WARNINGS: This example was only intented to be a technical demonstration to show how to combine totally different kinds of coarse-grained molecules (with different kinds of force-fields) together in the same simulation in LAMMPS. Tuning the force-field parameters to get realistic results was not the goal. 1) If my understand is correct, this lipid model was not originally designed to be used in vesicles with such high curvature. It is not clear whether it will behave realistically when placed in such a small vesicle. The inner layer shows signs of being in the gel phase at 300K, so for the example, I raised the temperature to 345K to reduce this. I suspect other lipid models could behave differently at these temperatures and curvatures. 2) The protein model was originally globular and was heavily modified to make it stable in a lipid bilayer. Its behavior in a membrane is has not been characterized. In addition, I have noticed that newer versions of PACKMOL do not always succeed at generating a spherical vesicle in a reasonable amount of time. (You may have to play with the .inp files in the packmol_files directory to get PACKMOL to produce any files at all.)