I have a system on N peptides. They are tightly packed into the unit cell, so peptide A thinks peptide Z is its neighbour, due to PBCs. A bit like lipids in a bilayer.
I would like to take the distance over time between two residues, one on peptide A and the other on peptide Z. Unfortunately, gromacs is returning the distance within the unit cell e.g., 10 nm, rather than the shortest distance over the PBC , 0.3 nm.
I can do this in VMD as I simply layout periodic images, however, it won't calculate a centre-of-mass distance between two groups of atoms. I keep going between pairdist, mindist and distance thinking that I've misread the help, however, I've yet to find an answer.
Hi Anthony,
have you tried playing with the gmx distance with options -pbc no? It should be toggled yes authomatically and for me it gives the shortest distance (direct or to a periodic image) when used on lipid bilayers.
If you were using gmx distance just take care of how it reads positions: within one group of the index, containing say atom0,...,atom9, it computes distances between atom0 and atom1, then atom2 and atom3 and so on. With the -selrpos res_com it can compute the distance between the COM of the residues the atoms belong to, but still uses this pairwise logic.
Hope this is useful.
Irene
This is a good question for the gmx-users mailing list. It'll be easier to get useful help if you report the command line that you used. But I assume you didn't give a .tpr file with the -s option, so the tool might not understand enough to handle the -pbc option the way you'd like.
Hi Mark, I've tried several GMX commands with no success, so I decided not to uploaded any and I'm now convinced it could be something that I can't do in Gromacs and may require a custom distance script. I've attached a drawn image to aid my original description.
The picture shows the unit cell with an illustration of the period image either side along the x-dimension. Inside the unit cell are a few fibril like proteins, not connected (this is actually collagen, the crystal structure and crystallographic dimensions yield a dense packed and staggered environment). I want to measure the distance between two residues as indicated. However, so far using Gromacs I only seem to be able to measure with distance within the unit cell, which is actually longer than if I was to measure the distance passing through the unit cell boundary into the nearest image. I hope that makes sense?
Thanks!
- Badges
- Science topic