Carleton University

Published purpose

AI systems that can describe and predict behavior of biological networks of cells, tissues and organisms will allow more accurate, faster and less expensive innovations in life sciences while at the same time ensuring predictable outcomes. In particular, the full potential for safe and efficient utilization of gene editing and live cell therapies requires an ability for controlled design of these modalities with simulations that allow testing and optimization under different conditions in both production and utilization. The project will combine machine learning and mechanistic modeling for the development of simulation of biological systems and link these simulations with learning algorithms for testing and optimizing different designs. The focus of this 2 year project, will be on the development of methods for the simulation and design of desired characteristics of exosomes, as carriers of gene therapies, and development of methodologies for building simulations for HEK293, as major cells for bioreactor production and iPSC cells as major cells in stem and cell therapy development. This project will focus on a probabilistic framework for merging data-driven methods with theory-based models. The project will also evaluate the applicability of state-of-the-art automatic metabolic quantification methods from the literature to NMR data currently being collected by the NRC.