Yapısal protein-protein ağlarında aradalık analizi

Protein-protein interaction networks (PPIN) that are maps of physical interactions between proteins are important and essential abstractions to understand biological functions and processes. Nodes represent proteins and edges represent pairwise interactions between proteins in a PPIN. Analysis of the proteins with high betweenness score is important in terms of their role in biological processes, molecular functions and cancer. PPINs do not indicate structural and chemical features of interactions. Our goal is to construct a structural PPI network (SPPIN) and analyze the proteins with high betweenness score found by shortest path, flow and random walk betweenness algorithms. To construct the SPPIN, we assigned weights on the edges of the network to reflect the effect of the simultaneous and mutually exclusive interactions and adjusted these weights for different betweenness algorithms separately. We also generated random samples of the network by selecting a set of interactions that can co-occur for each sample and calculated betweenness scores. We observed that %37.5 of bottleneck proteins of SPPIN are different from bottleneck proteins of PPIN found by shortest path betweenness centrality algorithm. This percentage is %18.75 for flow betweenness and % 31.25 for random walk betweenness algorithm. Bottleneck proteins of SPPIN which are different than bottleneck nodes of PPIN have roles in biological processes such as DNA repair and stimulus and have molecular functions such as hormone receptor binding and transcription activation. In addition, we observed that betweenness is positively correlated with number of mutations on interfaces.