Cellular functions are shaped by reaction networks whose dynamics are determined by the concentrations of components. However, it is unclear to which extent the structure of the underlying biochemical networks constrains concentrations, since calculation of concentration bounds from large-scale models is not currently possible. I will present a set of approaches which allow us to predict concentration ranges in networks endowed with particular types of kinetics given limited information about well-defined parameters. I will illustrate that these approaches provides insights about how diverse organisms, and plants in particular, can control their concentration profiles and thus affect growth. In addition, I will provide our insights for integration of metabolic concentration in the constraint-based modelling framework for prediction of complex phenotypes.