A machine learning approach to decipher protein-protein interactions in human plasma

Principal Investigator: Benoit Coulombe
Theme : Health
Competition : 2017 Competition: IVADO's Grants for fundamental research projects
Status : Completed
Start : Apr. 1, 2018
End: Mar. 31, 2020
Budget : $158,625.00


Certain proteins circulating in human blood could be used as clinical biomarkers given that when making a diagnosis or a prognosis or identifying people who will benefit from precision medical treatment, blood and its proteins are easily accessible compared to other tissues and organs. Generally speaking, to perform their functions, proteins interact with other molecules, including other proteins. These protein-to-protein interactions provide valuable information about the role and function of a protein and may also lead to the discovery of new biomarkers for diseases involving the protein of interest. Although methods for identifying protein-to-protein interactions in cell models using biochemical approaches are well established, in human plasma this identification has so far proved very difficult. The lack of appropriate biochemical controls, which by their nature involve significant variability, makes it difficult to assess these interactions. We therefore propose to develop a new machine learning approach that will extract the relevant signal from the variable controls in order to confidently decipher the interactome of clinically relevant proteins in the human bloodstream with the ultimate goal of identifying new biomarkers.


Lead Genome Centre: Génome Québec


Partner: IVADO



Mathieu Lavallée-Adam University of Ottawa
Marie-Soleil Gauthier IRCM