[gmx-users] Question on gmx editconf, pbc and NPT

[Sebastian Muraru]

Hi there! Please help me figure something out.

I have a setup made out of two molecules and I would like to investigate when happens when they meet (graphene + biomolecule). One of them is fixed by position restraints, one of them is mobile. When running a simulation under NPT conditions I assumed pressure is applied so that everything is 'slightly guided' towards the center of the simulation box (given a high enough pressure value). Would that assumption be correct? Or is pressure presumably applied on the faces of the simulation box?

Thus, I figured, since I want to check what happens when the molecules meet, I want molecule_1 to be in the center of the simulation box and have molecule_2 slightly guided towards it. However, given that the distance between the molecules is known and fixed, placing the box so that molecule_1 is at the center leaves a part of molecule_2 out of the box through the top face. 

I assumed this would not be a problem during a simulation due to pbc being on, thus allowing the part of molecule_2 that's out of the box to reenter through the bottom layer. However, I also need to go through a solvation step and that I think this complicates things because a part of molecule_2 will not be in solvent. Would you know a workaround for this issue? I try to avoid making the box taller / bigger.

Thanks and a Happy Upcoming New Year to the Gromacs Community!