The study of modularity in complex biological network is important for understanding the structure of complex networks and the mechanism of biological systems. It is one of the essential aspects of studies in systems biology. In this thesis, I studied systematically this problem in the following two ways: the computational methods of network modularity and the analysis of modularity in specific biological networks. The main contributions of this thesis are as follows: 1. Computational method for extracting hierarchical and overlapping modularity of biological networks: Biological networks often have both hierarchical and overlapping modularity organizations. These aspects are the intrinsic properties of biological networks. Thus a computational method of modularity is said to be effective only when it can detect both structures. The results like this can reflect the fact of biological systems and be more practical and predictive. We proposed a new framework to solve the problems based on consensus clustering and non-negative factorization. The method extracted the correct level of hierarchy and reflected the overlapping between modules. 2. Systematical analysis of modularity in the disease phenotypic network: Similar disease phenotypes are begotten by the modular nature of gene network, thus we suppose that all human disease phenotypes also appear in a modular style. We gave a network representation of phenotypes using the dataset of phenotypic similarity. We investigated the modularity of the disease phenotype network. We computationally extracted significant phenotype modules and found that the modularity is correlated well with the physiological classification of human diseases. We also found relationships between the modularity and function genomics as well as drug discovery 3. The modularity in protein-protein interaction networks and the biological implications: Firstly we investigated the interactome of yeast. We used quantities introduced by the modularity to describe network properties of genes and to investigate the correlations between duplicate genes and genetic robustness in yeast. Secondly, we confirmed the existence of the structural modularity in human protein-protein interaction network and evaluated the biological significance of the modularity from the similarity of gene’s functions and that of human disease phenotypes. At last, we defined the roles of proteins based on the modularity and investigated the relationship between the rol...
修改评论