University of Ottawa

Published purpose

Inferring the structural properties of a protein from its amino acid sequence is a challenging yet important problem in biology. Discovering these structures is important, since they determine a wide array of protein functions. However, experimental structure determination is costly and, as a result, atomic structures have only been determined for a tiny fraction of known proteins. The objective of this project is to develop an AI system capable of predicting protein-protein interaction and physicochemical properties of proteins solely from their amino acid sequences without any recourse to expensive, time-consuming and computationally prohibitive methods such as homology modelling, crystallography, nuclear magnetic resonance, molecular dynamics and macromolecular docking. The system, will allow for the design of proteins with specific properties by means of AI-based reverse engineering.