QSAR and Library Design in Discovery Studio

Advanced tools forcreating optimalchemical librariesDiscovery StudioQSAR and Library Design in Discovery StudioPareto Optimizationfor simultaneousmultivariableenhancementAn extensive set ofQSAR descriptors andseamless integrationof additional datainto QSAR modelsThe Power of QSAR and Library DesignThe ultimate objective of many computational modeling studies is to identify compoundsthat could potentially become new drugs. With appropriate molecular descriptors, thelarge quantity of relatively easily available data inherent in chemical libraries can bemined, analyzed and used to select compounds that can become drug leads.DS QSAR provides easy access to the hundreds of molecular descriptors, proven in biologicalsystems to correlate with activity. The streamlined Discovery Studio® (DS) interface presentsthese descriptors and advanced modeling and visualization methods in an easy-to-useenvironment. DS Library Design applies these capabilities together with similarity and diversitymethods specifically tailored for chemical library design to guide optimal library design.Accelrys ScienceExtensive Set of Proven Descriptors to EffectivelyCapture Critical Properties in DS QSAR• Describe billions of structural features presentin molecules using Extended ConnectivityFingerprint (ECFP) descriptors.• Access traditional descriptors for basicchemical features, physical properties, ADMEcharacteristics, and experimental data.• Optionally add VAMP and DMol 3 for efficientimplementations of semi-empirical quantummechanical methods to rapidly calculatehighly accurate electronic properties forthousands of candidate compounds.Advanced Modeling Tools for Easy Analysis ofComplex Data in DS QSAR• Easily apply modeling techniques such asBayesian models, multiple linear regression,Partial Least Squares (PLS), Genetic FunctionalAnalysis and more.• Extend the basic functionality of the packageby adding an advanced neural networkcomponent and quantum mechanical baseddescriptors.Powerful, Customizable, and Easily AccessibleSAR Tools in DS QSAR• Integrated Discovery Studio environmentprovides easy access to QSAR tools along sidelibrary design and other tools.• Using the Pipeline Pilot Platform, QSARmodels can easily be deployed to and sharedamong large groups of chemists.• Advanced 3D graphing and molecular dataviews guide users in drawing conclusionsfrom complex data.Easily Design Targeted Chemical Libraries withDS Library Design• Maximize multiple properties simultaneouslyusing Pareto Optimization methods.• Prioritize chemical libraries using readilyaccessible metrics for diversity, similarity, andhundreds of physical and chemical properties.• Interactively customize chemical librariesto optimize the value of each member of aselected set of compounds.

Discovery StudioFigure 1 An interactive MultipleLinear Regression QSAR modelshowing correlation betweenactual and predicted activities.Selection between the plot andtable are synchronized.The Gold Standard in TechnologyComprehensive – The QSAR and library design toolsin Discovery Studio include hundreds of usefuldescriptors, multiple well validated model buildingtechniques and tools specially tailored for customlibrary selection and design. These packages caneasily be augmented with powerful add-ons forrapidly generating quantum mechanical baseddescriptors and advanced modeling techniques.Proven history – The core technology has undergoneover a dozen years of continuous innovationand customer driven improvement, and hasdemonstrated dependable performance in thepharmaceutical industry with dozens of publications.Cutting edge – Accelrys is incorporating newscientific tools to meet current pharmaceuticalneeds and we are continuously working with ourcustomers to plan for future innovation.Easy to use interface – DS 2.0 provides apowerful and intuitive user interface. DS 2.0can be deployed either in a complete standalonesolution for individual modelers or as part of anenterprise-level client server installation for easierprotocol sharing and administration in largermodeling groups.Integrated solution – The DS 2.0 environment, basedon the Pipeline Pilot open operating platform,integrates protein modeling, pharmacophoreanalysis, and virtual screening as well as thirdpartyapplications for an infinitely extensiblevirtual discovery platform. Well-tested applicationsincluding CHARMm, MODELER, Catalyst, and othersare accessible in the graphical DS environment, thePipeline Pilot scripting and protocol developmentenvironment and from command-line prompts.Parallel computing – The DS 2.0 platform isoptimized to take advantage of grid and clustercomputing as well as multi-core processors torapidly process large tasks.Accelrys is Your Partner in ResearchUser community – Accelrys scientific forumsand presentations at scientific meetings worldwideprovide opportunities for Accelrys users toexchange ideas and share new research.Scientific consulting – Accelrys has dozensof experienced Ph.D.s with expertise inimplementing scientific solutions for drugdesign that are available for short or long-termengagements to create tailored solutions orperform modeling experiments.

Discovery StudioFigure 2 An interactive 3D plot ofthe first three principal componentsfrom a Principal ComponentAnalysis using several moleculardescriptors as input. The plot canoptionally be colored by a specifiedproperty.Customer support – Accelrys customers report a98% satisfaction rate with our support team.Committed to innovation – With over 100 Ph.D.s inthe field working daily with researchers in industryand academia, Accelrys is committed to deliveringcutting-edge technology to our customersWorld leading scientific advisors – Through ourin-licensing agreements, partnerships, and scientificadvisors many of the world’s foremost experts incomputational drug design are involved in settingour direction.Biological Validation andComparison2007 – Genetic Functional Analysis module used tostudy fluoroquinolone antibacterials. The resultssuggest specific chemical modifications andphysical properties for future drug design efforts. 12006 – Effect of 2-aminothiazole derivatives onNeuro-cell apoptosis studied using a QSAR modelthat has a 97% correlation to EC 5022003 – A descriptor-based QSAR model for hOCT1predicts IC 50values with 95% correlation toobserved data. 3References:1.2.3.Cheg, D. et al., “Relationship of quantitative structure and pharmacokinetics in fluoroquinolone antibacterials”,World J Gastroenterol, 2007, 13(17), 2496-503Jiang FC., et al., “The design and synthesis of 2-aminothiazole derivatives and their inhibitory activity on apoptosis”,Yao Xue Xue Bao, 2006, 41(8), 727-34Bednarczyk D., et al., “Influence of molecular structure on substrate binding to the human organic cation transporter, hOCT”,Mol. Pharmacol, 2003, 63(3), 489-980907