[gmx-users] Why density increase with increasing the cutoff length?

[Mark Abraham]

Arden Perkins wrote:
>  From what my Professor told me it is my understanding that cutoff 
> length is somewhat a trade-off between accuracy of the simulation and 
> length of time to generate the simulation. A higher cut-off indicates 
> more accuracy but will take longer to simulate. I use low cut-offs for 
> less important simulations like energy minimizations.

A higher cut-off does not necessarily indicate higher accuracy, for the 
parameterization process used a particular cut-off. The model physics is 
defined by all of the functional form, parameters, cut-offs, etc. The 
validity of the parameters is intrinsically linked to that cut-off. One 
might be able to demonstrate that one can get equivalently valid results 
with a different (i.e. longer) cut-off, but then there's not yet a 
demonstrated *increase* in accuracy. If the same parameters can produce 
a better model physics at a longer cut-off, then there's probably a case 
for further parameterization to do equivalently well for lower cost.

All this assumes a non-Ewald method. PME is a different matter entirely.

> An increase in density would mean a larger number of simulated molecules 
> and therefore a need for a higher cut-off for more accurate data. That 
> is my best theory anyway.

A higher density for a given cut-off increases the number of interaction 
partners for each atom, but that implies nothing about the accuracy of 
the model of that system at that density. A move from one density to 
another during equilibration at a given cut-off tends to indicate the 
unsuitability of the model physics at the former density.

Xavier Periole:
 > But I think we all agree on these issues: the treatment of long-range 
 > interactions are delicate :))

Agreed, treating long-range interactions is delicate :-)

Mark

> On Sun, Dec 20, 2009 at 11:03 PM, Yanmei Song <ysong30 at asu.edu 
> <mailto:ysong30 at asu.edu>> wrote:
> 
>     Dear Users:
> 
>     Anyone can explain why the density of the water models increase with
>     increase the cutoff length. I tried a couple water models in
>     reaction-field, PME simulations.The cutoff length ranged from 0.9 to
>     1.5. They all show the same trend. Then there must be some reasons.
>     Anyone can tell me why?
> 
>     -- 
>     Yanmei Song
>     Ph.D. Candidate
>     Department of Chemical Engineering
>     Arizona State University
> 
>     --
>     gmx-users mailing list    gmx-users at gromacs.org
>     <mailto:gmx-users at gromacs.org>
>     http://lists.gromacs.org/mailman/listinfo/gmx-users
>     Please search the archive at http://www.gromacs.org/search before
>     posting!
>     Please don't post (un)subscribe requests to the list. Use the
>     www interface or send it to gmx-users-request at gromacs.org
>     <mailto:gmx-users-request at gromacs.org>.
>     Can't post? Read http://www.gromacs.org/mailing_lists/users.php
> 
>