Diversity in minimal periodic distance in different 200-ns MD simulations

User discussions
1

GROMACS version: 2022.4
GROMACS modification: Yes/No no
Here post your question
Dear GROMACS users

I have completed a protein (WT & mutant) simulation for 200 ns in triplicates using Amber99sb force filed and TIP3P water model. I have used both cubic and dodecahedron boxes with a distance of 1.0 nm from the box edges (-d 1.0) in differenet runs. The vdw cut-off is set to 1.2 nm (rvdw =1.2) and long-range electrostatic cut-off is set to 1.2 nm (rcoulomb=1.2). Now I am interested in analysing the minimal periodic distance using gmx mindist module of GROMACS to check whether the protein has seen its periodic image during the simulation.
The used command:
gmx mindist -s .tpr -f nopbc.xtc -od minimal_periodic_dist.xvg -pi
then the protein was selected.
The results for the first run and the second run with a cubic box were respectively:
The shortest periodic distance is 2.15047 (nm) at time 144990 (ps) and the shortest periodic distance is 2.1 nm at 163585 ps.
But the result for the third run with a dodecahedron box was lower than 2 nm:
The shortest periodic distance is 1.74 nm at 22800 ps.
In addition, the shortest periodic distance for the mutant protein in the first run with a cubic box was 2.09 at 113866 ps while it was 1.8 nm at 420.168 ps in the second run with the same box.
I don’t know why the minimal periodic distance has been changed to lower than 2 nm in different 200-ns MD simulations. My question is how can I interpret the results of gmx mindist according to my system?

Thanks in advance
Farial Tavakoli

2

Proteins are flexible, and the difference between a value of 2.15 nm and 1.74 nm can be one between a slightly more extended and more compact conformation of a loop. It’s not great to have the minimal distance lower than 2 nm, but it’s not terrible either (check out a great discussion of box-size effects here), and overall it’s nothing surprising if you know that proteins can be non-spherical and change shape over time.