| New Methods for Statistical Analysis and Data Correction in HTS |
|
Presentation: P01015 Session: Cheminformatics, Library Design and Virtual HTS - Poster SessionDmytro Kevorkov, Vladimir Makarenkov and Pablo Zentilli, Universite du Quebec a Montreal Presenting Author: Pablo Zentilli, University of Quebec in Montreal - Canada The hit selection procedure in HTS screens requires routines for quality control and data correction. We present new methods and corresponding software, called Biomatrix, intended for statistical analysis and visualization of HTS data. These methods are based on the statistical analysis of signal variation within plates of an assay. They allow researchers to analyze experimental HTS assays, detecting and characterizing systematic errors. A k-means partitioning procedure divides original dataset on several subsets with similar systematic patterns. Two algorithms for data correction are implemented in Biomatrix. The sigma distribution analysis allows users to adjust sigma level for hit selection for each particular well. Biomatrix displays color plate maps, tables, and histograms. The detailed description of the implemented algorithms can be found in the paper by Kevorkov and Makarenkov (Journal of Biomolecular Screening, 2005). The executable file for Windows and the detailed description of the software are freely available on our website: http://www.labunix.uqam.ca/~makarenv/hts.html. |