Re: [AMBER] (no subject)

From: Nilmeier, Jerome P. <>
Date: Thu, 2 Dec 2010 16:05:26 -0800

Thank you...I would have never thought to try this approach. This may be a way to get a good test system working prior to coding something up. I appreciate the help.

Best, Jerome

-----Original Message-----
From: case []
Sent: Thursday, December 02, 2010 5:31 AM
To: AMBER Mailing List
Subject: Re: [AMBER] (no subject)

On Wed, Dec 01, 2010, Nilmeier, Jerome P. wrote:
> Thank you for the advice and assistance...I'll work through some of the
> equations some more. My current thought is that I can get it to work
> using either a static lambda or a gradient, as long as I'm careful about
> the bookkeeping. Perhaps I can put together a test run by patching
> some TI runs together prior to modifying the code. I'll dig into the
> technical side a bit more, and try to come up with a good plan.

If you really want to pursue this, check out the "iscale" variable in sander.
It allows "extra" dynamical variables beyond just the coordinates. Basically,
you could set iscale=1, and arrange to have your lambda value copied to
x(3*natom+1) and its force component to f(3*natom+1). The MD step will then
update these values along with the coordinates. You can see an example in how
the alignment tensor is handled for NMR residual dipolar coupling restraints.

...good luck...dac

AMBER mailing list

AMBER mailing list
Received on Thu Dec 02 2010 - 16:30:03 PST
Custom Search