English: Figure 2 from the article Functional and topological characterization of protein interaction networks. Caption: Large-scale characteristics of the protein interaction databases. (a) Degree distribution of the four databases, shown on a log-log plot. Note that all datasets have a power law tail, indicating that the underlying network has a scale-free topology. The solid line is obtained from the fitting to the function P(k)~k^(-gamma) to the DIP data, the best fit indicating gamma~2.5 for DIP data set. (b) Distribution of the clustering coefficient for the four studied databases shown on a log-log plot. The straight line has slope -2. (c) Cluster size distribution for the four databases shown on a log-log plot. Apart from the points corresponding to the giant component (for right) the P(n) curves follow a power law. The solid line is obtained from the least square fitting to P(n)~n^(-alpha) for the MIPS dataset, providing alpha = 3.4.