HCS: z-score with respect to negative control

I’ve got a question with regard to plate normalisation in the context of high-content screening.

In the presence of control wells people typically calculate normalised percentage inhibition (NPI) or percent of control (POC) [1]. Without controls a z-score normalisation is used, where the sample mean and standard deviation is calculated from all (non-control) wells. The disadvantage of the z-score is that it’s biased when the library is quite active and cannot effectively act as the negative control.

But I’m wondering whether it’d be “statistically kosher” to calculate the z-score of sample wells with respect to the actual negative control? Thus, the mean and sd used to calculate the z-score would be calculated from negative control wells. This would be an alternative to POC when only the negative control was present on a plate.

This is perfectly legitimate. The score would then indicate how different from control a test well is.