# Amber solvatebox: How do I remove water that has traveled away

I used Amber's solvatebox function with a small polypeptide and tip3p water with a distance of 15 angstroms between the molecule and the edge of the box.

After I ran dynamics, I tried to view the output by removing water from view in Jmol, but I had a really hard time rotating the polypeptide. When I viewed it with water, I found that the difficulty in rotating was because the radius of rotation (are those the right words?) was extremely large because some waters floated off in the distance.

I want to remove these stray waters. How would I go about doing that? Would I just crack open the output and remove those waters by deleting lines that contain information for those particular waters?

• That looks pretty peculiar. I've had similar things happen when trying to visualize a large system using jmol, and it usually is a problem with going from the MD to jmol and not actually a problem with the simulation itself. It usually has something to do with periodic boundary conditions... I would maybe try using solvateShell instead of solvateBox which creates a shell around the peptide of some specified thickness which will probably avoid problems with the boundaries that solvateBox might create. – jheindel Nov 21 '16 at 1:29
• If you use solvateShell, you can use a harmonic restraining potential to keep your solvent from flying off, but if I remember correctly the simulation is no longer periodic. – pentavalentcarbon Nov 21 '16 at 13:33

• Sounds like you know what you're talking about here. Can you expand on how one would use the iwrap, cpptraj, and autoimage keywords, perhaps by mocking up an input file containing them? See here for a tutorial on Markdown formatting to aid comprehensibility. Welcome to Chem.SE -- thanks for posting! – hBy2Py Nov 21 '16 at 18:52