On dynamically identifying elementary flux modes for a poly-pathway model of metabolic reaction networks
(English)Manuscript (preprint) (Other academic)
he aim with poly-pathway models is to model variations in the metabolic behavior of cells in response to changes in their external environment.By considering the elementary flux modes of a metabolic network, the network can be reduced to a set of macroscopic reactions. The macroscopic reactions connect external substrates to products, where each reaction is associated with a kinetic equation.Since enumerating all elementary flux modes is prohibitive for complex networks, these types of models are usually limited to simple networks. In this work we consider an algorithm for identifying elementary flux modes for a poly-pathway model. First we consider a dynamic identification of elementary flux modes and model parameters using column generation. However, due to non-linearity in one optimization problem involved in column generation, elementary flux mode identification can not be guaranteed with that column generation approach. In order to still be able to identify elementary flux modes, an approximation algorithm is derived and tested for the model identification. In a case study, the algorithm is shown to work well in practice and obtains a near-optimal solution compared to a method in which all elementary flux modes are enumerated beforehand.
Bioinformatics and Systems Biology Computational Mathematics
Research subject Applied and Computational Mathematics
IdentifiersURN: urn:nbn:se:kth:diva-172376OAI: oai:DiVA.org:kth-172376DiVA: diva2:847419
FunderSwedish Research CouncilVINNOVA
QS 20152015-08-202015-08-202015-08-27Bibliographically approved