[gmx-users] HELP：how could the simulation restarted from md*_prev.cpt rather than the md*.cpt

[张慧玲]

Hi, Justin I have 17 replicas, so I got {md0.cpt,md0_prev.cpt},   ......{md16.cpt,md16_prev.cpt}. Then I told gromacs to restart from the _prev.cpt using the following command:  mpirun -np 17 mdrun -deffnm md -multi 17 -noddcheck -replex 50000 -nsteps 100000000 -cpi md_prev
However, it seems that Gromacs neglect the cpi option and rerun from step 0...


> On 10/26/16 11:05 PM, ??? wrote:
> > Hi?
> >     I have 17 replicas for a simulation, and Gromacs outputs two cpt files for each replica: md*_prev.cpt and md*.cpt
> >     I know I can restart the simulation from checkpoint using the command "mpirun -np 17 mdrun-deffnm md -multi 17 -noddcheck -replex 50000 -nsteps 100000000 -cpi md", but how could I restart the simulatio from md*_prev.cpt rather than the md*.cpt?
> >     Thanks a lot for your help.
> >
> 
> Tell mdrun to do that with -cpi md_prev.cpt instead of what you're doing now, 
> which is -cpi md.cpt
> 
> -Justin
> 
> 
> -- 
> ==================================================
> 
> Justin A. Lemkul, Ph.D.
> Ruth L. Kirschstein NRSA Postdoctoral Fellow
> 
> Department of Pharmaceutical Sciences
> School of Pharmacy
> Health Sciences Facility II, Room 629
> University of Maryland, Baltimore
> 20 Penn St.
> Baltimore, MD 21201
> 
> jalemkul at outerbanks.umaryland.edu | (410) 706-7441
> http://mackerell.umaryland.edu/~jalemkul
> 
> ==================================================