Re: [AMBER] problem with "closest" waters being wrapped

From: Thomas Cheatham <>
Date: Thu, 3 Dec 2015 15:53:25 -0700 (Mountain Standard Time)

> You should try using the "noimage" keyword with the closest command; I think that should do what you want.

No, this is a more complex problem. The closest water IS actually in the
periodic image and this has to do with the fact that the molecule is
likely not fully centered or there is not enough water in the box to have
all the close waters NOT in periodic images.

If noimage is done, it will do what is wanted in terms of visualization
(only put waters close to the solute close to the solute), however these
will *not* necessarily be the true closest waters.

The only good solution I see is to keep the box information (which is done
by default) and do not turn on noimage in subsequent analysis. Note that,
if you are using a othorhombic aka rectangular box (not truncated oct)
then familiar (which images to put things closest to the solute) is not on
by default in cpptraj, so force it and see if this gives the better

autoimage familiar

If you still have periodic waters, then likely the level of solvation is
too small to keep all "near" waters in the same unit cell (i.e. make the
box bigger).


AMBER mailing list
Received on Thu Dec 03 2015 - 15:00:06 PST
Custom Search