VFFVA: dynamic load balancing enables large-scale flux variability analysis

摘要

Constraint-based reconstruction and analysis (COBRA) methods enable the study of metabolic pathways in bacterial [1] and human [2] systems, in time and space [3]. The metabolic models are usually formulated as linear systems [4] that are often under-determined [5], therefore several solutions could satisfy the subjected constraints. The set of alternate optimal solutions (AOS) describes the range of reaction rates that achieve the optimal objective such as biomass production. The AOS space is quantified using flux variability analysis (FVA) [5], which provides a range of minimum and maximum values for each variable of the system. FVA has been applied to find blocked reactions in the network [6], quantify the fitness of macrophages after the infection of Mycobacterium tuberculosis [7], resolve thermodynamically infeasible loops [8], and compute the essentiality of reactions [9].