From: JC Gumbart (gumbart_at_ks.uiuc.edu)
Date: Fri Jun 20 2008 - 21:23:44 CDT
That's incredibly high. Typically, 1000 steps is sufficient with a
few thousand more being required to handle large clashes, etc. To
know how well your minimization is proceeding, watch the gradient
tolerance. Once it gets to single digits (more or less), it should
be ready to simulate.
Equilibration will depend highly on your system; we usually run a few
nanoseconds, watching things like the RMSD to judge. I happened to
find a great post from a couple years ago regarding this same issue:
http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l/4364.html
On Jun 20, 2008, at 2:05 PM, S.K. Ghosh wrote:
>
> Hi NAMD users,
>
> I have set the no. of steps of minimization for my job to be
> 500000. Is that a bit too high? This job is to minimize a system of
> biotin-streptavidin immersed in water sphere.
>
> Total atoms: 47000+
> No.of biotin-streptavidin atoms ~ 7000
> Water ~ 40000
>
> Could someone please give an estimate of no. of steps for
> minimization and equilibration for such a system (nearest order or
> magnitude).
>
> Your suggestions will be much appreciated.
>
> Thanks,
> Sourav
This archive was generated by hypermail 2.1.6 : Fri Oct 10 2008 - 23:18:19 CDT