Predicted high-confidence human－HSV-1 protein-protein interactions

To complement experimental identification of human-HSV-1 protein-protein interactions (PPIs), we developed an integrative strategy to predict proteome-wide PPIs between human and HSV-1. For each human-HSV-1 protein pair, four popular PPI inference methods, including interolog mapping, the domain-domain interaction-based method, domain-motif interaction-based method and machine learning-based method, were optimally implemented and employed to generate four interaction probability scores, and then these four scores were further integrated into a final probability score. Only the high-confidence PPIs we predicted (i.e. probability score > 0.5) are displayed in the database, which covers 10,432 interactions between 4,546 human proteins and 72 HSV-1 proteins.