|M.Sc Student||Vitkin Edward|
|Subject||Functional Genomics Based Approach for Reconstruction of|
Genome Scale Metabolic Network Models
|Department||Department of Computer Science||Supervisor||Professor Tomer Shlomi|
|Full Thesis text|
Introduction: Genome-scale metabolic network reconstructions are considered as key step in quantifying the genotype-phenotype relationship. There are two major computational challenges for automatic network reconstructions: (i) the identification of missing reactions in a metabolic network (gap-filling), and (ii) the association of genes with the network reactions (gene-assignment).Flux analysis techniques are used to address the first challenge, while the second one is commonly addressed by functional genomics data based approaches.
Algorithm: Here, we present MIRAGE - a novel gap-filling approach, which identifies missing network reactions by integrating metabolic flux analysis and functional genomics data. Specifically, to reconstruct a metabolic network model for an organism of interest, MIRAGE starts from a core set of reactions, whose presence is established via strong genomic evidence, and identifies missing reactions from other species that are required to activate the latter core reactions, whose presence is further supported by functional genomics data.
Results: The performance of MIRAGE, in comparison to previous methods, is demonstrated on the reconstruction of network models for E. coli and the cyanobacteria Synechocystis sp. PCC 6803, validated via existing networks for these species. Then, it is applied to reconstruct genome-scale metabolic network models for 36 sequenced cyanobacteria (supplied via standard SBML files; www.cs.technion.ac.il/~tomersh/tools), amenable for constraint-based modeling analysis and specifically for metabolic engineering. To demonstrate the utility of the reconstructed cyanobacteria networks, a strain design method was applied to predict gene knockouts whose implementation is expected to significantly elevate the production rate of an important nutritional product, astaxanthin.
Conclusion: The improved gap-filling performance of MIRAGE in comparison to other state-of-the-art model reconstruction methods supports the advantage of integrating of functional genomic data as part of model reconstruction. We expect MIRAGE to be used for automatic reconstructions of many other species, leading to a significant boost in the understanding of their metabolism.