(301b) Application of Tabu Search to Metabolic Flux Analysis Based on Labeling Balances

Metabolic Flux Analysis (MFA) is powerful tool to quantify the flux of biochemical reactions in a metabolic network. The first step in MFA is to create a model to predict the behavior of metabolic network (based on metabolite and isotope balancing), and the second step is to employ an optimization algorithm to find unknown fluxes which minimizes the difference between experimental measurements and model predictions. When labeling balances are used to create a model for a network, a nonconvex optimization problem with bilinear constraints results for which the existence of multiple local minima is a major difficulty. In this work, we propose the use of Tabu Search (TS), a stochastic optimization technique, to determine the unknown fluxes. TS has been implemented in this work to find fluxes of central metabolism of Sacchoromyces cerevisiae. Tabu Search is a parallelizable stochastic optimization algorithm which uses adaptive memory to avoid becoming trapped in local optima. Experimental data was taken from literature sources and fluxes were quantified under different conditions of glucose repression. Results show the effectiveness of the algorithm in terms of determination of local and global optima, both for serial and parallel computation.