Abstract
Protein-protein interactions (PPIs) underlie most cellular functions, where pathogens interact with hubs and bottlenecks of the host PPI network. Modelling proteins as graphs allows us to study PPI as phenomena on irregular but structure geometry. Graph-based machine learning allows inference of structure and exploitation for PPI prediction. This project will generate a host-pathogen interactome map using graph-based machine learning for coccidiosis, a disease caused by Eimeria with annual losses exceeding €2 billion. Current vaccines are sub-optimal and new subunit vaccines are required. The model generated will significantly advance our ability to identify vaccine targets and utilize host-networks to optimize responses.
References
1 Blake DP. “Eimeria genomics: Where are we now and where are we going?” Vet Parasitol. 2015 Aug 15;212(1-2):68-74.
2 Pastor-Fernández I., Kim S. et al. “Development of cross-protective Eimeria-vectored vaccines based on apical membrane antigens” Int J Parasitol. 2018 Jun;48(7):505-518.
3 Nepusz T., Yu H., et al. “Detecting overlapping protein complexes in protein-protein interaction networks.” Nature Methods. 2012 Mar 18;9(5):471-2.
4 Bronstein, Michael M., et al. “Geometric deep learning: going beyond euclidean data.” IEEE Signal Processing Magazine 34.4 (2017): 18-42.
5 Cuesta-Astroz Y., Oliveira G. “Computational and Experimental Approaches to Predict Host-Parasite Protein-Protein Interactions.” Methods Mol Biol. 2018;1819:153-173.