I’ve completed a 250 ns molecular dynamics simulation. Even after applying re-centering and ‘nojump’ corrections during trajectory processing, I still have 6 atoms outside the simulation box. Does this mean I need to enlarge the simulation box and rerun the entire 250 ns simulation? Alternatively, can I resize the box post-simulation and proceed with analysis (e.g., RMSD), assuming the dynamics are unaffected? For context, RMSD and other analyses appear normal even with these 6 atoms out.
There is no outside the box in MD simulations as PBCs wrap the full 3D space. What do you mean with 6 atoms out?
Thank you so much for getting back to me .. so when i run
gmx check -c md_0_250.tpr -f md_0_250_centerNoJump.xtc
It ends up with a notes that some atoms are outside the box. it happens as well when i try to check only the tpr file gmx check -c md_0_250.tpr
I thought if I did re centering and some atoms are still outside the box it means the box is small for my system and I need to resize it.
Sorry for the late reply.
In all honesty, I have never seen a gmx check report. Anyway, as the output says, These may occur often and are normally not a problem. I would go ahead and not worry about them, as there is no in or out of a MD box, this should just change where the original (not PBC) copy of the atoms appears.
On the other hand, atoms crossing PBCs boundaries can generate artifact in analysis like RMSD. Have you tried gmx trjconv to clean up the trajectory from PBCs and recenter only the group for RMSD calculation? Another couple of tools to check if the box is large enough should be gmx mindist, specifically with the -pi flag and visualizing it with software like VMD along with the box.
Thank you so much for your detailed reply. I am sorry for my late reply. Yes I did gmx trajconv
I will try gmx mindist to figure out about the box size.