Re: Water Hole

From: Jim Phillips (
Date: Tue Sep 19 2006 - 23:47:10 CDT

You need to equilibrate periodic simulations with constant pressure to
prevent bubbles from forming (or over-compressing the system). Be sure to
use the extendedSystem option to read the modified basis vectors from the
.xsc file when restarting a run. Once you've equilibrated both pressure
and temperature you can switch to NVE if that's necessary for some reason.


On Tue, 19 Sep 2006, Jacob Poehlsgaard wrote:

> Hi NAMD users
> I'm posting in this old topic because I've also run into this problem now.
> I think I pretty much understand the concept of what Marcos is saying, but I
> don't quite get how to implement it correctly.
> I have a RNA/protein system in a water box. I ran a lot of equilibrium runs
> with constant temperature (using langvin), with gradually smaller restraints
> on the various atoms to eqilibrate them and everything went well. When I
> then wanted to do an actual production run (this is the first time I have
> ever tried this "for real"), I switched the temperature control off. I did
> this because I've read that production runs should run int the NVE ensemble,
> using no environmental controls at all. In this run, a huge water hole
> appeared rather rapidly.
> Should I have been using constant pressure control during my equilibrium
> runs and then switched if off for production runs? If so should I switch off
> both temperature and pressure controls at the same time?
> If pressure control works by rescaling the box, should I somehow alter the
> cell basis vectors to fit these new dimensions in subsequent runs with the
> output files?
> Jacob Poehlsgaad
> SDU, Denmark
>> Running each part independently won't fix anything. The holes
>> appear because your system is not at the appropriate density
>> and you are not using constant pressure. Just turn on the
>> constant pressure control (either Nose-Hoover Langevin Piston
>> or Berendsen, look at the NAMD user's guide and NAMD tutorial
>> for more details in for at least 100 ps or
>> more. The size of your box will eventually shrink and the
>> holes will disappear once the right density is reached.
>> Marcos
>> On Sun, 21 Apr 2002, Sting wrote:
>>> Thanks snoze pa! I'll try it.
>>> --Sting
>>> =================================
>>> try to run each part seperately.
>>> 1. minimization.
>>> 2. equlib.
>>> 3. free md
>>> goodluck
>>> On 4/19/02, Sting <> wrote:
>>> Dear All,
>>> I am using namd to study trihelices in the water box.
>> Minimization and
>> 500ps CA restrained dynamics were performed. I found a strange
>> phenomenon: sevral holes without water formed in the water
>> box(even afer
>> adding the water layer to 15 angstrom),and with the rotating of the
>> trihelices,the holes seem moving accordingly. Still when performing
>> SMD,the water holes remain and adjust while going on. I
>> wondered how to
>> explain it and how to tackle the problem. Could anyone
>> offer some help?
>> Thanks.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:42:36 CST