Review of methods for catalytic reaction-pathway identification at steady state

Two major approaches are available for algorithmic identification of reaction-pathways; one is based on linear algebraic or convex analysis, and the other is rooted in graph theory. The former approach generates pathways through a series of basis transformations; however, an effective search strategy is absent in any of the methods of this approach, thus apparently rendering them inefficient in exhaustively generating the feasible pathways. In contrast, any of the methods of the latter approach requires additional tools, for example, solvers of linear equation systems or linear programming problems, to validate the stoichiometrical feasibility of combinatorially feasible reaction pathways in the light of stoichiometry. The present work reviews the available methods and highlights analogies in the aforementioned major approaches for reaction-pathway identification (RPI). Moreover, the capabilities and theoretical backgrounds of a wide range of RPI methods are summarized.