[gmx-users] Re: MD run - gen_vel

[Justin Lemkul]

On 6/17/13 4:13 PM, Dariush Mohammadyani wrote:
> Dear Justin,
>
> I have sent a post on GROMACS-user and did not get any answer.
> Could you please help me out to find the answer?
>

I did not see the original post, but that doesn't really matter.  Please keep 
all Gromacs-related questions on the gmx-users list.  I am not a private help 
service, nor am I the only person capable of answering this question.

I am CC'ing this message to the list and ask that all further discussion occur 
there.

> I am going to run three MD runs on a well-equilibrated system (NPT and NVT) to
> make sure what I have seen on first run is similar to others.
> Actually, I use "gen_vel=no" in md.mdp. However, in this case I should use
> "gen_vel=yes" with different numbers for seed to generate different trajectories
> to compare.
> Am I right?
>

One should only initialize velocities at the outset of a new simulation (i.e. 
the beginning of equilibration).  Re-generating velocities afterwards is not 
normally productive and can often induce instability.

> Second, If I want to continue previous MD run for longer time, "gen_vel" should
> be yes or no?
>

Per the information above, it should be "no."  Proper continuations are done by 
providing .cpt files to grompp -t.

-Justin

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

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

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