Single assay-wide variance experimental (SAVE) design for high-throughput screening

• DOI: 10.1093/bioinformatics/btt538
• Publication Date: 2013-09-21T04:54:29Z
• Authors: Carl Murie; Caroline Barette; Laurence Lafanechère; Robert Nadon
• Keywords: Replication; False Discovery Rate; Design of experiments

Abstract Motivation: Advantages of statistical testing high-throughput screens include P-values, which provide objective benchmarks compound activity, and false discovery rate estimation. The cost replication required for testing, however, may often be prohibitive. We introduce the single assay-wide variance experimental (SAVE) design whereby a small replicated subset an entire screen is used to derive empirical Bayes random error estimates, are applied remaining majority unreplicated measurements. Results: SAVE able generate P-values comparable with those generated full data. It performs almost as well model t-test duplicate data outperforms commonly Z-scores standard t-test. illustrate approach simulated molecule interfering RNA screens. provides substantial performance improvements over only slight increases in cost. Contact: [e-mail] Supplementary information: available at Bioinformatics online.