Abstract
Biomolecular simulations are increasingly being used to generate detailed structural models to aid interpretation of neutron reflectometry (NR) data obtained from model biological membranes. Unlike globular systems, often studied by small-angle scattering, simulations of two-dimensional layers are sensitive to the simulation cell used which constrains the system laterally. We perform a careful analysis of NR data obtained from a monolayer of the lipid distearoylphosphatidylcholine at the air-water interface and show that the fit of number density profiles obtained from atomistic molecular dynamics simulation to the experimental data is very sensitive to the assumed area per lipid (APL). We propose a protocol for obtaining a realistic isotherm by combining the experimental surface pressure corresponding to a reflectometry measurement with an APL obtained from the simulation that best fits that data. Finally, we demonstrate how downstream interpretation of the experimental sample, derived from structural and dynamic properties of the atomistic model, depends strongly on the correct choice of simulation cell.