Analyzing Protein Interaction Networks via Random Graph Model

Many complex systems may best be described as networks, which we can use graph theory to analyze their topological properties. In an organism, protein-protein interactions may also be mapped into complex network. Here we use random graph theory to analyze seven different organism protein interaction networks. Three topological properties (degree distribution, clustering coefficient and average shortest path) were used to characterize these networks. The logarithm of the node degree distribution vs. the logarithm of the node degree plot shows that all seven species follow a power-law distribution quite well. In addition, we also obtained the relatively high clustering coefficient of these protein interaction networks. The distance between two nodes of these protein interaction networks indicates that it is quite short comparing with the large network size. The plot of the logarithm of the frequency vs. the shortest path length also indicates that the shortest path length distribution follows a normal distribution quite well.