Abstract
The research on host-pathogen protein-protein interactions (PPIs), specifically HIV1-human PPIs becomes one of the most challenging areas of medical science for antiviral drug invention. In this paper, we propose a pattern mining based approach to predict novel interactions between HIV-1 and human proteins with an estimated confidence based on the experimentally validated known interactions integrated with protein signatures and gene ontology (GO) annotations (biological process, cellular component and molecular function) of human proteins. It results in predicting more potential interactions along with the corresponding signatures and GO terms. We validate our predicted interactions by finding evidences from the literature and comparing with the predictions made by different computational approaches. We believe that our predicted information on PPIs enlightens the PPI research field with greater knowledge and better understanding of viral replication process; subsequently enhancing the discovery of new drug targets.
