Ideally, the molecule is centered initially and then the equilbration steps are done. Say, I've created a combined .gro file for my molecule in water and during equilbration run ,I notice that my molecule is diffusing out of the box. Should I try and take the .trr file and center the molecule before production run or I may run it first and then apply pbc and centering option through trjconv command before analysis?
During a typical long simulation, your molecule will cross the box boundary hundreds of times. However, since the box is periodic, there is no physical effect associated with these crossings; they are perfectly fine, unless you e.g. try to manually change the geometry of the box. Typically, you will only want to recenter/make whole/fit your solute for the purpose of visualization and analyses, and - as you suggest yourself - this can easily be done as post-processing.
Thank you for clarifying.I centered the molecule using trjconv option and added -pbc mol ,-ur compact, -center flags. But my water molecules are still diffused around the edges of the box. Will that be a problem?What do I do to correct this?
What exactly do you mean by 'diffusing around the edges'? Nevertheless, as I said, this should never be a problem if you just extend your run, or even start your simulation anew after energy minimization.
Whenever in doubt, you can use e.g. VMD (Graphics -> Representations -> Periodic) to show the periodic images explicitly; then you should see the system as continuous (use the Dynamic Bonds representation to avoid visualization artifacts).
- Badges
- Science topic