FIVE STEPS TO PHENOTYPIC DATA FROM AUTOMATED QUANTITATIVE IMAGING
Typically, each assay generates hundreds of output features that describe the imaged phenotypes.
We use optimization algorithms to select the most robust output features for each assay developed. In the simple example shown, the loss of nuclear puncta was the most striking change between control and treated.
The output feature “Puncta Intensity” captures the total amount of signal from the protein that formed puncta within the nucleus. Since this output feature is influenced by the number, area and brightness of puncta it provides a larger assay window (the statistical significance of the difference between control and treatment conditions) than other features such as puncta count.