[gmx-users] generating new velocities and exploring more of the phase space

[Justin A. Lemkul]

Hassan Shallal wrote:
> Dear Gromacs users,
> 
> -The usual target is: hopefully exploring different areas of the phase space
> by running at least three independent MD experiments starting with three
> different velocities
> 
> -The usual scenario is: (Minimization) then (equilibration) then (production)
> 
> 
> I generate velocities at the beginning of the equilibration (gen_vel is set
> to yes, and gen_seed is set to -1 in the equlibration .mdp) and continue to
> the production (gen_vel is set to no in the production .mdp).
> 
> -The assumption is: By running a new equilibration for each MD experiment, I
> guarantee ending up having different statrting velocities for the production
> phases of these independent experiment and achieve the above target. The
> confusing part for me here is that the protein is position-restrained during
> the equlibration and only the solution is allowed to equilibrate using the
> assigned velocities.
> 
> -Is the above assumption right? If yes, is there a better way? If no, how
> could I achieve the above target?
> 

Partly.  You can confirm that the velocities are different using either gmxcheck 
or gmxdump.  You should not assume that only the solvent is equilibrating, so to 
speak.  See below.

> -Does the protein atoms get assigned velocities at the start of the
> equlibration (different velocities by running the equlibration again) and not
> allowed to use it until the are released from the position-restrain at the
> beginning of the production?
> 

Velocities are assigned, and the protein atoms can, in fact, move even while 
under the influence of position restraints.  The use of restraints simply adds 
an energy penalty to disfavor motion, but does not preclude the possibility of 
motion entirely.  Again, gmxcheck or gmxdump will tell you what these velocities 
are.

-Justin

> I would appreciate any help responding to the above questions and thoughts.
> 
> Hassan
> 
> 
> 
> 

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================