Re: [AMBER] PBC in umbrella sampling

From: case <>
Date: Fri, 18 Jun 2010 18:15:41 -0400

On Fri, Jun 18, 2010, dhacademic wrote:

> I have a question about PBC in umbrella sampling.
> The distance between active site and ligand is defined as the reaction
> coordinate. PBC with truncated octahedron as well as iwrap=1 are
> used. During the simulations of some windows in Umbrella sampling, the
> protein and its ligand will drift over time out of the primary unit
> cell. It looks to stay on one side of the cell box while the ligand
> appears on the other side of the box. In principle, this should be OK as
> the implement of PBC.
> However, the calculated coordinate is much longer than the restraint one
> (as shown on the output attached below). It seems that the calculated
> coordinate is the distance between protein and one image structure of
> ligand. What is wrong? Thanks in advance.
> 138000 7.070
> 138500 7.172
> 139000 61.263
> 139500 7.239
> 140000 7.118
> 140500 7.126

I think your analysis is fine. You can't use iwrap=1 and this sort of
restraint, at least in the code as it stands.

[Aside to developers: can someone step up to the plate and see if there
is any reason not to allow unformatted restart files? Basically, someone
just needs to test it out in a variety of circumstatnces, and study the code
to see if anything looks broken. As people now routinely run long
simulations, we need some option other than just telling everyone to set


AMBER mailing list
Received on Fri Jun 18 2010 - 15:30:03 PDT
Custom Search