[gmx-users] replicates with different random seed

[Shreyas Kaptan]

If you use the .cpt file then the velocities will be provided as from the
checkpoint. This would effectually only lengthen the simulation. In order
to truly sample the simulation space from a "random" seed, you
should reassign velocities from scratch as in the .mdp file. In *principle*
the starting coordinates, if used from the end point of your 100ns data,
would actually bias the new trajectory. But that should not be a
substantial bias if, for example, the simulation system did not undergo
drastic conformational change during your 100ns run.  if it did, then I
would rather restart the whole simulation from the same starting
coordinates and different velocities to test the dynamics of the system
i.e. if the conformational change is reproducible or just a flue of N=1.

On Tue, May 8, 2018 at 11:13 AM Alex <nedomacho at gmail.com> wrote:

> Ideally, you should repeat your entire procedure with a new seed from the
> step that first generates velocities.
>
> Repeating hundreds of nanoseconds of simulation a few times probably would
> not be my response to your likely 85 year old reviewer, but I hope this
> helps.
>
> Alex
>
> On May 8, 2018 2:58 AM, "vijayakumar gosu" <vijayakumargosu at gmail.com>
> wrote:
>
> Hello all,
>
> I have submitted a paper with 100ns simulations data, however reviewer
> suggested to provide replicates with different random seeds. I have used
> below lines in the mdp file of equilibration step for my first simulation.
> now i am confused  for replicates (minimum 2) whether i have to start from
> equilibration step or i can start from production step using the npt.gro
> and npt.cpt from first simulation. additionally, I assume that when we have
> ( for example 3 replicates) we have to consider the average from the 3
> simulations data. Please advise me whether my approach is reliable...
>
>
> nvt.mdp
>
> ; Velocity generation
> gen_vel = yes
> gen_temp = 300
> gen_seed = -1
>
> Thanks in advance
>
>
> Best,
>
> Vijayakumar Gosu
>
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send
> a mail to gmx-users-request at gromacs.org.
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>


-- 
Shreyas Sanjay Kaptan