Our knowledge of the properties and dynamics of complex molecular reaction networks, for example those found in living systems, considerably lags behind the understanding of elementary chemical reactions. In part, this is because chemical reactions networks are nonlinear systems that operate under conditions far from equilibrium. Of particular interest is the role of individual reaction rates on the stability of the network output. In this research we use a rational approach combined with computational methods, to produce complex behavior (in our case oscillations) and show that small changes in molecular structure are sufficient to impart large changes in network behavior.