[gmx-users] Re: Continuing a simulation plus another error

[Justin Lemkul]

On 12/24/12 12:43 PM, Ankita naithani wrote:
> Hi Justin,
>
> Thank you so much for your response.
>
> I also noticed that when I type in gmxcheck -f md.cpt to see which
> version was used, I get the same fatal error of "Attempting to read a
> checkpoint file of version 13 with code of version 12".
>
> In my linux machine, the gromacs version is 4.5.5 and the one on the
> cluster is also 4.5.5. I just offloaded the data from the cluster and
> was trying to run the test for continuing simulation before submitting
> it to the cluster for continuation.
>

There is a mismatch somewhere.  Run any command simply with the -h flag or read 
the printed information of any command invocation to see the version.  Your 
executables are newer than the version of Gromacs used to do the original run.

> Could you please suggest me something in this regard too. I understand
> that the versions would be different but my system and cluster have
> the same version of gromacs. Also, is version 13 being used  by
> gromacs 4.6? In that case, how do we continue a simulation with that
> version?
>

You should use a consistent version throughout.  Newer versions should always be 
backwards compatible, but unless you have a specific reason for upgrading 
(especially between major releases), I would avoid it.

> If somehow someone updated the software or switched it to gromacs 4.6
> in the cluster, I will then directly invoke mdrun there but I would
> need to know if the command would be the same for continuing the
> simulation in gromacs 4.6?
>
> i.e.
>
> mdrun -cpi md.cpt -s md.tpr
>

The commands haven't changed in this respect.  Adding -cpi -append to whatever 
you ran before is sufficient to continue the simulation if everything else is 
consistent.

> Sorry for bothering with silly questions, my time on the cluster runs
> out in 2 days so I really could do with as much help I can for now
>

You can always install your own version of Gromacs in a location accessible to 
you, like your home directory, and run using those executables.  Silently 
replacing a version of software and nuking the old one is a great way to upset 
users, so hopefully the other version is still available somewhere, even if not 
in the default $PATH.  Only your sysadmins know that, though.

-Justin

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

Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

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