Skip to content

Latest commit

 

History

History
99 lines (79 loc) · 3.4 KB

GROMACS_procedure.md

File metadata and controls

99 lines (79 loc) · 3.4 KB

The following is a procedure for performing GROMACS simulations and ACFCalculator calculations on a system composed of a DPPC membrane with a harmonically restrained solute.

###GROMACS Simulations

All the necessary files for running a GROMACS simulation are in the permeation/ directory which is set up as:

permeation/  
  down/  
    sim_0/  
      /* GROMACS input files (ex: .mdp, .gro, ...)  
    sim_1/ 
    .... 
    sim_39/  
    makemdp.py  
    gromacssubmit-list
    pull-template.mdp  
  down-nve/  
    sim_0/ 
    ... 
    sim_39/  
    makemdp.py  
    gromacssubmit-list  
    gromacssubmit-list-restart  
    pull-template.mdp  
  up/  
    ...  
  up-nve/  
    ...

Sample GROMACS input files are included in this Git. Note that these are not sufficent to set up the system, for that see a GROMACS guide. The following commands create individual .mdp files for each equilibration simulation with their unique restraint reference positions, and submits the jobs to the queue:

python makemdp.py
./gromacssubmit-list [job name e.g., down/up] -np [number of processors per simulation, 8] -c [number of windows, 40]

Run these for both the down/ and up/ directories. This procedure assumes use of the PBS queueing system. The gromacssubmit-list file will have to be modified if using a different queueing system.

After the equilibration simulations are complete (check with qstat), copy the directories:

cp -r up up-nve
cp -r down down-nve

The following commands regenerate the necessary .mdp files and submits the NVE jobs to the queue:

python makemdp.py
./gromacssubmit-list [job name e.g., down-nve/up-nve] -np [number of processors per simulation, 8] -c [number of windows, 40]

The NVE jobs might not complete (check using wc -l pullx.xvg, the files should contain 5000016 lines). If they're incomplete, resubmit with:

./gromacssubmit-list-restart [job name e.g., down-nve/up-nve] -np [number of processors per simulation, 8] -c 
[number of windows, 40]

Once the simulations are complete, we need to copy the .xvg files into a separate directory to run ACFCalculator:

cd permeation/down-nve
mkdir xvg_files
find . -name "pullx.xvg" -exec cp --parents \{\} \xvg_files \;

cd xvg_files
rm -r down
rm -r xvg_files

The find command adds the .xvg files from the down/ directory, and those it already copied to xvg_files, which is why we remove the redundant files. Repeat this process for up-nve/.

###ACFCalculator

sh setup.sh

The script setup.sh runs ACFCalculator and is located in ACFCalculator/build/ with the ACFCalculator executable. It assumes the input files can be found as /${HOME}/permeation/down-nve/xvg_files/sim_${i}/pullx.xvg for i:[0,39] (this will be true if following the above procedure). It will save the ACF calculations as /${HOME}/acf_down/acf_down_sim${i}.dat and the output from ACFCalculator as /${HOME}/out_down/out_down_sim${i}.out. Explanation of the options used in this file can be found in the README. ACFCalculator prints some data to screen, all of which is saved in the .out files.

python getDall.py

The getDall.py script compiles all of the VACF D(s) values from the sim_${i}/ directories into a single out_ds.out file. The output file is a space-delineated file containing two columns: z-position (in Angstroms) and D(s) in cm^2/s. Any graphing utility can be used to easily plot this data.