Computation of elementary modes: a unifying framework and the new binary approach
A number of tools for analyzing constraint-based metabolic models are available to the scientific community [1, 2, 3]. However, these tools have important limitations: some of them, based on floating-point arithmetic, may yield incorrect qualitative predictions, while others are too computationally demanding for large genome-scale metabolic models, and few are able to answer all the questions a researcher may ask. We address these limitations in the following way. First, we restate important questions in more tractable terms using new structural insights. Second, we incorporate procedures in exact arithmetic to prevent numerical inaccuracies. Finally, we propose an integrated pipeline for performing a complete qualitative analysis of constraint-based models. This analysis is facilitated by the MONGOOSE (MetabOlic Network GrOwth Optimization Solved Exactly) package we developed in-house.