Re: two stage equilibration in single configuration file

From: Brian Radak (
Date: Thu Aug 03 2017 - 12:45:14 CDT

If you want rigorous constraints as in, these atoms have zero
velocities, then no, you cannot. If you are satisfied with harmonic
restraints with respect to a reference structure, then yes, this is very
simple. For example I start pretty much all of my simulations from
crystal structures this way:

constraints on
consref <pdbfilename>
conskfile <pdbfilename>
conskcol B
constraintScaling 1.0

minimize 500
reinitvels [temperature]
firstTimestep 0
foreach scale [list 1.0 0.5 0.25 0.125 0.0] {
     constraintScaling $scale
     run 100000

where <pdbfilename> is the PDB from the complete system (usually the
same protein structure as in the deposited PDB) and the B-factor column
has been modified to flag all heavy atoms in the protein, etc. The
harmonic restraint will have a force constant equal to the value in that
column (I usually use 10 - this is arbitrary). The Tcl foreach loop
gradually reduces the force constant by a factor of 2 (you can do more
or less). You can also run additional dynamics after the loop is complete.

Minor caution - I'm not sure that constant pressure makes complete
rigorous sense in this case. It should work fine, but I consider the
dynamics in this period completely artificial and rarely analyze it.


On 08/03/2017 01:03 AM, wrote:
> I want to carry out a two stage equilibration in the first stage i don't
> want my protein to move only the water molecules will take part in the
> dynamics and in the second stage whole system will equilibrate,my question
> is that can it be done in a single configuration file?

Brian Radak
Postdoctoral Appointee
Leadership Computing Facility
Argonne National Laboratory
9700 South Cass Avenue, Bldg. 240
Argonne, IL 60439-4854
(630) 252-8643

This archive was generated by hypermail 2.1.6 : Sun Dec 31 2017 - 23:21:31 CST