Visualization and analysis of an interconnected network of genomic elements

In the context of studying the relationships among bacterial replicons, i.e., chromosomes and plasmids), we investigate various methodological approaches using a test dataset. Standard method fail to describe the complexity of the genetic events that occur in the evolution and adaptation of these elements. Given a set of genes of interest linked functionally and used as variables, the organization of beta-proteobacterial replicons was studied using several dimension reduction methods, graphs as well as supervised classification. Combinations of methods prove indispensable to characterize the relationships between replicons, permitting to identify both global trends among replicons as well as specific features of single replicon. Furthermore, our study underlines the inherent difficulty to explore exhaustively genomic data using a single tool.