From: Jérôme Hénin (
Date: Fri Dec 09 2005 - 02:23:36 CST

Le Jeudi 08 Décembre 2005 22:43, Marc Q. Ma a écrit :
> if the hbonds of interest lie on the inter-image of the PBC boxes, it
> already indicates that the simulation is at risk -- proteins move
> across the boundary of the box, ....

Well first of all, using NAMD with PBC, I would not be very surprised if a
solvated protein did move across the boundary of the box during a very long
simulation, especially with the usual Langevin + PME combination. And I would
not worry too much either - what kind of risk do you see for the simulation,
except for the drift itself?

> if the objects of interests are all contained in the main image, then
> the suggested additions are not useful.
> maybe i am missing something here?

What if my system is a homogeneous solution of some small solute? Or if it is
a lipid bilayer? Or if I am interested in water-water H-bonds? In many cases,
the system of interest simply cannot be confined to the center of a waterbox.


Jérôme Hénin
Equipe de Dynamique des Assemblages Membranaires
Université Henri Poincaré / CNRS
Tel : (33) 3 83 68 43 91        Fax : (33) 3 83 68 43 71