Response to comment by Almeida et al.: free area theories for lipid bilayers--predictive or not?

Free area theories for lateral diffusion in lipid bilayers are reviewed and discussed. It has been suggested by Almeida et al. that free area theories yield quantitative predictions for lateral diffusion coefficients of lipids. We investigate the plausibility of this suggestion by first sketching what is to be expected of a quantitative theory with predictive power, and subsequently examining whether existing free area theories comply with these expectations. Our conclusion is that current free area theories for lipid bilayers are not quantitative theories with predictive power. They involve a number of adjustable parameters, all of which are not estimated independently, but derived from fitting the theory to the very data whose behavior the theory is supposed to predict. Further, the interpretation and behavior of some of the parameters are ambiguous. The best example is the so-called activation barrier, whose qualitative behavior with the cholesterol concentration in a DMPC bilayer varies depending on the experimental method used to generate the input data and the exact assumptions made to formulate the theory. Independent determination of the activation barrier from numerical simulations or experiments appears to be very difficult.