Construction Of A New Package For Flux Balance Analysis On Genome-scale Metabolic Networks

Metabolic network has been proved to be an amazing research tool for biologicalproperties, and reconstruction of genome-scale metabolic network is playing a moreand more important role in providing systematic metabolic engineering strategies.Many e fforts ha ve be en made on t he de velopment o f methods f or metabolicnetwork a nalysis and f lux ba lance a nalysis(FBA) i s one of t he most i nformativemethods i n qua ntitative s tudy of genome-scale metabolic networks. FBA uses t heoptimization theory of mathematics to solve the biological problem, which is to findboth the optimal objective value and the solution named flux distribution in systembiology, i.e. the maximal/minimal rate of some reaction as well as t he rates of theother reactions in the model.So fa r some a pplicable t ools f or f lux ba lance a nalysis have be en p ublished,however, these methods still have some problems in the application aspect. Here weproposed a new open and user-friendly python package for flux balance analysis ofmetabolic networks. This package has the basic predictive ability such as prediction ofgrowth rate, robust analysis, and phenotype phase plane analysis. Furthermore, anoptimization based procedure is proposed to identify and eliminate network gaps inthe reconstructed networks. This package also includes methods to identify the deadends and the precursors that cannot be synthesized in a metabolic model, as well as toidentify the reactions from a reaction database that restores the connectivity of thenetwork by filling the gaps, which is curial for easy debugging of a model. Then wecompared the result of this package with COBRA using the core E.col metabolicmodel and verified its capability of prediction and further simulation. This package iswritten in Python and can be easily extended with more functions from other pythonpackages in the future for automated model reconstruction and reconciliation.