Analysis and application of large-scale protein-protein interaction data sets

Protein-protein interactions play key roles in cells. Lots of experimental approaches and in silica methods have been developed to identify and predict large-scale protein-protein interactions. However, compared with the traditionally experimental results, the high-throughput protein-protein interaction data often contain the false positives in high probability. In order to fully utilize the large-scale data, it is necessary to develop bioinformatic methods for systematically evaluating those data in order to further improve the data reliability and mine biological information. This review summarizes the methodologies of analysis and application of high-throughput protein-protein interaction data, including the evaluation methods, the relationship between protein-protein interaction data and other protein biological information, and their applications in biological study. In addition, this paper also suggests some interesting topics on mining high-throughput protein-protein interaction data.