Membrane proteins are involved in fundamental physiological mechanisms such as transport and signalling in living cells. Membrane proteins are dynamical structures whose properties vary according to mediators, such as transmembrane potential, ionic concentration and pH, used by the cell to convey information. Permeation, selectivity and gating in the ubiquitous families of potassium channels and ammonium transporters are of special interest. Potassium channels are involved in the regulation of action potentials in excitable tissues, such as the heart and brain. This function of the potassium channels is in part controlled by a mechanism known as slow inactivation, which consists in the spontaneous closing of their selectivity filter in response to various stimuli. Despite the extensive electrophysiological and structural data on potassium channels, the atomistic mechanism behind slow (C-type) inactivation remains to be elucidated. On their side, the ammonium transport proteins (Amt/Rh) are essential to nitrogen metabolism in all domains of life. The recent determination of the first X-ray structure of a member of that family, the AmtB transporter, has given some insights on the possible transport mechanism of ammonium. Though, without information on the dynamics of the protein, the detailed mechanism remains unknown. By combining explicit molecular dynamics simulations and advance statistical physics principles with functional and X-ray crystallography data provided by collaborators, we propose to elucidate the molecular basis of these mechanisms in the potassium channels and ammonium transporters. Our main approach consists in calculating multidimensional free energy profiles that govern conformational changes underlying protein activity. Complementary techniques include homology modelling, Poisson-Boltzmann electrostatic calculations and stochastic simulations. The development of a multi-scale simulation framework will be initiated to ultimately bridge atomistic simulations with macroscopic simulations of excitable tissues, which are based on empirically adjusted parameters and thus neglect microscopic structural details. Such a framework could be essential to fully understand the mechanisms that allow protein structures to react to their environment and to transfer information accordingly.