VMD-L Mailing List

From: Celia Bremer (fuwsle_at_hotmail.com)
Date: Wed Nov 14 2012 - 10:23:18 CST

Hello,

I want to use VMD to visualize results of some simulations that I did in LAMMPS. I can get my data into VMD with Topotools. The problem is, my simulations use periodic boundary conditions, and I have a lot of bonds that are between atoms that lie on opposite sides of the periodic boundaries. What I see as a result is, I get bonds that go all the way across the box. It looks like the plaid pattern in 3D. That makes sense, but it's not what I'm looking for. (Unwrapping doesn't make sense for my data.) I've seen many questions about periodicity on this list, but I don't think this one has already been covered.

1) Is it possible to select these bonds in VMD - say, select bonds that are longer than a certain length - and either hide or delete them? I looked through the GUI selection options, but they were all about atoms, not bonds.

2) This one is probably a "no", but I wanted to make sure. What would be even better is if I could see those bonds, but they went to atoms that are wrapped to the other side of the box. (This would make the most sense if visualizing only bonds, not atoms.) For example, let's say I have a bond between two atoms, one near the +X face of the box, the other near the -X face. The atom near the -X face of the box is a little bit closer to the box boundary than the +X atom. What I'd like to see is, the -X atom gets *replicated* to its image near the +X face, and I see the bond between these two atoms. (This way, there are two copies of this particular atom.) One consequence of this is, if I replicate data, e.g. with topotools::replicatemol, it would look like a continuous network, as long as all bonds are the same color.

3) One last item, also probably a "no". I tried replicating my data with topotools::replicatemol, and there was no bonding across the faces of the different replicas. If I made 3 replicas, I got 3 independent plaid boxes. Is this correct, or is there an option I missed to account for that?

I figured I'll have to process my data to some extent before getting it into VMD. I'm just trying to make sure I'm not reinventing the wheel.

Thank you!
Celia