The Use of Design Specificity in Standardized Mean Difference for Analysis of High throughput RNA Interference Screens

RNA interference (RNAi) high-content screening (HCS) enables massive parallel gene silencing and is increasingly being used to reveal novel connections between genes and disease-relevant phenotypes. The application of genome-scale RNAi relies on the development of high quality HCS assays. Strictly standardized mean difference (SSMD), introduced by Zhang et al. [1], provides a possibility for hit selection in HCS experiments. This method has relied on normal approximation, which works in the primary screens considering positive and negative controls. This paper describes a new extension of the SSMD, which integrates bioinformatics RNAi on-target analysis results for both the SSMD-based testing process and the use of SSMD as a ranking metric for hit selection by using additional controls generated from RNAi libraries.