- From: Rick Venable <rvenable)at(pollux.cber.nih.gov>
- Subject: Re: CCL:CCL
- Date: Wed, 7 Jan 2004 14:14:59 -0500
You make my points clearly-- don't use just one probe to define
equilibration, and it may require longer simulation times than 10-50 ps
to achieve equilibration.
The timescale plots are interesting, but I do wonder about such a large
change in backbone RMSD; over 40 A suggests quite a large change from
the xtal coords. What were the simulation conditions for this? Is the
"equilibrated" structure obtained at long times significant, i.e. is
there any evidence to indicate agreement with epxt data?
> Rick Venable wrote:
> > For a solvated protein simulation, I'd monitor the potential energy,
> > radius of gyration, and the bacbone RMSD relative to the starting
> > structure. When all 3 of these become flat, one could argue that
> > the system is equilibrated; this could require 100s of picoseconds.
On Wed, 7 Jan 2004, Anton Feenstra wrote:
> Although indeed one can argue this, it is hard to tell what 'flat'
> means in a practical sense. For a seond-year lecture I give I have
> three slides each giving the rmsd of a protein w/r to the x-ray
> structure. All three of them level off to a nice plateau at the end,
> and appear to be genuinly equilibrated in the way you define that.
> But, the three plots are of the same simulation, the first of 0-100ps,
> the second of 0-1ns, and the last of 0-10ns! If you overlay them, they
> are almost the same, except for the scale of the axes! Those of you
> who'd like to see them, please ask me and I can send you the three jpg
> files (~50k each).
> While I don't regularly monitor r_gyr or e_pot, I don't know but
> suspect that they would show similar behaviour... (although they may
> equilibrate much faster than the rmsd).
Rick Venable 29/500
FDA/CBER/OVRR Biophysics Lab
1401 Rockville Pike HFM-419
Rockville, MD 20852-1448 U.S.A.
(301) 496-1905 Rick_Venable)at(nih.gov
ALT email: rvenable)at(speakeasy.org
"Don't blame me, I voted for Kang."