I am simulating a 7TM protein using HIPPO. Cavites in the protein are filled with water molecules/ligand.
After ~80 million MC steps, the center of the protein located in the membrane region starts to collapse: polar/charged residues in the middle of the protein start to interact. This gives rise to unwinding of helices. I think the reason for the collapse is that the applied dielectric constant in the protein center is to low. I tried to avoid the problem by filling out the protein interior with water, but apparently it is not enough. It would be helpful with an option for changing the dielectric constant in the middle of the protein (by introducing a "water bubble" there).