Representativeness of a set of metabolic pathways

Abstract

Pathways and more precisely Elementary Flux Modes (EFM) are artefacts extracted from metabolic networks that are very useful to achieve the comprehension of a very specific metabolic function or dysfunction. Manymethods to extract pathways have already been developed and all of them have to deal with common problems like the production of infeasible subnetworks and the production of the samesolution repetitively. Although some strategies have been incorporated to thosemethods in order to mitigate the problems, they get already a high ratio of repetitions and the insistent presence of the same reactions in the solutions.We do a proposal focused onlinearprogramming (LP)methods for pathway extraction. It aims to improve the representation of every reaction in the set of computed pathways by penalizing the most often included reactions during the extraction.