Institute of Information Science, Academia Sinica

Protein-protein interaction network offers a conceptual framework for better understanding the functional organization of the proteome. However, difficulties may arise in the process of systematic analysis due to the nature of network complexity. Since the cellular network, like the genome, was developed through evolution, the protein phylogenetic information could be utilized as a powerful tool in dissecting a protein interaction network. Here, we adopted a phylogenic grouping method combined with force-directed graph drawing in the topology space, successfully decomposing the human protein-protein interaction network in a multi-dimensional manner. First, we found that ancient proteins tend to occupy the core of the network, forming the main hierarchy of the structure, whereas young proteins reside at the periphery. Second, topological analysis also suggested a positive correlation between protein ages and a variety of network centrality measures. A great majority of network hubs were played by aged proteins, which in turn largely contribute to the scale-free and small-world properties of the network. Third, the temporal pattern of network interaction density implied a selection pressure acting on the duplication and divergence process during network evolution, in which proteins with higher centrality were selected to avoid perturbation. Lastly, functional analysis revealed that each phylogenic group possessed high specificity of enriched biological processes and pathway engagements, which could correspond to their evolutionary roles inside a cell. More interestingly, the network landscape closely coincided with the cellular localization of proteins. Together these findings suggest the potential of the conceptual framework to mimic the real functional organization in a living cell.