In the biotechnology industry the trade-off between speed to market and process efficiency is constrained by good manufacturing practices and regulation, which effectively fix future manufacturing costs early in process development. Given the challenge of applying predictive process engineering technologies to the complexity of upstream bioprocesses along with the high public health and economic benefits of speed to market, the industry has consequently been slow to adopt challenging predictive technologies, preferring phenomenological approaches to process development. As predictive technologies have evolved however, it’s become advantageous to apply them to upstream bioprocess design and optimization.<br/><br/>Here we present a computational framework for mechanistic modeling and simulation of CHO cell culture allowing faster, more robust and optimal in-silico process development. As an exemplary application of this framework, we have developed and calibrated a simple CHO cell culture model consisting of: A) a constraint-based model of the primary carbon-nitrogen metabolism, B) transmembrane transport kinetics with temperature, dissolved oxygen, pH and osmolarity dependence, C) stress-dependent energy of maintenance, and D) a mass balance bioreactor model. We have applied the model to a dynamic Flux Balance Analysis (dFBA) problem and calibrated it with experimental process characterization data for a pipeline process using a non-dominated sorted Genetic Algorithm (NSGA2).

Computational framework for mechanistic modeling and simulation of upstream bioprocesses

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.