a high shape and electrostatic similarity are viewed, they clearly look alike and one can see that they should interact in a similar fashion. Shape and electrostatics are not only whole molecule properties but also fundamental metrics, giving them advantages over 2D descriptors and simple pharmacophore models. Several recent publications validate the use of ROCS and EON in drug discovery. The first publication used ROCS to ‘scaffold hop’ to several active compounds without the same toxicity and potential intellectual property issues as their original lead compound. The new actives had been missed by the original HTS campaign. The second study, using both ROCS and EON to screen a virtual combinatorial library, showed the use of electrostatic similarity to be critical. One of the resulting leads was three-fold more active than the original query. In the third paper, shape and electrostatics were shown to be important variables in discriminating classes of active and inactive compounds for the diverse cases of COX2, progesterone, dopamine and the calcium ion channel. Key new features in ROCS 2.1 and EON 1.1 include:
Referenced Publications
PR-9 : Mogul: Rapid Retrieval of Molecular Geometry Information from a Crystallographic Database Susan Robertson, Cambridge Crystallographic Data Centre (CCDC) The Cambridge Structural Database (CSD) currently contains over 345,000 structure determinations for small molecule organic and organometallic compounds, and thus is an excellent source of information on molecular geometries and conformational preferences. The program Mogul [1] (Molecular Geometry Library) has been developed by the CCDC to provide rapid access to this information with full statistics. To retrieve information on a bond length, valence angle or acyclic torsion, the user inputs a molecule of interest into Mogul and selects the geometric feature to be investigated. The molecular environment of the selected fragment is used to generate atom- and bond-based "keys", which are then used to retrieve all relevant CSD hits. The use of a search tree optimises search speeds without the need for graph-based atom-by-atom matching typically used by other retrieval programs such as ConQuest [2]. Key statistics and histograms are derived and displayed along with an indication of how well the geometry of the feature in the input molecule matches observed geometries in the CSD. Mogul searches can be run automatically via an instructions file with results written to an output file, allowing easy integration with other programs. Mogul has already been integrated with CRYSTALS [3] in this way to allow validation of molecular geometries of newly refined crystal structures. Mogul can also be used with DASH [4], software for structure solution from powder diffraction data.
PR-10 : KnowItAll®: An Integrated Solution for ADME/Tox, Spectroscopy, Cheminformatics, and Custom Software Development Holger Ruchatz, Bio-Rad Laboratories, Inc - Informatics Division Bio-Rad Laboratories, Inc. offers both desktop and enterprise-wide solutions to facilitate multiple aspects of pharmaceutical drug design and screening, industrial, and academic research including:
Bio-Rad’s KnowItAll offers a complete suite of award-winning tools for the in silico prediction of a potential drug’s ADME/Tox profile, including over 30 predictive models, applications to validate and build predictive models (using SVM technology), and experimental ADME/Tox data. Bio-Rad’s unique environment for ADME/Tox features consensus modeling, which involves the combination of multiple, complementary models to improve the accuracy of prediction over single models. This presentation will demonstrate how drug discovery professionals can improve the accuracy of their in silico investigations using this technology and will review tools for global models and local models. Bio-Rad offers multiple solutions for multiple spectral techniques including spectral data management for proprietary data, processing, analysis, and access over 885,000 high-quality MS, NMR, IR, and Raman spectra. At ICCS, Bio-Rad will feature its new multi-technique spectral searching. Unlike other search software, KnowItAll searches data from ALL techniques at the same time. The system seamlessly consolidates all spectral data available to yield a single, intuitive result. By visualizing all spectral information simultaneously, KnowItAll is able to offer accuracy and efficiency beyond systems that handle only one spectral technique at a time. For more information, please visit www.knowitall.com.
