A New Effective Method for Elimination of Systematic Error in Experimental High-Throughput Screening

Abstract

High-throughput screening (HTS) is a critical step of the drug discovery process. It involves measuring the activity levels of thousands of chemical compounds. Several technical and environmental factors can affect an experimental HTS campaign and thus cause systematic deviations from correct results. A number of error correction methods have been designed to address this issue in the context of experimental HTS (Brideau et al., J Biomol Screen 8:634–647, 2003; Kevorkov and Makarenkov, J Biomol Screen 10:557–567, 2005; Makarenkov et al., Bioinformatics 23:1648–1657, 2007; Malo et al., Nat Biotechnol 24:167–175, 2006). Despite their power to reduce the impact of systematic noise, all these methods introduce a bias when applied to data not containing any systematic error. We will present a new method, proceeding by finding an approximate solution of an overdetermined system of linear equations, for eliminating systematic error from HTS screens by using a prior knowledge on its exact location. This is an important improvement over the popular B-score method designed by Merck Frosst researchers (Brideau et al., J Biomol Screen 8:634–647, 2003) and widely used in the modern HTS.