Real-time estimation of biomass and specific growth rate in physiologically variable recombinant fed-batch processes

Area of Science:

• Biotechnology
• Bioprocess Engineering
• Real-time Analytics

Background:

• Accurate real-time biomass measurement is crucial for bioprocess optimization.
• Protein expression in recombinant processes significantly alters cell physiology, complicating biomass estimation.
• Existing methods often require offline sampling and extensive training data, limiting their applicability in dynamic environments.

Purpose of the Study:

• To develop a generally applicable, real-time soft-sensor for biomass quantification in induced recombinant fed-batch cultures.
• To overcome limitations of current biomass estimation techniques, particularly regarding cell variability and dynamic process changes.
• To provide a robust method for process development and control by enabling real-time monitoring without offline sampling.

Main Methods:

• Implementation of a first-principles-based soft-sensor model.
• Real-time quantification of biomass in induced recombinant fed-batch processes.
• Comparison with state-of-the-art methods for biomass concentration and specific growth rate (µ) estimation.

Main Results:

• The soft-sensor accurately quantifies biomass in real-time, avoiding offline sampling and the need for large training datasets.
• The method demonstrated good generalization capabilities, adapting to different growth stoichiometries and induced culture modes.
• Automatic detection of gross errors, such as incorrect stoichiometric assumptions or sensor failures, was achieved.

Conclusions:

• The developed soft-sensor provides a reliable and adaptable solution for real-time biomass quantification in complex recombinant bioprocesses.
• This approach enhances process understanding and control by providing a key variable without traditional limitations.
• The method's ability to handle variable model coefficients and adapt to different conditions makes it valuable for dynamic process development.