GROMACS version: 2018.8
GROMACS modification: No
Hello everyone,
I have a polymer system that is an assembly of several single chains. I have modified the .itp file for the single chain, in this way the chain has bonds, angles, etc. across the periodic boundary only along the z axis. I am aware that I should use the periodic-molecules = yes in the mdp options.
The problem for me is that I found the polymer will rotate around the z-axis since it is free. The rotation is not what I want for my further simulation. I have tried to use comm-mode = Angular to remove the rotation. But with periodic-molecules = yes, Gromacs will give an error.
I only need to remove the rotational movement around the z-axis. and actually I think this is the only rotation that could happen in my system. And this should not violate the periodic connection along the z-axis. But it seems Gromacs does not support this. Could you please tell me how can I achieve this purpose?
Another thing is that actually I did not find observable differences with or without using periodic-molecules = yes. A simple try of mine is that I turned off the option periodic-molecules = yes in my first several steps, then comm-mode = Angular could work. And later I will use periodic-molecules = yes again. But I am not very clear about the role periodic-molecules = yes plays and how big the impact is on the infinite long chain calculation. If the function of periodic-molecules = yes could be explained a bit, it also would be much appreciated.
Thanks a lot in advance!
Hi,
A while ago I run a double strand DNA as periodic molecule at NPT using mdp options
periodic_molecules = yes
pcoupltype = semiisotropic
maybe it helps
Alessandra
Hello,
Thank you very much for the reply!
I am very sorry for the later reply. I forgot to check the messages.
I am sorry I did not quite understand your reply. I also used semiisotropic in the NPT runs, but the molecules still rotates. And it seems if periodic_molecules = yes is used, the comm-mode = Angular can not be used at the same time.