Classification of phenotypic subpopulations in isogenic bacterial cultures

Identification of different phenotypes in bacterial cultures and distinguishing between them is often a cruvial step when analyzing subpopulations of bacterial colonies. Often the calculation or estimation of a threshold value for data obtained from a fluorescence marker is sufficient for a proper differentiation. Unfortunately, a threshold based classification might fluctuate strongly due to effects caused by stochastic gene expression. Therefore we propose a statistical strategy for the determination of cell types based on gene expression profiles over a whole generation time which is less susceptible to noise than a fixed threshold value.