[gmx-users] gromacs.org_gmx-users Digest, Vol 120, Issue 95

[Justin Lemkul]

On 4/22/14, 7:18 AM, sunita gupta wrote:
> Hello,
>
> Several time I have extended my jobs using the protocol you have mentioned
> above. I always used to get my output appended to the previous one, which
> never required any concatenation.
> This time my aim was to extend the job using new.mdp files (as the previous
> one contained simulated annealing protocols to reach 500K temp and the new
> job to have trajectories at 500K only)
> So, I used 1st protocol of Changing .mdp file given in
> http://www.gromacs.org/Documentation/How-tos/Extending_Simulations. grompp
> -f new.mdp -c old.tpr -o new.tpr -t old.cpt and mdrun -s new.tpr
> As, a result I got new files this time, creating backup files for the
> previous ones. I renamed those backup files as old files and trying to
> concatenate with the new ones. I have only one .cpt file now, as its backup
> file was not created.
> Please let me know, If I am wrong or any alternate way to do this.
>

If you re-run grompp, the default values of init_t and init_step are zero.  If 
you didn't reset those values, your trajectory will start over from time = 0. 
If you want to concatenate the resulting output to a previous run, you should 
indeed reset the start time with trjcat -settime, as stated before.

-Justin

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

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

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