Good Morrow! I wanted to ask if anyone had knowledge of how to apply repeating boundary conditions in the x/y dimensions but not in the z?
I am simulating an ion channel in a lipid membrane and I need to ensure that ions that flow through the channel do not “loop around” back to the interior/exterior of the membrane. I am fine with repeating boundary conditions in the x-y dimensions, but I would rather have a hard barrier in the z-dimensions since my simulations do not care about edge effects far away from the channel.
If anyone has any advice on how to do this for energy minimization/NVT/NPT steps, it would be very appreciated. Thanks!
I think that you could use flat-bottomed position restraints.
But such a setup is likely problematic. You would build up an electric field if only of one kind pass through and are then not allowed to loop around. This will create a counter-force.
I am trying to show the movement of charge through an ion channel and the conformational changes of the ion channel based on charge presence above and below the channel, so electric field build up is perfectly fine.
Just to make sure, you are aware of the existence of the computational electrophysiology module? That’s one way in which you can set up a stable charge imbalance without edge effects.
I was unaware of such a feature. I have previously attempted manually mirroring my simulation environment, but had run into issues with its execution.
The process looks incredible, though I am weary of using it since I would not like to rebuild my topology. If you have familiarity with the software, do you think you could answer a few questions I had from looking through it?
Is this a separate run that I would have to do for electrophysiology or could I run it with energy minimization?
If it is compatible with energy minimization, what would the commands that I add look like?
There’s no need for rebuilding your topology, you just duplicate the system, which on the topology side means duplicating the lines in [ molecules ], and on the structure side doubling the size along Z-axis, copying the structure, translating it by the length of the Z-axis (e.g. with gmx editconf), and merging them into one.
I don’t think it can affect the minimization though, the whole point is about swapping the ions in a dynamical way to keep the imbalance (that’s initially generated in an automatic manner). There are a few detailed tutorials/discussions dedicated to that if you google the name of the method.
That is what I did before. Unfortunately, I was unaware GROMACS had a protocol for this so I did it manually. Regardless, that is not a method I would like to use since it produced all kinds of problems when I did it before.