[gmx-users] cutoffs

[Erik Lindahl]

On Friday, Mar 7, 2003, at 17:37 US/Pacific, Michael Shirts wrote:

>
> So, my understanding is, if I'm simulation a thre-point water, I put 
> rlist =
> rvdw = rcoulomb = 0.8, and nstlist = 1, then the coulombic interation 
> will
> only be calculated if the center of masses are within 0.9 nm of each 
> other.
>
> Is there some way to adjust this so that the O-O distance is the cutoff
> distance used?  I am planning on using PME for my later research, but 
> I want
> to try to make sure I can replicate some old data, and the original 
> TIP papers
> used an O-O cutoff.   I've tried it with the COM-COM distance, and the
> densities are off of the original results.
>

Not trivially, but if you only have water in your system you could of 
course just hack
the rotine calc_cgcm in src/gmxlib/pbc.c to use the first atom as the 
charge group center.
(It is actually a center of geometry and not mass, since we sometimes 
use massless particles like dummies).

Cheers,

Erik