Statistical evaluation of a self-deconvoluting matrix strategy for high-throughput screening of the CXCR3 receptor

In high-throughput screening (HTS), compounds can be tested in self-deconvoluting matrices (SDMs) of 10 compounds per well. The SDM setup is based upon a systematic mixing of compound samples such that each compound appears twice in the screening assay, in two independent mixtures. In order to test the quality of the SDM approach, we compared it with a standard single-compound screening approach. In a CXCR3 scintillation proximity assay, we performed five multiple screening trials of 26,400 compounds at a 10 mu M screening concentration to estimate false positive and false negative rates in the compound population. No potent hits ( < 6.2 mu M IC50) were missed in any screening method. Forty-eight percent of all actives were found in every screening trial independent of compound handling method. The SDM strategy had an average of 25 false positives and 15 false negatives as compared with an average of 34 false positives and 15 false negatives with a more conventional single-compound screening approach. Most of the variability resulted from day-to-day variation around the hit cutoff criterion, rather than from any particular screening technique. In the two most extreme examples, a compound with a 7.5 mu M IC50 was missed in one out of two mixture trials, and a compound with a 6.2 mu M IC50 was missed in one out of three single-compound trials. In the CXCR3 assay presented herein, the SDM screening method had better predictive value than the single-compound screening approach.