Comparisons of the protein-protein interaction networks constructed from the DIP database with different version

Protein-protein interactions (PPI) play the most important role in the interactomics systems in living cells. The protein- protein interaction network (PIN) is displayed as an undirected graph, in which the biological functions of proteins were properly encoded. In this paper, the complex network theory was used in analyzing the protein-protein interactions from Database of Interacting Protein (DIP). The general topological features: right-skewed degree distributions, relatively small clustering coefficients and short characteristic path lengths were introduced to character the properties of the giant component of PIN (GCOP). While, tests were done on the degree distributions of GCOPs in order to find the best-fit power low model. At last, the Comparisons were made between the different PINs constructed on the DIP database with different version. The results led to the findings that the PINs lack modularity and tree structure should be the framework of these topologies. Moreover, the PINs are not well connected, the PPI data is substantially incomplete, the PPI between different components should be substantially explored. © 2011 IEEE.