AMBER Archive (2003)

Subject: RE: non-periodic system

From: Yong Duan (yduan_at_udel.edu)
Date: Wed May 21 2003 - 22:17:03 CDT

That does sound a little odd.
Without knowing any detail (my ignorance), I may guess one case that
such an approach could make sense.

For a protein surrounded by water, the surface tension should keep water
molecules around the protein and will impose roughly 1000-3000 atm
pressure to the protein. In these cases, by scaling the COMs of
molecules, one may be able to control the pressure. This is technically
possible because virial can always be calculated and volume can be
approximated by overlapping VDW spheres.

Physically, this does not make sense. But it practice, when one uses
such a system, the purpose is not to simulate a protein with a few water
molecules in gas-phase (because not that many people are interested in
the gas-phase properties of proteins). The purpose is still to mimic
solvated environment, yet without the cost of periodic box of water.
Afterall, one of the most convincing justification of a periodic box is
that the surface tension imposes an enormous pressure. So, if one has a
way to adjust the pressure, one can get away without a periodic box.

yong

> -----Original Message-----
> From: Michael Crowley [mailto:crowley_at_scripps.edu]
> Sent: Wednesday, May 21, 2003 7:27 PM
> To: Ioana Cozmuta
> Cc: amber_at_heimdal.compchem.ucsf.edu
> Subject: Re: non-periodic system
>
>
> Dear Ioana,
> I am curious to know what the meaning of pressure is
> in a non-periodic system. How does one model the "pressure"
> in a non-periodic system? Where is the external (constant)
> pressure coming from? What is the non-periodic NPT model in NAMD?
>
> THanks
> Mike
>
> -----------------------------------------------------------------
> Physical mail: Dr. Michael F. Crowley
> Department of Molecular Biology, TPC6
> The Scripps Research Institute
> 10550 North Torrey Pines Road
> La Jolla, California 92037
>
> Electronic mail: crowley_at_scripps.edu
> Telephone: 858/784-9290
> Fax: 858/784-8688
> -----------------------------------------------------------------
>
>
>