Validation of metabolic networks in Arabidopsis thaliana

The growing amount of experimental data in Systems Biology leads to the development of large complex networks. Heterogenous data of different quality from various experiments are compiled to larger and larger networks. Some of the data are true only with a given probability score, or are just erronous. The model as a whole has to be self-consistent. Bioinformatic analysis provides qualitative insights into the structure and the dynamics of the network. Several techniques have been established to validate the network structure and to identify missing or incorrect reactions. Standard approaches adopt algorithms which were developed over the last decades in Mathematics and Technical Informatics for the analysis of concurrent systems. Here, especially Petri nets have been shown as a valuable modeling methodology. The task to be solved has a high complexity. This fact confines the application to small to medium sized networks. In order to enable the analysis of large networks we developed a reduction technique applicable to metabolic networks. Reduction and analysis of a network is demonstrated for a metabolic network of Arabidopsis thaliana. Reduction techniques can also be proven to be worthwhile for the biological interpretation of the network.