[gmx-users] RE: Avoid centering in genbox

[Tsjerk Wassenaar]

Hi Manoj,

> >My problem is unique, as I am setting up an AFM pulling in one
> direction.

Do you really think no one did that before?

> >If I could keep the solute to one side of the box, there is sufficient
> solvent
> >box length on the other side to pull into. ( because the molecule is
> long - a triglyceride)
> >Otherwise if I make it in center,  it would call for additional expense
> of computation
> >time (because of extra water molecules on one side - not required)

That's a misconception. The dimensions do not depend on the
orientation of either component with respect to the definition of the
lattice. Remember you're working with PBC.

However, there may be an issue when pulling and one of the groups
jumps over the periodic boundaries or the distance in one direction
suddenly becomes shorter than in the other direction. There was such
an issue some while ago at least; I'm not sure whether it persists in
the current version.

So, a) with PBC you have to have enough space to pull, wherever you
put your molecules. You can't get any gain whatsoever by translation
and b) to avoid jumps over the periodic boundaries you have to place
the first atom (of each group) as close to the center of the box as
possible, e.g. with editconf, using and index group with the first
atom of each molecule of interest for centering.
And while you're at it.., read Chapter 3 of the manual (the parts on
periodic boundary conditions).

Cheers,

Tsjerk


-- 
Tsjerk A. Wassenaar, Ph.D.
Junior UD (post-doc)
Biomolecular NMR, Bijvoet Center
Utrecht University
Padualaan 8
3584 CH Utrecht
The Netherlands
P: +31-30-2539931
F: +31-30-2537623