[gmx-users] mdrun -rerun problems

George Abadir georgea at ece.ubc.ca
Thu Jul 19 07:12:11 CEST 2007

Yes of course I looked at the output but I thought is I specify energy 
groups it will output only those specified and not the whole system. 
However I tried a short run and it did output both. For the crash it 
turned out it is because all the nodes try to read the huge trajectory 
file at the same time if this may help somebody in the future. We are 
still trying to figure out how to solve this problem anyway.
Thanks for your time.

Mark Abraham wrote:

> George Abadir wrote:
>> Hi all,
>>         I am running simulations of carbon nanotubes and amino acids 
>> in water. I need to calculate the potential energies for the whole 
>> system as well as the interaction energy between the CNT  and amino 
>> acid. What I am tried to do is to calculate the energy for the whole 
>> system first (i.e. I did not introduce any energygrps in the run.mdp 
>> file) and then use the rerun option to calculate the interaction 
>> energy after introducing "energygrps= CNT Protein SOL" into the 
>> run.mdp file. However, each time I try to run this second step the 
>> cluster on which I run the simulation crashes with a message "Kernel 
>> panicked"!
>> Does any body know about this problem?
> Yeah, something's broken that's probably got nothing to do with 
> GROMACS, or the fact you're doing a re-run.
>> Is there another way to get both the potential energy of the system 
>> as well as the interaction potential energy? I understand that if I 
>> write "energygrps= CNT Protein SOL" from the very beginning the 
>> energy of the whole system won't be written, is that right?
> mdrun always calculates the total potential and kinetic energy, and 
> outputs it according to nstxxx flags.
> Have you not looked at typical output, or typical .edr file contents?
> Mark
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/search before 
> posting!
> Please don't post (un)subscribe requests to the list. Use the www 
> interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php

More information about the gromacs.org_gmx-users mailing list