Max Delbrück Center for Molecular Medicine,
1Department of Bioinformatics
Robert-Rössle-Str 10,
13092 Berlin-Buch, Germany
Email: zevedei@mdc-berlin.de stschust@mdc-berlin.de
Pathway analysis has become an important tool in biochemical modelling [1, 2]. It is helpful in functional genomics [3, 4] and biotechnology [2, 5]. A central concept in this analysis is that of elementary flux modes [2, 6]. In metabolic networks, there are often enzymes having more than one function (e.g. transketolase, transaldolase, aldolase, nucleoside diphosphokinase, uridine kinase) [7]. These multifunctional enzymes operate according to various mechanisms. Here, we analyse the effect of multifunctional enzymes on the number of elementary flux modes (pathways). We also study the benefits of two different approaches to describing multifunctional enzymes. The usual description is in terms of (overall) enzymatic reactions. At a more detailed level, the reaction steps (half-reactions, hemi-reactions) [8] of the formation and conversion of enzyme-substrate complexes are considered. Several software tools for understanding the interrelations between these two descriptions and metabolic pathway analysis were developed. We show that the results of pathway analysis are independent of the level of description if all reactions are irreversible while they may not if some reactions are reversible. In this situation, the latter description provides a better characteristic of the metabolic system, without complicating it. The analysis is illustrated by several biochemical examples taken from human sugar metabolism and nucleotide metabolism.