[gmx-users] Possible to stop mdrun changing file names when usingcheckpoint files?
gmx3 at hotmail.com
Tue Feb 17 07:55:44 CET 2009
> Date: Tue, 17 Feb 2009 16:59:42 +1100
> From: Dallas.Warren at pharm.monash.edu.au
> Subject: RE: [gmx-users] Possible to stop mdrun changing file names when usingcheckpoint files?
> To: gmx-users at gromacs.org
> > One option is to avoid checkpoint files all together. I personally use
> > a grompp/mdrun cycle of ~2h segments that was historically done in
> > order to utilize the -sort -shuffle options to gromacs 3. I found it
> > simpler to keep on doing the exact same thing in gromacs 4, even
> > though I no longer -sort or -shuffle. It appears to be working fine
> > for me. I suspect that you could do the same type of thing via tpbconv.
> Yes, that is what I was doing previously and are continuing to do so with some at the moment with 4.0.X I suppose the question is now, is it a more continuous simulation to be using the checkpoint files, or does tpbconv with the energy and trajectory files provide the same degree of accuracy?
The trajectory and energy file do not store all the state variables for all
thermostats and barostats. But resetting these variables introduces
such a small error in most cases that it is negligible.
I think I'll add a -rename (or something like that) flag to mdrun,
so you can control the renaming.
> Totally unrelated and a little nugget for people that happen to have an iPhone or iPod touch, you can get an application called Molecules for free that allows you to view .pdb files, download direct from the pdb database or from custom location.
> Catch ya,
> Dr Dallas Warren
> Pharmacy and Pharmaceutical Sciences
> Monash University
> A polar bear is a Cartesian bear that has undergone a polar transformation
Express yourself instantly with MSN Messenger! Download today it's FREE!
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the gromacs.org_gmx-users