AMBER Archive (2007)

Subject: Re: AMBER: How to modify the velocity of a water molecule?

From: Thomas Cheatham (tec3_at_utah.edu)
Date: Thu Feb 15 2007 - 21:47:24 CST


> I am doing targeted md. My protein is solvated by a shell of water. The
> whole has been equilibrated. When the TMD runs for a period of time,
> sometimes 300 ps, or 500ps, there is a water molecule escapes from the
> surface. It is headachy.

Yes, but it is probably real.

> Now, I want to modify the velocity of the escaping water, and let it run
> to the opposite direction. I think it is physical. Correct? If yes, how to
> modify it form the *.rst file?

I think it is unphysical to reverse a water molecules direction. Maybe,
if you imagined that the water droplet was imbedded in a concentrated
water vapor, you would see waters come back to the droplet, but in vacuo,
this water should evaporate and move away...

> Is there someone know the realy meaning of the RST file?

The formats of the AMBER files are on the Amber WWW page at
http://amber.scripps.edu

   http://amber.scripps.edu/formats.html

Water shell models are full of complications; there are a number of
reviews and primary literature that discuss the issues and I do not think
the issues are resolved.

The simplest way to handle your current situation-- that I can think of--
is to run the system in a large (constant volume) periodic box that has
your solvated protein plus water vapor. The periodicity will effectively
bring waters that evaporate back into the box. However, you have to be
careful since the standard water models were developed for the liquid
phase. They may not properly represent the water/vapor interface. The
other alternative is simply to set the velocity of the evaporated waters
to zero (rather than reverse it). But in any event, you are treading into
new research areas and would be wise to consult the literature to see what
other's have done...

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu