Transcript SS IBC V3
Schedule 8:30-9:30 Introduction 9:40- 10:45 Analysis Methods 10:55-12:00 Design and Analysis 12:00 Lunch 1:00-2:05 2:15-3:15 3:25-4:00 4:00-4:30 Design and Analysis I (Will and Stan) Pooling Pooling Discussion General Discussion 1 HTS Directions NISS Affiliates Workshop 25 October 2002 2 Sequential Screening Paradigm Corporate Collection Stat Methods Chem Descriptors Model Data Sets Commercial Compounds Filters Virtual Compounds Initial Compounds Statistical Analysis Statistical Model Active Compounds Screen Compounds Selected Compounds 3 Data / Descriptors / Stat Method Data : Protein binding Cell-based Whole Animal, e.g. Toxicology Descriptors : BCUT, Atom Type Counts, Topological, 3D, etc. Analysis : LR, PLS, RP, NN, Latent Class, SVM, etc. Combined (e.g. RP and LR) 4 Data Sets – Need for benchmarking data sets Binding : NCI Cell-based: none Tox : Mutagenicity, others Pooled : none 5 Software Needs Compound viewer/editor: MolViewer : JMP/smiles Descriptor Calculations: Dragon? Environment to code new algorithms: R, MatLab, MOE Standard stat methods: SAS, S, R, JMP Special Data Mining Code : ChemTree, CompChem vendors LeadScope, BioReason, etc. 6 Sequential Screening Paradigm Corporate Collection Stat Methods Chem Descriptors Model Data Sets Commercial Compounds Filters Virtual Compounds Initial Compounds Statistical Analysis Statistical Model Active Compounds Screen Compounds Selected Compounds 7 Initial Sample 1. 2. 3. 4. 5. Focused, knowledge-based No knowledge => Diverse Redundant to support analysis Large enough to start process Random is generally OK! 8 Rational Screening Goals 1. Optimize bioactivity, (increase selectivity) 2. Reduce the cost of experimental data 3. Increase the number of “active” classes 4. Increase speed of knowledge acquisition 9 Irrational Screening Goals 1. Find all active compounds. 2. Cost is no object. 3. Time is no object. 4. Find every active class. 10 Find Multiple Chemical Classes 11 Literature DM Hawkins, SS Young and A Rusinko. “Analysis of a Large Structure-Activity Data Set Using Recursive Partitioning” QSAR 16:296-302 (1997). MJ Valler and D Green. “Diversity screening versus focussed screening in drug discovery” Drug Discovery Today 5:286-293 (2001). MFM Engels and P Venkatarangan. “Smart screening: Approaches to efficient HTS” Current Opinions in Drug Discovery&Development 4:275-283 (2001) 12