I would imagine that you would need forces along the alchemical parameter,
since SMD requires you to know the work being done by the biasing potential
(along any coordinate, even an alchemical one). I don't think you can get
around this via scripting; you'll actually need to modify the code. You'll
probably need the gradients along this coordinate, and as Ross pointed out,
the TI facility is based on a static lambda value. His advice about working
out the equations before you jump into it is probably also well-advised.
Good luck!
Jason
On Wed, Dec 1, 2010 at 4:38 PM, Ross Walker <ross.rosswalker.co.uk> wrote:
> Hi Jerome,
>
> The value of lambda is fixed during the TI simulation so there is no way to
> modify it during a simulation without getting your hands dirty. One 'hack'
> would be to run a few hundred steps and stop writing a restart file and
> then
> start again with a new lambda value, wash, rinse, repeat. This could be
> easily scripted in a shell script although the performance will not be
> great
> as you will have to pay the startup penalty on each restart.
>
> A cleaner way would be to modify the source code to change the value of
> lambda. I haven't looked into it but this should not be too hard to change,
> just trace it through from where it is first read from the namelist to find
> where one would change it. In principal this would work. The question that
> may need thinking about is whether it is valid to do this. Is simply
> changing lambda ok or is there some gradient with respect to lambda that
> you
> would need to take into account. This will need some thinking about while
> looking at the equations so you might want to consider that first.
>
> In summary you will need to modify the code but it shouldn't be too
> difficult assuming there are not some strange derivatives that have to be
> taken into account.
>
> All the best
> Ross
>
> > -----Original Message-----
> > From: Nilmeier, Jerome P. [mailto:nilmeier1.llnl.gov]
> > Sent: Wednesday, December 01, 2010 1:20 PM
> > To: AMBER Mailing List
> > Subject: Re: [AMBER] (no subject)
> >
> > Hello:
> >
> > I am interested in applications very similar to what a steered md run
> might
> > provide, but with alchemical parameters as the time varying function. I
> am
> > just now learning the details of Amber, and I thought that maybe there
> > would be a way to generate these types of trajectories using existing
> > machinery. Perhaps we can talk more about it when you have time?
> >
> > Best, J
> >
> > -----Original Message-----
> > From: Adrian Roitberg [mailto:roitberg.qtp.ufl.edu]
> > Sent: Wednesday, December 01, 2010 8:10 AM
> > To: AMBER Mailing List
> > Subject: Re: [AMBER] (no subject)
> >
> > I am pretty sure there is no code like that in amber, but I would be
> > VERY interested in why you would do such a thing.
> >
> > Adrian
> >
> > On 12/1/10 5:05 PM, case wrote:
> > > On Tue, Nov 30, 2010, Nilmeier, Jerome P. wrote:
> > >
> > >> I am interested in generating trajectories whereby an alchemical
> > >> potential is continuously altered over the course of the trajectory.
> > >> (Something like steered MD, but with lambda varying rather than a
> > >> biasing potential). Is there a straightforward way to do this?
> > >
> > > I'm pretty sure there is no current way to do this, unless Volodymyr
> stuck
> > > something in the ABMD stuff that I missed....
> > >
> > > ...dac
> > >
> > >
> > > _______________________________________________
> > > AMBER mailing list
> > > AMBER.ambermd.org
> > > http://lists.ambermd.org/mailman/listinfo/amber
> > >
> >
> > --
> > Dr. Adrian E. Roitberg
> > Associate Professor
> > Quantum Theory Project, Department of Chemistry
> > University of Florida
> >
> > Senior Editor. Journal of Physical Chemistry.
> >
> > on Sabbatical in Barcelona until August 2011.
> > Email roitberg.ufl.edu
> >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Graduate Student
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Dec 01 2010 - 14:00:03 PST
