Stochastic modelling of bioreactors for Escherichia coli fermentation with recombinant protein production

Mariana Assunção Albino

Key information

Authors:

Mariana Assunção Albino (Mariana Assunção Albino)

Supervisors:

Susana de Almeida Mendes Vinga Martins (Susana de Almeida Mendes Vinga Martins); Krist V. Gernaey

Published in

11/30/2022

Abstract

The need to develop models capable of describing key processes in the biomanufacturing industry has been rising as this industry is in a rapid transition towards more automation and digitalisation. Fermentation for recombinant protein production is one key process, and as all biological processes, it is subject to the stochasticity of living beings. Because of this, it is important that these models take this stochasticity into account. Gillespie’s algorithm is a method of doing so, with exact numerical calculations within the framework of the stochastic formulation without the need of solving the Chemical Master Equation (CME). The aim of this project was to develop a model for fed-batch fermentation of Escherichia coli for recombinant protein production that was able to capture the uncertainty associated with this bioprocess. This model was developed for two case study proteins, Green Fluorescent Protein (GFP) and recombinant human Growth Hormone (rhGH). This goal was achieved by using deterministic models available in the literature and adapting them to the Stochastic Simulation Algorithm (SSA). It was shown that the developed model was applicable and could describe well the production dynamics. By confronting the model with experimental data, some parameters were re-estimated using the ABC Rejection sampler method. In summary, it is possible to derive conclusions regarding the uncertainty associated with each species present in the model. These results provide one more successful application of this algorithm, and stand as a foundation to expand and improve its use for industry relevant processes.