The Reconstruction, Analysis And Decomposition Of The Transcriptional Regulatory Network Of Bacillus Subtilis

Various cellular components interact with each other forming complex networks to execute specific biological functions. Structural and functional analyses of these networks are key issues in systems biology. The increasingly available data on cellular interactions generated from high-throughput technologies builds solid base for the reconstruction and analysis of biological networks at genome level. Transcriptional regulatory network (TRN) is a key network in cell regulation, especially in Prokaryote organisms. Bacillus subtilis is the best-characterized Gram-positive bacteria. It represents a paradigm of gene regulation in bacteria due to its complex life style. However, the analysis in the TRN of B. subtilis is just starting up and far from complete, there is a clear need to get biological insights from topology analysis of the network.We first reconstructed TRN of B. subtilis by integrating information from two different databases: DBTBS and Prodoric. Considering the different function of sigma factors and transcription factor, we analyzed the network in two different ways: one with sigma factors and the other without sigma factors. For the network with sigma factors, we calculated the out degree and in degree distributions and found that the TRN of B. subtilis is a scale-free network, which is the same as other biological networks. By comparing it with the regularoy network of Saccharomyces cerevisiae, we concluded that they have different structure features. Then we developed two decomposition methods for the network, after discussed the characteristics of each, we determine the solution for the strongly connected components and the shared genes between modules. For the network without sigma factors, we analyzed the weakly and strongly connected components and the path lengthes of the network. Then a distance-based decomposition method is presented focusing on the giant weakly connected component. Applying this method to the TRN of B. subtilis without sigma factors, 32 modules are identified, and the biological functions of these modules have been clearly defined. This result indicates that the new decomposition method is useful in identifying biologically meaningful modules in the TRN of B. subtilis.