CARTE Industry Speaker Seminar: Active Learning for Optimizing RNA Based Vaccines and Therapeutics


Friday, October 28, 2022
12:00pm-1:00pm


Myhal Centre, Room 380
55 St George St.


CARTE Industry Speaker Seminar Series welcome Michael Bailey, Computational Scientist at Sanofi Data and AI Center of Excellence in Toronto, for the first in-person seminar of 2022-23 academic year.

Registration: To register, please see here. Capacity is limited. Please register early to secure your spot.

Abstract: RNA vaccines saved the world from COVID. But vaccines are just one of several potential uses for this breakthrough technology. As the world leading vaccine manufacturer and one of the largest pharma companies, Sanofi has recently launched its RNA Center of Excellence to lead the way in the development and use of RNAs for vaccines and therapeutics. While promising, the use of mRNA raises several new computational challenges. These involve issues related to representation and search in the exponential space of mRNA molecules, the design and optimization of their lipid vehicles and the ability to predict human response from non-human models. I will discuss these challenges and will also present methods we developed to address these issues. Our methods use deep language models and graph neural networks for representation and couple them with active learning approaches for optimization. By developing an experimental-computational strategy we were able to obtain more accurate RNA and lipid combinations while still reducing the time and cost to optimize vaccines for new variants.

Speaker Bio: Michael Bailey is a Computational Scientist at Sanofi. With a background in Mathematics (doing a Ph.D. at the University of Toronto), he transitioned into Machine Learning later in his career, looking to work on real-world problems. In his current role at the new Sanofi Data and AI Center of Excellence in Toronto, he works on Machine Learning problems to support the discovery of new therapies.

 

© 2024 Faculty of Applied Science & Engineering