Computation and Analysis of Metabolic Pathways

Presented by Dimitrije Jevremovic

Abstract

A metabolic network can be decomposed into a set of unique Elementary Flux Modes (EFMs). Each EFM represents a minimal set of enzymes that can operate at steady state with all  irreversible  reactions used  in  the  appropriate  direction. Knowledge of all elementary modes is of significant value as it provides a rigorous basis for rationally designing cells with precisely defined metabolic capabilities. Among the algorithms to enumerate all elementary modes, we use the most efficient, the so-called Nullspace Algorithm. Because the number of intermediate EFMs can be almost exponential in the combined size of input and output data, a parallel algorithm must distribute the work without needing to share all the data. We have developed parallel algorithm that exhibits good scalability while reaching sizes beyond the memory capacity, as well as incorporated a divide-and-conquer feature in it. We also propose algorithms for the elucidation of reaction deletions that may lead to efficient metabolic networks with increased metabolite production, coupled with a cellular growth.