[gmx-users] mdrun -c option "did not work"

[Mark Abraham]

Jérôme Baffreau wrote:
> Thanks for your answer.
> 
>>> Actually no my run did not finish normally the first time (it has 
>>> been stopped after 7 days because of the time limitation on the 
>>> cluster it was running on). So I started it again (I ran the same 
>>> command and added the option "-append yes") and this time the 
>>> calculation finished normally, but when I look at the md.log file it 
>>> looks like the second run did not add anything since it stops in the 
>>> middle of a line (in the middle of a word actually!)...
> 
>> Sounds like a filesystem problem.
> 
> 
> No, that's because the Linux cluster's system terminated my calculation 
> when it reached max. allowed time.
> 
> 
>>> Maybe I just did not use the correct command to start this 
>>> calculation again, that would explain the end of the md.log and the 
>>> "no-generation" of the final structure.
>>> What option(s) should I use with trjconv? Actually I already tried to 
>>> use this command, but it generated a 200 MB pdb file which was not my 
>>> aim!
> 
>> Read trjconv -h and see how to choose the only final frame.
> 
> 
> I have a stupid question: how do I know how many frames there are in my 
> calculation?

gmxcheck and most analysis tools will mention the number of frames they 
analyse.

Mark