Presentation: P03026

Z'-factor (Zheng et al, 1999) is a useful tool for performing quality control during a screen, where a Z'-factor is often calculated for each plate. How many control wells per plate are required to get an accurate estimate of Z'-factor? We have explored this question using a simulation program. We considered two relationships between the variance in the high and low controls: constant variance (i.e. the standard deviations are similar for both sets of controls) and constant C.V (i.e. the standard deviations are proportional to the means of the controls). The constant CV situation gave a larger spread of Z'-factors for a given mean Z'-factor and number of control wells. For assays with a mean Z'-factor >=0.8, four wells of each control were sufficient for a good estimate. Similarly, eight wells of each control were sufficient for Z'-factors >=0.6, but 16 or more wells were required for poorer assays.

Close Window X