AMBER Archive (2002)

Subject: Re: sander.LES questions

From: Carlos Simmerling (carlos.simmerling_at_sunysb.edu)
Date: Thu Sep 12 2002 - 12:24:25 CDT

see answers embedded below...
carlos

----- Original Message -----
From: "Peter Varnai" <Peter.Varnai_at_ibpc.fr>
To: <amber_at_heimdal.compchem.ucsf.edu>
Sent: Thursday, September 12, 2002 12:27 PM
Subject: sander.LES questions

> Dear Amber-list,
>
> I have started using LES in amber7 and some questions arose:
>
> 1, after a LES-MD I would like to choose the copy that exhibits the
> lowest Epot. Where should I plug this printing in the code?

you can't do this yet.
the problem is that the energies are not stored for each copy
separately. We are trying to work on this kind of analysis
but don't want to make the sander code too messy.
the only option is to split the trajectory into separate single
copy trajectories and run mm-pbsa type analysis.
here's my suggestion: download my "analysis" program
from the moil-view page, and split the trajectory. then
use the imin=5 option in sander (amber7) to evaluate the
energies of each copy. use that data to decide what you
want to do.

>
> 2, would it be useful to use VRAND to increase chances
> (with temp0=temp0les)?

increase chances of what? this isn't clear to me.

>
> 3, what are the advantages of lower temperature copies and thus separate
> heat bath?

this has been studied in the past, for example Ulitsky and Elber,
Straub, and others. Technically the LES copies represent a higher T
ensemble than the non-LES system. In practice we reduce the T of
the copies somewhat, say to 200-250K. exactly how much to do this
can take some trial and error. sorry I can't be more help right now.
>
> 4, I use MOIL-View on an SGI: is there any chance to export it to my
> linux?
>

We are working on this, it is far from trivial. The current Linux
version will load pdb files and let you rotate them, but there isn't
much more functionality yet.

> 5, with the default temp0les=-1 I get a message that DOF in LES region is
> zero - this might have to do with heat bath coupling but not with
> actual movements of the copies, is that correct?
>
correct, it just means no degrees of freedom have been placed
in the LES bath, the LES copies are all in the regular bath.

> Thanks,
> Peter Varnai
>
>