Modeling And Analysis Of Gene Networks In Brain Tumor

Genome is a complex system, and forms gene networks in terms of complicated interactions to realize its functions. Studying these relationships to capture the biology rules is an important topic in post-genomic era. In this thesis, based on genes'expression profiles data, the impact of three-type genes in biological processing are discussed. Three-type genes contain housekeeping gene (HKG), tissue-specific gene (TSG) and disease-related gene (DRG) in brain. The details are as follows:Using genes' expressions profiles data of three-type genes in brain tumor at normal and gradeⅠ,Ⅱ,ⅢandⅣ, the corresponding gene mutual information networks are constructed by reverse network modeling. It is believed that the structure of a network dictates its functions. The comparison of the structures of gene networks for brain tissues with and without tumor can provide better understanding of formation and development of cancer at the molecular level. Through comparing and calculation the network structural parameters of gene networks for brain tissues with and without cancer, the significant difference of the structures of gene networks corresponding to tissues with and without cancer is found. For the networks of normal and brain tumor, there is no significant difference in the structures of HKG networks corresponding to normal and gradeⅠ,Ⅱ,ⅢandⅣ. Therefore we predict that HKG may be no relation with the formation and development of brain tumor, but TSG and DRG have closely relationships. Furthermore, for TSG and DRG, according to the network structural parameters of gene networks, a method to find structural key genes is proposed that may have significant impacts on the formation of cancer from the network structure point of view. Here, twenty key genes are given. Our literature review shows that fifteen of these genes are closely related to the formation and development of brain tumor, leaving the other five genes open. Furthermore, we predict that these five genes may play important roles in the formation of brain tumor.Traditional pairwise relationships cannot adequately illustrate the complexities that arise in cellular networks. In this thesis, we study and analysis logical relationship of HKG, TSG and DRG Using genes'expressions profiles data of three-type genes, the corresponding gene logical networks are constructed. First, the network structure parameters will be extended to logical network. Through comparing and calculation the network structural parameters of gene networks for brain tissues with and without cancer, the significant difference of the structures of gene networks corresponding to tissues with and without cancer is found. These differences inspire researchers:HKG is no relation to the formation and development of brain tumor, but TSG and DRG have closely relationship. Through contrasting of the network structures between the networks for tissue with and without cancer, obvious difference in the distribution of logic motifs of 2-order of networks are found in the logical relationships. We speculate the cause leading to difference may be the differences of the variation of reaction modes, genes effects and the abnormal of few genes.The obtained results applying the above two methods researching three-type genes in brain tumor can provide beneficial enlightenment for diagnosis and treatment of brain tumor.