University of Ottawa

Purpose

Bioreactors play major role in manufacturing processes of biologics. While their design has been well documented for traditional fermentations for production of simple biomolecules, their use for production of complex biologics using eukaryotic systems such as HEK-293 human cells remains very empirical. Biologics production in bioreactors is performed in a high dimensional design space requiring a high and broad level of expert knowledge not necessary captured systematically in the prevailing scientific and technical literature. Additionally, biological products such as exosomes and viral vectors (AAV) are complex biological structures that have demonstrated recently significant value in the treatment of a number of cancers and hereditary diseases. Their manufacturing in bioreactor at scale in quantity and quality that meets the preclinical and clinical needs significantly limits their use and growth potential as therapeutic tools. This create an opportunity to integrate the expert knowledge accumulated in the databases and exploit the advances in AI to accelerate the design and optimization of the production of these two key biologics (exosomes and AAV) through the implementation and operation of a digital-twin bioreactor connected to a physical bioreactor using a series of sensors in constant communication and feed-back. This project will focus on AI and machine learning for development of insillico digital twin.