Young-Gyun Oh

Multiobjective flux balancing using the NISE method for metabolic network analysis.

Flux balance analysis (FBA) is well acknowledged as an analysis tool of metabolic networks in the framework of metabolic engineering. However, FBA has a limitation for solving a multiobjective optimization problem which considers multiple conflicting objectives. In this study, we propose a novel multiobjective flux balance analysis method, which adapts the noninferior set estimation (NISE) method (Solanki et al., 1993) for multiobjective linear programming (MOLP) problems. NISE method can generate an approximation of the Pareto curve for conflicting objectives without redundant iterations of single objective optimization. Furthermore, the flux distributions at each Pareto optimal solution can be obtained for understanding the internal flux changes in the metabolic network. The functionality of this approach is shown by applying it to a genome‐scale in silico model of E. coli. Multiple objectives for the poly(3‐hydroxybutyrate) [P(3HB)] production are considered simultaneously, and relationships among them are identified. The Pareto curve for maximizing succinic acid production vs. maximizing biomass production is used for the in silico analysis of various combinatorial knockout strains. This proposed method accelerates the strain improvement in the metabolic engineering by reducing computation time of obtaining the Pareto curve and analysis time of flux distribution at each Pareto optimal solution.

Determination of the Metabolic Networks Fluxes Using Carbon Isotopomer Labeling and Metabolic Flux Analysis

To determine intracellular fluxes using carbon labeled experimental data, adequate techniques are needed. Metabolic flux analysis (MFA) is a useful method to simulate metabolic networks. Because the measurable extracellular flux data is always insufficient, there are more unknowns than equations. Further information or constraints are required to make fully determined system of which the degrees of freedom (DOF) is zero. It is possible to obtain mass distribution data using the carbon isotope labeling experiment for fermentation experiments. Carbon isotope labeled data can be obtained from C isotope tracer technique and GC-MS (Gas Chromatography-Mass Spectrometry) measurements. In this work, we have developed a metabolic networks simulation tool which can exactly determine intracellular fluxes using carbon isotopomer labeling data. The result can provide strict insight into complex metabolic networks

Multi-product trade-off analysis of E. coli by multiobjective flux balance analysis

In this study, we developed a novel multiobjective linear programming (MOLP) strategy based on the noninferior set estimation (NISE) method (Solanki et al., 1993), whereby Pareto solutions for the given set of conflicting objectives and corresponding flux distribution profiles are generated to understand how the internal fluxes are changed in the metabolic system. Furthermore, this MOLP approach was integrated as a new module into the program package, MetaFluxNet, which was developed for metabolic pathway construction and analysis (Lee et al., 2003). As a result, this package enables users to implement the multi-product trade-off analysis as well as the single product optimization. The efficacy and efficiency of the approach were demonstrated by applying it to the in silico E. coli model. Consequently, multiple objectives such as the maximization of succinic acid production and the maximization of NADP were considered simultaneously. The result can provide new insight into the relationship among the measurements, the objective criteria and the possible solutions.