A Network Perspective To Natural Selection

Many factors contribute to species formation and biological evolution, including different environment and genome sequences between individuals. A variety of reasons cause gene’s allele frequency rise or fall, which affects relevant protein’s expression level and function.After layers of amplification, it may affect individual’s adaptation to environment, that is, produce the so-called "positively selected genes". During long and repeated interactions between organism and environment, the difference will ultimately have effects on the species’survival mode. In the recent studies, positive selection along the human lineage has attracted strong interests of scientists and many levels of datasets have been generated. At the same time, the rapid development of proteomics and network provide us the powerful tools and favorable conditions to learn the physical characteristics of biological molecules at the protein level, give researchers strong support for the reliability of the corresponding results. In this paper, combining population genetics with proteomics together, we collected positive selected genes and protein interaction data as the given information to survey the property of selected genes in the protein-protein interaction network. Using many network relevant coefficient (e.g, the clustering coefficient, betweenness, network diameter as well as the shortest path), we found the positively selected genes are more likely to gather together and form a compact cluster. Compared with network background genes, selected genes’average degree is larger. When we divide the selected genes into long term selected genes and recently selected genes.We found that long term selected genes scatter in the network periphery, while recently selected genes cluster in the network center regions. In order to further understand this phenomenon, we explored a few features of these genes and found that recently selected gene appear more early in time (show in more species) and a proportion of recently selected genes are driven by fine-tuning expressional differences between population (as supposed by coding changes). With several sources of data for selected genes and different sizes of protein networks, we get consistent results. Our study integrated information from natural selection together with protein-protein interaction network and explored several interesting characteristics of positively selected gene in the light of the protein network。...