RE: extended system or Basis Vector

From: JC Gumbart (
Date: Tue Jun 20 2006 - 01:05:54 CDT

It depends on the ensemble you wish to simulate in.

You have constant pressure enabled now, but it seems like you would prefer
constant volume. If you comment out all the pressure commands in your file,
you will run with constant volume.

In general, it's best to use the extendedSystem file when continuing the
same simulation. By enforcing a certain box size upon restart that is not
realistic, you can have very unnatural steric clashes between your atoms or,
if too big a box, a very unnatural vacuum present.

-----Original Message-----
From: [] On Behalf
Of Ching Wong
Sent: Monday, June 19, 2006 8:29 PM
Subject: namd-l: extended system or Basis Vector

Hi all,

    When running MD, should we use extendedSystem or cellBasisVector?
Which do you use after getting xsc file from the first step?
Do you always check the size of the system and reset when going on to
the next steps?

Please give me some clues (thanks a lot! )

I did
1. waterMD, (heat from 0K to 600K, then minimize)
2. then Emin,
3. then go on to HE (350K)
4. production run

Originally the box size was 76 76 80
It always get bigger after heating, so I ran water MD with the same
size and found out the system got bigger, i went back reset the size
like 80 80 82 and rerun waterMD. It created spaces (lower water
density) around the edges. I went on to step 2 (Emin) and there were
spaces there too. At the end of step 3 (HE), the water density was
pretty even.

Should we use extendedSystem instead of cellBasisVector determined by
the size of each step?



This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:42:13 CST