Plate Statistics

We are running 384 well plates but we do not use the 2 outer rows/columns free because of concern related to edge effects and hence have not imaged them. I noted that in CellProfiler it can calculate the Z factor but it seems to need the a whole 96 or 384 well format. I was wondering if I am correct in my understanding and if there is a way to work around of having an incomplete plate. This may also be of use for late experiments as we are moving to spotting 1500 spots for siRNA later in the year.


For the calculation of the Z’-factor using the CalculateStatistics module, a plate layout isn’t required (although, of course, that’s the format this statistic is most often calculated for). What is required for CalculateStatistics is the identity of the positive and negative controls, typically through the use of per-image metadata. But otherwise, you are free to use any number of images that fall into each control category, with the caveat that you want to use as many images as possible to insure that the statistics are robust.

Could you let us know where you got the impression that the whole plates’ worth of images was needed. If it’s in the documentation, we could change it to make it more clear…