The viral receptor is key to the viral entry of host cells as well as the development of anti-viral drugs. Unfortunately, identification of virus receptors is challenging and only a small proportion of human viruses has the viral receptor identified due to the limitation of current experimental methods of identifying viral receptors, which hinders the human virus research. Based on our previous studies, we have built a machine-learning method for predicting the receptorome of human viruses. However, it is still difficult to identify receptors for a given virus. Here, we investigated the prediction of human virus-receptor interactions by integrating protein-protein interaction (PPI) prediction methods. Three computational methods were used in the study: CAMP, InterPred and HostMimic. The known interactions between human viruses and receptors were used in evaluating the performance of these computational methods and in integrating them. For each method, it was used to predict the interaction potential between a viral RBP and each human cell membrane protein in the human virus recepterome. Then, all these human cell membrane proteins were ranked by the interaction potential (a score) and the rank of the known virus receptors were obtained.