Analysis of High-Throughput Screening (HTS) Assay Data Based on Preliminary Test Inference

• Activity Number: 619
• Type: Contributed
• Date/Time: Thursday, August 2, 2012 : 8:30 AM to 10:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Changwon Lim*+ and Pranab K. Sen and Shyamal Peddada
• Companies: Loyola University Chicago and The University of North Carolina at Chapel Hill and National Institute of Environmental Health Sciences
• Keywords: Dose-response study ; False discovery rate (FDR) ; Heteroscedasticity ; Hill model ; M-estimation procedure ; Nonlinear regression model

Abstract

Quantitative HTS assays use cells to screen thousands of compounds rapidly. Data from qHTS assays are then evaluated using nonlinear regression models and decisions regarding toxicity are made using the estimates of the parameters. Since thousands of compounds are simultaneously evaluated in a qHTS assay, it is not practically feasible for an investigator to perform residual analysis to determine the variance structure before performing statistical inferences on each compound. In this talk I describe preliminary test estimation based methodology which is robust to the variance structure as well as any potential outliers. Performance of the the proposed methodology is evaluated in terms of FDR and power using a simulation study mimicking a real qHTS data. Of the two methods currently in use, the simulations studies suggest that one is extremely conservative with very small power in comparison to the proposed PTE based method whereas the other method is very liberal with FDR. In contrast, the proposed PTE based methodology achieves a better control of FDR while maintaining good power. The proposed methodology is illustrated using a data set obtained from National Toxicology Program.