|Marc Zimmermann1, Sally Ann Hindle2, Thorsten Naumann3, Hans Matter3, Gerhard Hessler3, Karl-Heinz Baringhaus3, Christian Lemmen2, Marcus Gastreich2, and Matthias Rarey4. (1) SCAI, Fraunhofer Gesellschaft, Schloss Birlinghoven, Sankt Augustin, 53757, Germany, (2) Chemoinformatics, BioSolveIT GmbH, An der Ziegelei 75, 53757 St. Augustin, Germany, (3) Molecular Modelling, aventis pharma Deutschland GmbH, Industriepark Höchst, Frankfurt/Main, 65926, Germany, (4) Zentrum fuer Bioinformatik, Univ. Hamburg, Bundesstrasse 43, Hamburg, 20146, Germany|
|The pursuit of innovative drug candidates has driven technological progress in the field of high throughput screening (HTS). However, the HTS process generates vast amounts of data often plagued with noise. There is increasing demand for approaches that sensibly interpret such data and to this end we have developed a novel tool called "HTSview". The software combines data mining techniques with pharmacophoric concepts. Based on the Feature Trees descriptor and lacking the necessity of 3D alignments, HTSview is extremely fast. Clustering algorithms and classification methods quickly facilitate the focussing of data. Extraction of SAR information to form biophore models optimized in conjuction with machine learning techniques is also a central capability of the tool. We tested the tool on data available from both in- house sources and the literature. Virtual screening studies with biophore models demonstrated the scaffold hopping potential of HTSview – an important concept in lead idea generation.|
Computational Chemistry in Drug Discovery: Are High Information Content Calculations Better than Low Information Content Calculations?
8:00 AM-11:40 AM, Monday, September 8, 2003 Javits Convention Center -- 1E04, Oral