[gmx-users] Umberella sampling

Justin A. Lemkul jalemkul at vt.edu
Sun May 8 16:33:51 CEST 2011

mohsen ramezanpour wrote:
> Dear Dr.Justin
> I didn't any equilibration for any windows! Because:
> As I know the thermodynamics of the system had wroten in NPT.cpt file 
> (for example the velocities and etc).

Then where does NPT.cpt come from?  Presumably, if you have this file, you did 
some equilibration.  I don't understand what you've done.  If you want to rely 
on previous, equilibrium dynamics under an NPT ensemble, you must do equilibrium 
MD (not steered) for each configuration under the desired conditions, then pass 
the .cpt file to grompp -t when building the umbrella sampling input files.

> Then when I use that  in grompp for each windows ,I expect  the same 
> characteristics of system as before(At least the same T)!

With "gen_vel = no" and no record of the prior state, this is not a reasonable 
expectation.  You will get a temperature based on the initial forces present in 
the system (which determine initial velocities).  The resulting temperature may 
or may not be identical to the previous state, but under the influence of a 
thermostat, you will reach the desired temperature.

If you're attempting to follow my tutorial directly, at this point I think you 
should stop and consider the discussion I have in the tutorial on maintaining 
the ensemble:


In the example system (one I know very well), it was satisfactory to simply use
"gen_vel = no" and allow these initial forces to govern the resulting dynamics. 
  In many other cases (probably most), you should:

1. Generate your configurations (SMD, etc)
2. Run individual NPT simulations (or whatever ensemble) for each desired 
starting configuration (and thus obtain npt0.cpt, npt1.cpt, etc)
3. Initiate umbrella sampling windows using the npt*.cpt files from each window.

Otherwise, skip step 2, use "gen_vel = yes" and discard the first few hundred ps 
as equilibration in each window.

> Actually after PULLING step I used the following commands:
> grompp  -f  umbrella.mdp   -c   config0.gro  -p  complex.top  -o  
> umbrella0.tpr  -n index.ndx  
> mdrun  -v  -deffnm  umbrella0    -px  pullx0.xvg   -pf   pullf0.xvg   
> -cpi   NPT.cpt
> what is your point of view now?

These commands do not necessarily preserve the ensemble.  Without providing 
suitable information to grompp -t, the initial state is lost.  I still don't 
know where NPT.cpt comes from.  You say you're running grompp immediately after 
pulling, which is a non-equilibrium ensemble that you don't want to preserve as 
input into umbrella sampling.  That, and likely the settings in SMD vs. US are 
different, so if NPT.cpt corresponds to your SMD simulation, you probably don't 
want to be using that for all your umbrella sampling windows, as the information 
it contains is for one single snapshot somewhere along the SMD run.  If this is 
not the case, and you've generated NPT.cpt in some other manner, you will have 
to explain where it came from.

> What do you think if I use NPT.enr  OR   NPT.trr   OR   in 
> grompp(especially  NPT.enr) ?

Not necessary.  Using .cpt files is more accurate.

> Because all of information about NPT step are included in this files.

Note entirely.  The differences between using .edr + .trr vs. .cpt alone are 
very small, but using a .cpt file is the most rigorously correct way to maintain 
the ensemble.


> Thanks in advance for your
> On Sun, May 8, 2011 at 5:47 PM, Justin A. Lemkul <jalemkul at vt.edu 
> <mailto:jalemkul at vt.edu>> wrote:
>     mohsen ramezanpour wrote:
>         Dear Dr.Justin
>         Regarding doing umbrella sampling:
>         I used gen_vel =no
>         and I prefer to use from thermodynamics of system in  NPT.cpt
>         Although the T and P were as I did set in NPT.mdp file (T=310
>         ,P=1 bar)
>         But  their values in my umbrela.log files are not as before (for
>         example  P=-8.434578e+1)
>         Where did I wrong?
>     I have no clue.  You didn't provide your commands or any evidence
>     that the prior ensemble was to be maintained.  If you have done NPT
>     equilibration in each window, then pass the .cpt file to grompp and
>     the ensemble should be preserved in each window.  Note that an
>     instantaneous pressure value means nothing, as discussed weekly on
>     this list and online:
>     http://www.gromacs.org/Documentation/Terminology/Pressure
>     It is the average that is most meaningful.
>     -Justin
>     -- 
>     ========================================
>     Justin A. Lemkul
>     Ph.D. Candidate
>     ICTAS Doctoral Scholar
>     MILES-IGERT Trainee
>     Department of Biochemistry
>     Virginia Tech
>     Blacksburg, VA
>     jalemkul[at]vt.edu <http://vt.edu> | (540) 231-9080
>     http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin
>     ========================================
>     -- 
>     gmx-users mailing list    gmx-users at gromacs.org
>     <mailto:gmx-users at gromacs.org>
>     http://lists.gromacs.org/mailman/listinfo/gmx-users
>     Please search the archive at
>     http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
>     Please don't post (un)subscribe requests to the list. Use the www
>     interface or send it to gmx-users-request at gromacs.org
>     <mailto:gmx-users-request at gromacs.org>.
>     Can't post? Read http://www.gromacs.org/Support/Mailing_Lists


Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080


More information about the gromacs.org_gmx-users mailing list