A transition probability cell cycle model simulation of bivariate DNA/bromodeoxyuridine distributions.

Abstract

The transition probability cell cycle model is extended to describe both cell cycle variability and incorporation of bromodeoxyuridine (BrdUrd). The model can simulate BrdUrd uptake in both pulse-chase and continuous-labeling experiments. With the use of a random transition, variability due to cell cycle progression is distinguished from dispersion due to staining and machine errors in the generation of bivariate DNA/BrdUrd distributions. In a comparative test with a compartmental model developed by Yanagisawa et al. (Cytometry 6:550-562, 1985), the present model is shown to provide realistic simulations with fewer model parameters and with the ability to describe gradual asynchronization of cell cycle cohorts. With model predictions as the basis, a simulated experiment is performed to illustrate the difficulty in analyzing bivariate distributions. The simulated experiment illustrated that it is very easy to overestimate unlabeled cell fractions, and, as a result, matching the periodicity of the cell cycle cohort movements is more reliable in the estimation of model parameters.