Difference between revisions of "ASPA Molecular Dynamics Simulation"

From Bioinformatikpedia
(Created page with "Coming soon to a Wiki page near you (this one, in fact): The results of the LRZ computation (We apologize for the inconvenience, but we are still waiting for our jobs to finish…")
 
 
Line 1: Line 1:
  +
We kept strictly to the [[Task 8 - Molecular Dynamics Simulations|tutorial]] for almost all of the steps needed to do this task and run a molecular dynamics simulation for the wild type ASPA reference structure, the CM994594 mutation and the RS104894553 mutation. We encountered minor difficulties when trying to start our computing jobs ob the LRZ cluster, but these were quickly taken care of by the LRZ via a software update.
Coming soon to a Wiki page near you (this one, in fact):
 
   
  +
We hit another obstacle, though: One of the three jobs, the one for RS104894553, refused to complete successfully, throwing the following error:
The results of the LRZ computation
 
   
  +
<pre> Fatal error:
(We apologize for the inconvenience, but we are still waiting for our jobs to finish. Sorry!)
 
  +
1 particles communicated to PME node 1 are more than 2/3 times the
  +
cut-off out of the domain decomposition cell of their charge group in
  +
dimension x.
  +
This usually means that your system is not well equilibrated.
  +
For more information and tips for troubleshooting, please check the GROMACS
  +
website at http://www.gromacs.org/Documentation/Errors</pre>
  +
  +
As it turned out, the minimization of the system had not converged earlier on; we tried using a smaller step size and more steps in total and managed to have the system converge to machine precision in 6500 steps. As this did not yield any improvement to the original issue, and as our local computing resources were unable to support further increases in either nsteps or step size, we finally had to abandon the RS104894553 mutation.
  +
  +
The analysis task will therefore concentrate on the wild type reference and the CM994594 mutation.

Latest revision as of 07:52, 4 November 2011

We kept strictly to the tutorial for almost all of the steps needed to do this task and run a molecular dynamics simulation for the wild type ASPA reference structure, the CM994594 mutation and the RS104894553 mutation. We encountered minor difficulties when trying to start our computing jobs ob the LRZ cluster, but these were quickly taken care of by the LRZ via a software update.

We hit another obstacle, though: One of the three jobs, the one for RS104894553, refused to complete successfully, throwing the following error:

 Fatal error:
1 particles communicated to PME node 1 are more than 2/3 times the
cut-off out of the domain decomposition cell of their charge group in
dimension x.
This usually means that your system is not well equilibrated.
For more information and tips for troubleshooting, please check the GROMACS
website at http://www.gromacs.org/Documentation/Errors

As it turned out, the minimization of the system had not converged earlier on; we tried using a smaller step size and more steps in total and managed to have the system converge to machine precision in 6500 steps. As this did not yield any improvement to the original issue, and as our local computing resources were unable to support further increases in either nsteps or step size, we finally had to abandon the RS104894553 mutation.

The analysis task will therefore concentrate on the wild type reference and the CM994594 mutation.