Abstract:[Objective] Staphylococcus aureus is a member of Gram positive bacteria, but is also one of common pathogens that are most difficult to deal with. It infects human skin, soft tissue, mucous membrane, bone and joint, especially in the nosocomial environment. Because studies on Staphylococcus aureus before were largely based on a single gene or protein, it is necessary to study this organism from the whole genome. [Methods] We used bioinformatics methods including five computational methods (phylogenetic profile, gene neighbor method, operon method, gene fusion method, interolog) to predict the protein interaction network of Staphylococcus aureus. [Results] We constructed the protein interaction network of Staphylococcus aureus and studied its function. [Conclusion] Through the network analysis, we found that the protein interaction network of Staphylococcus aureus was subject to scale-free property and a number of very important proteins, such as SA0939, SA0868, rplD. Through the analysis of the important cell wall synthesis and signal transduction and regulation local networks, we also found some very important proteins. Such information will help us better understand pathogenic mechanism and develop new drug targets of Staphylococcus aureus.