DNA-encoded Measurement and ML-based Prediction of Quantitative Protein Glycan Interactions

Glycans densely coat the surface of every cell in all Kingdoms of life but studying the function of glycans is challenging because glycans fundamentally lack of 1:1 correspondence between DNA and glycan structure, owing to the fact that cellular glycosylation is not directly encoded by the genome. To solve this problem, we developed genetically-encoding platforms that leverage DNA sequencing to study glycan-protein interactions on the surface of live cells ex vivo and in vivo. Using such glycan-protein interaction data in a meaningful model that can integrate known protein:glycan interactions and predict yet unknown ones is yet another challenge that we are solving with the help of machine learning models.

Featured Speaker

 

Ratmir Derda, PhD
Professor of Chemistry, University of Alberta, Canada