User Tools

Site Tools


exercises:2018_ethz_mmm:pmf

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Next revision
Previous revision
exercises:2018_ethz_mmm:pmf [2018/05/25 10:19] – created dpasseroneexercises:2018_ethz_mmm:pmf [2020/08/21 10:15] (current) – external edit 127.0.0.1
Line 43: Line 43:
 Then we will use integration of the average force between the molecule and the substrate, along the adsorption path, to extract the free energy difference between two configurations: the "very far" configuration and a configuration along the path. Then we will use integration of the average force between the molecule and the substrate, along the adsorption path, to extract the free energy difference between two configurations: the "very far" configuration and a configuration along the path.
  
 +<note important>  
 +  * After you copied the exercise_12 directory and entered it, look at the mol_sub.xyz file (by editing or vmd) and understand the geometry of the system: which range of distances should you consider for MD runs?
 +  * Create a directory T_300 and copy into it the following files and cd into the directory:
 +<code> cp run* *pot c1.topo md_temp.inp get_pot_mean_force T_300  ; cd T_300 </code> 
 +  * At this point edit the **run_distance_loop** script and insert the list of distances you want to simulate.
 +  * How many steps will you run per each distance? This will be decided by editing the input file.
 +</note> 
  
 +<code - md_temp.inp >
  
 +# PMF Input script template
 +# (1) Initialisation
 +units real
 +atom_style full
 +boundary p p p
 +pair_style hybrid/overlay lj/charmm/coul/long 10.0 12.0 buck/coul/long 12.0 morse 10.0
 +bond_style harmonic
 +angle_style harmonic
 +dihedral_style charmm
 +kspace_style pppm 0.0001
 +# (2) Define atoms
 +read_data c1.topo
 +group molecule molecule 1
 +group substrate molecule 2
 +group bottom molecule 3
 +# (3) Settings
 +# freeze bottom layer of substrate to prevent it from drifting
 +fix 2 bottom setforce 0 0 0
 +include new_defpot.pot
 +include best_all.pot
 +neighbor 2.0 bin
 +neigh_modify delay 0
 +timestep 1.0
 +# (4) NVT Dynamics
 +fix temp1 molecule nvt temp _TEMP_ _TEMP_ 100
 +fix temp2 substrate nvt temp _TEMP_ _TEMP_ 100
 +velocity                        all create _TEMP_ 293288
 +fix PMF molecule recenter NULL _Y_ NULL
 +compute temp_molecule molecule temp
 +compute yforce molecule group/group substrate kspace yes
 +thermo_style custom time c_yforce[2] etotal pe c_temp_molecule temp ke evdwl press enthalpy
 +thermo_modify flush yes
 +thermo 50
 +dump xyz all xyz 100000000 mol_sub.xyz
 +dump_modify xyz element C C H C N K Cl
 +dump coord all dcd 5000 trajectory.dcd
 +restart _NSTEPS_ TCB_PMF.restart
 +run _NSTEPS_             
 +</code>
 +
 +
 +<note important>
 +  - Edit the input file to run 50 picoseconds, to write the restart at the end, to thermalize at the wished temperature
 +  - **REPLACE _NSTEPS_ WITH THE NUMBER OF STEPS AND _TEMP_ WITH THE DESIRED TEMPERATURE IN THE md_temp.inp FILE** 
 +  - The length of the simulation is small to get converged averages, but the run will thus last a few minutes per distance making the exercise feasible on 16 cores.
 +  - Now run the chain of simulations
 +<code> qsub run_distance_loop </code>
 +  - Directories R_<distance> will be created (check with ** ls -ltr ** )
 +  - You can enter a certain directory (e.g., R_12.5), check the **log.lammps** for the evolution of a trajectory, and visualize the trajectory with: **vmd mol_sub.xyz trajectory.dcd**
 +  - At the end of everything, you can perform averages and integrals by understanding (and using) the script ** get_pot_mean_force **  FROM WITHIN THE DIRECTORY T_300:
 +<code>
 +./get_pot_mean_force
 +</code>
 +  - THIS PRODUCES THE POTENTIAL OF MEAN FORCE pot_mean_force for that temperature
 +  - Repeat the same for other temperatures (e.g.) 10 K and 800 K.
 +  - Potentials can be visualized by adapting the **gnuplot** script in the exercise_12 directory: pot_mean_force.gnu. Edit and adapt, then:
 +<code> gnuplot
 +gnuplot> load "pot_mean_force.gnu"
 +</code>
 +</note>
 +<note important>
 +Check the already finished runs in REF_300: 
 +<code>
 +cd REF_300
 +cp ../get_pot_mean_force .
 +./get_pot_mean_force
 +gnuplot
 +gnuplot > plot "pot_mean_force" w lp 
 +</code>
 +</note>
 +
 +<note warning>
 +  - Report adsorption (free) energy and equilibrium distance for the three temperatures. What do you note?
 +  - Compare the results with the paper
 +  - Compare your results with results extracted from REF_10, REF_300, REF_800 that contain longer equilibrations and averaging (about 0.5-1 ns each distance). Where do you get more differences? Why?
 +  - Can you say something about the entropy contributions?
 +</note>
  
exercises/2018_ethz_mmm/pmf.1527243542.txt.gz · Last modified: 2020/08/21 10:15 (external edit)