Klauda Lab Simulation Wiki

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
[[bilayer_simulations]]

==== Production run instructions for Zaratan ====


Listed below are some instructions for the production run simulation on Zaratan. As we discussed, this will automatically run your simulation in chunks with each being a certain length.  The files to copy (not the sub-directories) are in the following location on Zaratan to your directory with step6.1 to step6.6 namd files:

**/home/jbklauda/sim-inputs/namd**


There are key things to change in the dyn.csh file:

  - # number of nodes to run your job (talk to Dr. Klauda)
  - You should adjust the job name to something for your system. 
  - Adjust the SBATCH and namd run lines per the number of nodes (see examples in the the following directory: ~jbklauda/sim-inputs/namd)
  - Also there is a line with set  simt = 1000000. This is to run 1 Million steps of 2ns per block (with 2fs timesteps).  Discuss with Dr. Klauda and/or a graduate student on what to change this to and also change in the dyn-2.inp file.  

You then will need to adjust dyn-1.inp and dyn-2.inp to match with what step7_production.inp has. These files will be combined into a dyn.inp file with the firsttimestep defined (see dyn.csh file). The easiest way to do this is to copy and paste the psf/pdb lines in step7*inp to replace that in dyn-1.inp. Then, copy most of the step7*inp to replace the material in dyn-2.inp. One key important change will likely be the values for the "run" and "numsteps" lines in dyn-2.inp based on how many ns per dyn file. These should be set to the simt value.  

Note: dyn-1.inp inputname should stay as dyn

Once the changes are made you will also need to copy the step*6.6.xsc/.vel/.coor to dyn.xsc/vel/coor so the job will use these to restart from your equilibration runs. Then to run the production run, just type:

<code commend> ./dyn.scr </code>

Also note that there is an important file (last.seqno). This tells the daisy-chain script when to stop.  So if you want to run for a total of 100ns you'll need 50 files so this is the value that should be in the last.seqno file.  If you want this to be longer or short just change the number in the file.