AbstractsBiology & Animal Science

Saccharomyces cerevisiae is one of the yeasts with major economic, social, and health significance in human culture. Depending on the growth conditions experienced by the cell, S. cerevisiae growth can proceed via fermentative, respirative, or respirofermentative metabolism. Scar formation, unequal division, a limited replicative lifespan, and increase in cell size commensurate with the cell's replicative age are individual characteristics of this yeast affecting the performance of bioprocesses. These characteristics increase the complexity of predictive models and introducing them with ease into a continuous model is not realistic. Nevertheless, an individual-based model is able to accommodate this complexity in a single computational model. Once an individual model is implemented, it has to be parameterized, calibrated, and its adequacy assessed. All these processes ideally require a high number of both individual and system-level experimental observations. The aim of the present thesis is to advance the development of an individual-based methodology to tackle the study of microbial systems driven by the relevant yeast S. cerevisiae. The adequacy of INDISIM-YEAST, an existing individual-based model of a generic budding yeast, is first assessed. In order to obtain valuable individual-based observations to support the desired individual-based methodology, the diversity of S. cerevisiae in experimental individually-oriented observations under different growth conditions and at different stages of the growth curve is verified and assessed. A quantitative individual-based model focusing on the fermentative (anaerobic) growth of the yeast S. cerevisiae has been designed, implemented in Fortran 90, and termed INDISIM-Saccha. The developed model is parameterized, calibrated, its adequacy evaluated, and used to assess in silico ethanol production by means of virtual experiments. The calibration procedure, and the performance and analysis of the data from the virtual experiments is undertaken using the statistical programming language R. The model adequacy is assessed by testing several model predictions both at a system level (glucose depletion, population growth curves) and single-cell level (fraction of budded cells, genealogical age distribution, and cell diameter distribution evolutions). Individual cell diameter observations obtained within the present thesis play a significant role in this assessment. Results of the virtual experiments suggest that differences in cell size distribution can drastically affect the performance and productivity of fermentations, and encourage routine characterization of the inocula in the biotechnological industry. INDISIM-Saccha is also adapted to take into account the aerobic growth of S. cerevisiae and contrasted with two experimental trials with different oxygen levels in the medium. The preliminary simulated results achieved with the model suggest that the approach also has the potential for reproducing aerobic batch cultures of S. cerevisiae. This represents a further step in…