Network Structure And Function Of Yeast Biological Evolution

This thesis was focused on the relationship between network topology structure and protein function and evolution,including three chapters.Chapter 1 was an introduction to the application of complex network theory into system biology, emphasizing network static parameters and five structural characters of biology network,especially in yeast biology network.Chapter 2 was a study about the impacts of yeast metabolic network structure on enzyme function and evolution.We detected significant structural modularity in yeast metabolic network.The correlation between two centrality parameters degree and betweenness of enzymatic genes was substantially weakened due to the hierarchical structure.By a partial correlation analysis,we demonstrated that evolutionary rate of the enzymatic genes did correlate with betweenness directly and correlate with degree indirectly.We proposed betweenness as a graph centrality measure for its appropriateness in reflecting the global rather than local impact of nodes in a network with highly hierarchical modular structure such as the yeast metabolic network.We established that the enzymatic genes with higher betweenness,acting as a link between different modules,were more essential than those with lower betweenness. The enzymatic genes with higher degrees were less likely to be essential than those with lower degrees,for the protection from network compensation by alternative pathways and genetic redundancy by duplicate genes.These findings indicated that the structural features of the yeast metabolic network have significant impacts on the evolution of its enzymes.Chapter 3 was a study about the impacts of protein complexes structure on protein function and evolution in protein-protein interaction network.We detected significant correlation between network centrality parameter degree and protein essentiality in protein interaction network,named centrality-lethality rule,but the correlation was exceedingly weak.We found protein complexes occupying the centrality position in network,had larger degrees.Larger the size of protein complexes was,more essential proteins were in it.By a partial correlation analysis, we demonstrated that essentiality of the proteins did correlate with the size of complex directly and correlate with degree indirectly.We also found a significant negative correlation between evolution rate and the size of complex.We proposed that the proteins integrate into complex before joining protein interaction network,so protein complex was important link from network structure to protein function. Otherwise,we demonstrated the preferential attachments between essential proteins in PPI network.Since the subunits in complex display the same deletion phenotype (essential or non-essential),the interactions within complex almost explained the effect of preferential attachments between essential proteins.These findings indicated that the structure of protein complex have significant impacts on the function of protein in PPI network.