[gmx-users] What is the "gen-vel" used for?

[Justin Lemkul]

On 1/2/20 10:07 PM, Sun Yeping wrote:
> I guess I understand what you mean, but could you clarify it further? I
> usually perform the simulations by the following process:
>
> (1) energy minimization (EM); (2) NVT for 1 ns with heavy atom restraints;
> (3) NPT for 1 ns with heavy atom restrains; (4) production simulation
> without restraints.
>
> To prove repeatability, I think I need to set up several parallel
> simulations. Could I perform the NPT step for 3 different durations such as
> 1 ns, 2 ns, 5 ns to get three different velocities at the final frame of
> the three NPT simulations, and remove the restraints and continue the
> simulation from these velocities respectively? Thus I will get three
> production simulations start from different velocities.
>
> Or do I need to stop a so-called production simulations (without restraint)
> at different time points and get three configurations, and repeat the EM,
> NVT, NPT and production simulation steps for the three configurations
> respectively to get three independent trajectories?

The conventional approach is to generate velocities at the beginning of 
what you have defined as step (2) above. Continuing NPT is basically 
just taking time points of the same run and continuing it. One could 
argue that those simulations will eventually diverge, but I don't think 
you can truly call them independent runs. They all started from the 
exact same input/state but were just allowed to run for different 
amounts of time.

-Justin

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

Justin A. Lemkul, Ph.D.
Assistant Professor
Office: 301 Fralin Hall
Lab: 303 Engel Hall

Virginia Tech Department of Biochemistry
340 West Campus Dr.
Blacksburg, VA 24061

jalemkul at vt.edu | (540) 231-3129
http://www.thelemkullab.com

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