embed membrane proteins in the vesicles

[McHaillet]

Membrane proteins can easily be placed in the vesicle simulation. They just need to be oriented with the membrane on the xy-plane of their coordinate system. The user should also provide an input for the z-height of the membrane relative to the macromolecules coordinates.

Biggest issues for implementation:

• memory concerns for sampling of atoms to electrostatic potential for GPUs (and time for CPU)
• ground truth annotations of the particles: class mask needs to be generated per vesicle and classes need to remain unique for each simulation