Predicting novel protein-protein interactions between the HIV-1 virus and homo sapiens

Abstract

The HIV-1 virus affects millions of people around the world. Identifying novel protein-protein interactions (PPIs) between HIV and humans would lead to a better understanding of the virus and possibly to new treatment targets. The Proteinprotein Interaction Prediction Engine (PIPE) is a broadly applicable, highly precise, and computationally efficient method of predicting PPIs. Here, PIPE is used to predict new host-virus protein interactions in order to generate new testable hypotheses and to guide future biological experiments. In total, 229 new interactions were predicted at high confidence, with an estimated recall of 22.5% and specificity of 99.95%. Some of these interactions may be verified experimentally in the future.