Dan,
the current (Amber9) REMD does indeed restart sander after every
exchange. We have a prototype working now that doesn't do this and would
probably work better for your case. We're planning for Amber10 release but
if this is important for you to try now we can give you access to
the modified amber9 to test out. This will overcome the issue of
redoing the initialization every exchange. Contact me directly if
you're interested.
I'm not sure about the weighting, maybe Dave Case can comment
on that and RDC stability in general.
Carlos
On 6/22/07, McElheny, Dan <dmcelh1.uic.edu> wrote:
>
> Hi
>
> i'm trying to run remd refinements with rdc's as restraints. runs with
> noe's/torsion restraints seem fine. however, with rdc, things seem
> somewhat unstable. anyway, is it possible to weight the rdc's in a manner
> similar to noe's during each step? this might help in the stability.
> also, how is the alignment tensor treated when exchanges occur? it would
> be nice to retain an evolving tensor throughout the run and not have to
> recall the initial guess at each exchange. please advise.
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber.scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sun Jun 24 2007 - 06:07:57 PDT
