High-Throughput Screening Data Analysis

An overview over the role and past evolution of High Throughput Screening (HTS) in early drug discovery is given and the different screening phases which are sequentially executed to progressively filter out the samples with undesired activities and properties and identify the ones of interest are outlined. The goal of a complete HTS campaign is to identify a validated set of chemical probes from a larger library of small molecules, antibodies, siRNA, etc. which lead to a desired specific modulating effect on a biological target or pathway. The main focus of this chapter is on the description and illustration of practical assay and screening data quality assurance steps and on the diverse statistical data analysis aspects which need to be considered in every screening campaign to ensure best possible data quality and best quality of extracted information in the hit selection process. The most important data processing steps in this respect are the elimination of systematic response errors (pattern detection, categorization and correction), the detailed analysis of the assay response distribution (mixture distribution modeling) in order to limit the number of false negatives and false discoveries (false discovery rate and p-value analysis), as well as selecting appropriate models and efficient estimation methods for concentration-response curve analysis.