[gmx-users] Re: Molecule Jumps

[Mark Abraham]

darrellk at ece.ubc.ca wrote:
> Hi All,
> I just wanted to let you know how I resolved my problem, in case anyone
> else experiences a similar problem in the future.
> 
> My structure spanned all four quadrants around the origin (i.e. the 4
> vertices of my structure were (-x, -y), (-x,y), (x,-y), & (x,y)). I
> think because it spanned all 4 quandrants, this caused the molecule to
> jump from corner to corner within the box.
> 
> Once I translated my molecule such that its coordinates were entirely
> within the positive x,y quandrant, the molecule no longer jumped around
> in the box.

...only because the critical feature of the molecule (COM? first atom?) 
didn't drift outside the box in which you placed it. I don't know (or 
care) what the conventions of GROMACS or your visualization program are, 
but the general solution is to forget about it until you go to visualize 
your trajectory, and then fiddle with trjconv until you're happy with 
the result. :-)

Mark