Re: [AMBER] many initial systems for a two molecules system

From: Jason Swails <>
Date: Wed, 03 Sep 2014 11:07:32 -0400

On Wed, 2014-09-03 at 14:26 +0300, Fabian Glaser wrote:
> Hi,
> I would be grateful to get ideas or a sample script that does the
> following: I would like to create many initial systems as possible for
> two small molecules which are supposed to interact, including two
> small molecules, in the following manner:
> To start with the two molecules far away from each other (let's say
> 50 Angs). Start pulling them to each other slowly (let's say 1
> Angstrom in each 100 ps) until the final distance is 10 or 15
> Angstroms. Once this distance is reached, to pull them apart to 50
> Angstroms again, and continue doing this 50 times. In this way I would
> like to get 50 initial structures where the two molecules are in close
> contact. Then, I will perform 50 independent MD with these initial
> conformations.

You can do this using steered molecular dynamics if you wanted to (see
chapter 23 of the Amber 14 manual regarding NMR restraint -- in
particular the "jar" keyword). An example steered MD run (inspired by
the Jarzynski equation) can be found in $AMBERHOME/test/jar.

However, this seems like overkill just to get starting configurations.
Why not just do a single steered MD scan and take conformations along
the path? If you need more at a particular distance, you can run a
short MD simulation with the restraint fixed at that distance (like
umbrella sampling) and take multiple configurations that way.

Also, if you plan on using explicit solvent, a 50 A separation can be
challenging to _actually_ implement. Your box would have to be at least
100 A to maintain at least 50 A separation between all periodic images.


Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Wed Sep 03 2014 - 08:30:02 PDT
Custom Search