PR-11 : Automated Property-based Structure Design Gavin Shear, Advanced Chemistry Development, Inc. For the last 10 years, ACD/Labs has been dedicated to building integrated solutions that enable data transfer and connection within chemical organizations. Modern laboratories have to effectively manage multitudes of chemical structures and associated data. ACD/Labs offers chemical and analytical databasing solutions that enable chemists to manage their work effectively, and make stored chemical data available to other members worldwide. Our other key products include renowned predictors of physicochemical properties (pKa, logD, logP, solubility, PSA, boiling point, etc.), nomenclature software, and spectral predictors, processors, and databasing tools. Each year, we release enhanced versions of our software to provide more capabilities and superior integration between existing and new technologies. This year, we would like to present the following products: ACD/Name: Generate Name from Structure or Structure from Name
ACD/ChemFolder: Database Chemical Reactions, Structures and Related Information ACD/Structure Design Suite: Optimize Structures for Improved Properties
PR-12 : Applications for the Modeling of Chemistry and Chemical Properties Christof H. Schwab, Molecular Networks GmbH Molecular Networks' core expertise is in the development of computationally fast, empirical descriptors for the modelling of electronic properties of atoms and bonds that would usually require quantum mechanical calculations. These descriptors are the basis of software tools designed to support researchers in the pharmaceutical, biotechnology and chemical industry and have a wide range of applications: from the prediction of molecular properties used by computational chemists for discovering and optimizing lead compounds to the decision support applications that help chemists to plan a synthesis of chemical compounds. Molecular Networks product portfolio was recently extended with the applications
and will include an updated version of the current synthesis planning system WODCA in the very near future. Molecular Networks product portfolio also includes applications for processing, manipulating, visualizing and warehousing chemical structures and reactions. Due to Molecular Networks and SciTegic's collaboration, several applications (such as 2DCOOR, ADRIANA.Code, CONVERT and CORINA) are available as components for Pipeline Pilot. Molecular Networks will be demonstrating its suite of software applications and decision support tools at the exhibition at Booth #10. The booth personnel will be available to answer any questions regarding Molecular Networks' technology and will be glad to help you to find the right solution for your needs.
PR-13 : ClassPharmer™ Suite Automates Extraction of SAR to Maximize Antiviral Activity and Minimize Cytotoxicity Vincent Vivien; Bioreason Chemists involved in lead optimization use information mined from R-tables to aid in the design of new compounds. This is a labor-intensive and time-consuming process when identifying features and positions of groups that affect more than one property. Using accurate chemical classes as the starting point for extraction of statistically based SAR hypotheses can provide a computational solution for a difficult problem. ClassPharmer Suite meets the everyday development needs of the chemist by creating an easily accessible environment that is chemistry smart and not a black box. In this presentation, we show how an individual would use the information displayed in the R-tables and provided in the hypotheses to identify properties of R-groups that can be used to optimize compounds for two competing properties: high antiviral activity and low cytotoxicity. We used the public DTP-AIDS Antiviral Screening data with values for in vitro protection of HIV infected cells (EC50) and in vitro cytotoxicity of uninfected cells (IC50). Compound structures for molecules with a pyrimidine variation were selected and clustered using ClassPharmer™ Suite. The classification was then analyzed with two ClassPharmer™ Suite analysis modules: generation of R-tables and extraction of structure property relationships for both antiviral activity and cytotoxicity from R-group properties. Analyses were focused on two classes: one without a ring fused to the furan (with representative structures shown below) and one with a fused ring (not shown). All the compounds in the latter were “inactive.” Several of the SAR hypotheses for the former describing the “best” groups at R6 are summarized below. In addition, the activity distribution for one of these hypotheses is shown along with example supporting compounds.
Alexander Kos, AKos GmbH PASS: Prediction of Activity Spectra for Substances PASS predicts the biological activity spectra on the basis of the 2D structural formula. This provides the opportunity to select compounds with desirable effects, and without unwanted side effects in the early stage of drug discovery. PASS version 1.932.1 (January 2005) predicts 1000 kinds of biological activity with an average accuracy of 85% (leave one out cross-validation). Based on the calculated values of probability to be active and inactive (Pa and Pi respectively), one may define a flexible criteria for selecting the most promising leads with desirable level of novelty. Calculation of biological activity spectra for 100,000 compounds on a PC takes about 20 min. PASS can be effectively used to analyze large databases. PharmaExpert helps to analyze the prediction taking into account a huge knowledgebase of activity-activity relationships. It provides the means for interactive selecting the most advantageous compounds. Structures are visualized using the Chime plug-in. Compounds can be profiled based on user-defined criteria. Selected compounds can be exported in SD-file format. PASS CL is the command line version for inclusion in other in-house applications, or into programs like SciTegic’s Pipeline Pilot. The programs are working on PCs under Windows using MOL and SDF input formats; TXT, SDF and CSV output formats. Contact: Dr. Alexander Kos, AKos Consulting & Solutions GmbH, Postfach 141, CH-4010 Basel, [email protected], www.akosgmbh.com.
PR-15 : Fingerprints, Clustering and High-speed Virtual Library Analysis: BCI's Software Toolkits and Web Services Geoff Downs, Barnard Chemical Information This year, BCI is celebrating its 20th birthday. Over the past 20 years, it has established a unique reputation for the strength of its consultancy and specialist software for chemical structure representation, clustering, diversity analysis and Markush structure handling. During the past 18 months, BCI has increased the utility of its technology by releasing the functionality contained in the original standalone programs as a range of toolkits and SOAP services. The toolkits are available for an extensive range of operating systems and can be supplied with a number of language interfaces including C/C++ and Java. The SOAP services can be called from any client supporting the SOAP protocol, including Scitegic’s Pipeline Pilot. BCI has also increased the functionality of its products, with significant advances having been made. This is particularly evident in the Markush handling of combinatorial libraries, with very fast full, overlap and substructure searching now being provided.
PR-16 : LITIHUM Dock- A Virtual Assay System for Docking Ulrike Uhrig, Tripos, Inc. LITHIUM Dock allows laboratory scientists ready access to the results of a molecular modeler’s docking work, improving visibility and utilization of modeling results. Through centralized capture and communication of modeler produced docking information, LITIHUM Dock improves decision support in the laboratory by allowing chemists to rapidly run candidate molecular structures against stored docking procedures. LITHIUM Dock user benefits:
LITHIUM Dock capabilities:
PR-17 : SciFinder 2006: A Preview of Upcoming New SciFinder Features Paul Peters, CAS SciFinder was launched 10 years ago as a search tool which promised to change the way scientists conduct research. Over this period of time, the SciFinder family of products has been continually enhanced with new functionality and content. SciFinder is now an enterprise-wide research service for discovery scientists at leading research-driven companies, academic institutions and government agencies throughout the world. SciFinder has grown to become an essential part of the research process. This product review will offer a preview of the upcoming release of SciFinder scheduled for later this summer.
PR-18 : Accord Cheminfomatics Suite - Enhancements in v 6.0 Tim Aitken, Accelrys Ltd. The Accord suite of Cheminformatics software ranges from individual function-specific software components, to programming toolkits, desktop applications and enterprise-wide solutions, providing both off-the-shelf applications and tools to allow custom development. 2005 has seen the launch of version 6.0 of the Accord suite, including some of the following enhancements in chemical and biological data management. This review will focus on recent advances in the Accord Chemistry Engine underlying all Accord products. Accord Developer Tools Additionally, enhancements have been made to the Similarity & fingerprinting functionality, allowing fingerzoning & fingerlocations to be defined – this gives users a wide range of similarity search types, including molecule only, reactant vs product, reaction centre only etc. Bit density enhancements mean the fingerprints can be weighted according to the target data set allowing users to find previously unexpected chemically significant results from similarity searches on reaction databases. The Property calculations available within Accord products have been extended, to include Hepatotoxicity & CYP450-2D6 alongside BB permeability, Aqueous Solubility, #H bond donors etc Version 6.0 of the Accord Chemistry Cartridge allows partitioned indexes & support for parallel processors for enhanced index creation & search speeds, along with support for the Cost Based Optimiser to maximise performance of complex queries. Accord Enterprise Informatics
|
|||||||||||
|
|
|
Last updated 27 October, 2005 |