With the new C36 lipid force field, simulations can be run in the NPT ensemble. However, it is still a bad idea to use lattice types which are not symmetric in the xy plane. With ORTHorhombic, there is still the real possibility of a collapse along either the x or y axis, which is independent of the force field used. The prisms, either TETRagonal or HEXAgonal, remain the best choices; most of our work has been with the TETRagonal lattice.