Protein data analytics for protein-protein interaction prediction and drug design

Interactions among proteins are essential to many biological functions in living cells but experimentally detected interactions represent only a small fraction of the real interaction network. Computational protein interaction prediction methods have become important to augment the experimental methods (e.g. two-hybrid and TAP-tagging). Our parallel Protein Interaction Prediction Engine (PIPE) has been able to perform the first ever scan of the entire human proteome (3 months of 24/7 computation on a 1,000 processor cluster). Several previously unknown protein interactions that our method predicted have later been experimentally verified. An independent comparison study published in BMC Bioinformatics determined that PIPE outperforms other protein interaction prediction methods by a wide margin. Our most recent work is on engineering inhibitory proteins which interact with selected target proteins. Such inhibitory proteins play an important role in the design of new drugs such as biologics for cancer treatment.

Current research projects include:

  • A collaboration with the Ottawa Hospital on new stem cell therapies for Muscular Dystrophy.
  • A collaboration with Agriculture Canada on soy bean resistance to Canadian winter temperatures.
  • A collaboration with WITS University in South Africa on the role of protein interactions in the evolutionary development of multi-cellular organisms.

We have also launched a new spin-off company, Designed Biologics Inc, which contributes towards the design of new targeted drugs based on engineered synthetic inhibitory proteins (e.g. biologics for cancer treatment).