Analysis Of Arabidopsis Gene Logic Network Under Stimulus

As the whole function of a complex system depends on its structure, it is an important method to construct a mathematical model of the complex network to research the function of a complex system. From the system biology point of view, it is significant to make clear the reaction mechanism in Arabidopsis by the construction of the gene network model, especially the logic network with the higher-order logic, and the analysis of the causal relationships among genes. In this thesis,the data were collected from NCBI GEO DataSets and EBI ArrayExpress, got from either under the normal condition (control group) or under one of the four stimuli:the short day stimuli (experiment group I), the long day stimuli (experiment groupâ…¡), the bacterial stimuli (experiment group III) and the salt stimuli (experiment group IV). After having integrated the gene expression data on the same condition and the same platform, we got the abstract genes. The logic networks based on the abstract genes were reversely constructed, where there were higher-order logic relationships among genes. After having compared the structures of logic networks, the results were obtained that there were obvious differences in the distribution of 2-order logic and the property of the module structure:only the 3rd type of 2-order logic existed in the logic network of the control group, while other types except the 7th and the 8th turned up in any logic network of the experiment groups; the structure of module is more clearer in the network of the control group than in other networks. Further, the logic networks were constructed by use of the expression profile of the dominant genes defined here. Then the dynamics of these logic networks were stimulated, analyzed and discussed. The obvious differences were gained by comparing the state stability between the control group and each of the experiment groups. For example, the number of the attractors with the larger size of the attraction domain belonging to logic network of the control group was much larger than that of the ones belonging to each logic network of the experiment groups. The method of construction and analysis could also be applied to the gene profiles of other plants under stimuli, and it might be of general significance in understanding the intrinsic mechanism.