I am trying to rewrap the water in my system because my protein partially leaves the system boundary. The dcd trajectory files come from CHARMM version 36a6 and a truncated octahedron periodic boundary condition was used. I used the following as input for cpptraj (V4.12.1).

parm protein.psf
trajin protein.dcd
autoimage
trajout output.dcd
go
exit

Here is a snapshot from the output.dcd file so you can see the issue I am having. It appears that the truncated octahedron is offset a little bit and you are seeing corners of what might be the neighboring images.

I have tried changing the anchor selections, as well as trying origin and firstatom in the autoimage command, but each time I get the same issue to some degree. Any suggestions on how to fix this problem would be greatly appreciated.

I want the water rewrapped because I am interested in looking at the water-mediated hydrogen bonds in and around this particular protein. If it is true that the resulting dcd file is being reimaged (rewrapped) slightly off-center and that the waters immediately around the protein are correctly placed then the strange behavior seen around the edges of the rewrapped system would not lead to inaccuracy in my analysis. This is what looks to be the case, but I want to make sure that the water placements are not in general incorrect.

Thank you greatly.