Structured Modeling and Parameter Identification of Budding Yeast Populations

The present paper investigates the possibility to obtain, from experimental data, reasonable and plausible parameter estimates of a particular model of budding yeast population, which accounts for unequal division and structural heterogeneity by genealogical age. The optimal parameter estimates are produced by fitting distributions predicted by the model to protein distributions obtained by flow-cytometry from steady-state cultures of S. cerevisiae at different growth conditions. Calculations were carried out, with the aid of a standard Levenberg-Marquardt algorithm for nonlinear least-squares, by minimizing the sum of squares between measured and simulated data. The best fit values provide information, related to the cell cycle, which satisfactorily agrees with previous experimental studies. Moreover, the reliability of the model is investigated through a statistical analysis based on the accuracy of estimation.