[gmx-users] Continue run

Milan Melichercik melicher at cray.dbp.fmph.uniba.sk
Thu Sep 7 17:58:32 CEST 2006

Dňa Št 7. September 2006 15:27 Navratna Vajpai napísal:
> Dear All..
> to continue the run, as i understood, I modified the md.mdp and then
> use the .tpr and .trr for the further run. Can anyone suggest that it
> is the right way or not? actually in one of the tutorials note i have
> just found to change the options as -time $value -until $value.
> I was wondering whether my way is also OK or not.

Nearly right, but you don't need to modify the .mdb file. If your simulation 
crashed, simply execute 'tpbconv -s your_file.tpr -f your_file.trr -o 
your_new_file.tpr' (or you can add -e your_file.edr for energy) - this will 
generate new .tpr, which you use by running mdrun - that will generate 
something like your_new_file.trr and you can join then by 'trjcat -f 
your_file.trr your_new_file.trr -o your_full_file.trr, which will contain the 
whole simulation.
The -time $value generates .tpr, which will start simulation from $value time 
(normally it starts from the last possible positions, so this is in the most 
cases not necessary). -until and -extend parameters you can use for 
continuing finished simulation until the given time/for the given time.
The parameters of the simulation is allways the same.
I hope it will help.


More information about the gromacs.org_gmx-users mailing list