Enhanced kinetic model parameters for xylitol bioproduction from Candida mogii ATCC 18364

Xylitol is a common sweetener used in the dental and pharmaceutical industries. This molecule can be obtained from renewable sources and, is also a building block with the potential to produce a wide variety of chemical compounds. However, xylitol bioproduction is still economically unattractive and, there is a need for investigating optimal process design and operation. For that, a mathematical model that accurately predicts the xylitol fermentation is fundamental to reduce uncertainties during process production. In this study, a framework for structural and practical parameter identification is implemented to increase the number of identifiable kinetic parameters, using a mathematical model composed of 5 states and 11 parameters. The methodology uses well-established methods that unfortunately are not commonly used together such as a) experimental data processing, b) structural identifiability analysis, c) enhanced parameters identification using a nested optimization approach for self-tuning stochastic optimization and, d) validation. The employed methodology allows to determine that the full set of parameters is locally identifiable, and the number of identifiable parameters was increased from 4 to 10, also with reduced confidence intervals. The increased confidence in the model parameters brings interesting insights into this fermentation, which might permit peforming further and more certain analysis.