Calling Hits in a High-Throughput Screening Assay

• Activity Number: 513
• Type: Contributed
• Date/Time: Wednesday, August 4, 2010 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Minya Pu*+ and Karen Messer
• Companies: University of California, San Diego and University of California, San Diego
• Keywords: High-throughput screening ; False positive rate ; LOESS smoothing ; Mixture model ; Cancer vaccine adjuvant ; Computational chemistry

Abstract

A high-throughput screening (HTS) assay was performed to identify small molecule innate immune receptor in cancer vaccine adjuvant design. About 180,000 compounds were considered to find active substances ('hits'). False positive (FP) rates estimated from the raw data based on three types of control wells ranged from 0.5-1%. It was also observed from the positive control well there was plate to plate trend in z'. Cut-off thresholds were created using LOESS smoothed SDs from the positive control wells. Some existing methods including using a mixture model of Gaussian and gamma distributions were also implemented. FP rates were estimated from a set of validation runs and were compared among different methods. Final cut-off thresholds were adjusted based on both the FP rates and the target hit rate. We considered using computational chemistry methods to further screen hits in silico.