The problem of cell debris in pulse-cytophotometry: a probability theoretic model

The problem of cell debris in pulse-cytophotometry is considered on the basis of a probability-theoretic model and the results are compared with the model of a pure exponential function for the portions of the DNA fragments in the channels of the flow cyclometer, which is generally used in practice. The model is based on the assumption that the possible decomposition of the DNA content of a single cell into some parts as a consequence of the necessary preparation of the cell material for measurement can be interpreted as a Poisson point process. Therefore the number of divisions of the DNA content of a cell is assumed as a Poisson random variable. It follows a well-defined distribution function of the DNA content in the common population of intact cells and fragments depending on the apriori distribution before pretreatment. This distribution determines a theoretical histogram which can be compared with the measurements. The results corroborate the assumption of a pure exponential course for the portions of DNA fragments only in the area of very small DNA contents. For greater DNA contents there are differences between the pure exponential course and the model described here depending on the a priori distribution and on the intensity of the cell disintegration process. As a byproduct, an estimation of the fraction of disintegrated cells is obtained with respect to the fraction of fragments. A corresponding computer program was applied to a number of experimental histograms. The results have shown that the theory describes the measured histograms satisfactory in most cases.