Modification of the dihedral parameters

TO USE THE FUNCTION LIBRARY (VERSION UP TO DATE) IN THE INTERACTIVE SHELL: you@eulerX ~$ module load courses mmm ; mmm-init

Download the 2.3 exercise into your $HOME folder and unzip it.

you@eulerX ~$ wget http://www.cp2k.org/_media/exercises:2015_ethz_mmm:exercise_2.3.zip
you@eulerX ~$ unzip exercises:2015_ethz_mmm:exercise_2.3.zip
All files of this exercise ( input and scripts are all commented ) can be downloaded from the wiki: exercise_2.3.zip

Go to the directory “exercise_2.3/”

you@eulerX ~$ cd exercise_2.3
The relevant files are:
  1. For the non-restrained optimizations to get A and B configurations, inp.a and inp.b
  2. For the restrained optimization along a chain, ff_modify and inp_ff.templ, respectively the script to generate the “path” and the input file model for cp2k.
  3. For the potential with varying parameters for the Psi dihedral angle, pot_psi.templ, that will be used by ff_modify.

In this exercise, you are requested to start from the results of exercise 2, and perform the following steps

Choose two configurations A and B from the previously optimized grid (you can find them in the folder ~/exercise_2.2/Logs), close to the two minima. I suggest opt.1.3.pdb and opt.3.2.pdb.

Use m_pdbtorsion to measure the angles, but don't forget to load library in the memory first:
you@eulerX exercise_2.3$ . ~/Scripts/myfunctions.bash

To get help how to use the program simply type its name without any argument, and press “Enter”:

you@eulerX exercise_2.3$ m_pdbtorsion
Hint: Definition of PHI and PSI torsion angles for this particular system is the following…
PHI: 5 7 9 15
PSI: 7 9 15 17

The input file inp.a is similar to the one of exercise 2.2, but the “constraint” part was removed to perform a “free” geometry optimization. An important line is the initial configuration filename: ini.a.pdb Copy the opt.1.3.pdb into ini.a.pdb.

Run cp2k with:

you@eulerX exercise_2.3$ bsub cp2k.popt -i inp.a -o out.a 

The file a_opt-pos-1.pdb contains a row of configurations. To extract the last one:

you@eulerX exercise_2.3$ tail -25  a_opt-pos-1.pdb > amin.pdb

meaning that you get the “tail” of the file (last 25 lines) and you put these lines in the file amin.pdb. Check the final psi and phi angles using the script m_pdbtorsion, in the file amin.pdb. Note these angles on a piece of paper.

Do the same with inp.b (but now use opt.3.2.pdb as a starting point, not opt.1.3.pdb) , run cp2k in a similar way, and measure both torsion angles in the file bmin.pdb, that you may obtain using the “tail” command as before.

Check the final energies:
you@eulerX exercise_2.3$ grep 'E =' a_opt-pos-1.pdb
you@eulerX exercise_2.3$ grep 'E =' b_opt-pos-1.pdb

Is the energy becoming lower during the optimization?

Now copy the optimized “a” configuration into “ini.pdb”. Substitute the values of the angles in the ff_modify script:

PHI_A="some_value"
PHI_B="some_value"
PSI_A="some_value"
PSI_B="some_value"

and submit a new job:

you@eulerX exercise_2.3$ bsub < ff_modify

which will perform different jobs with the torsional term for the angle PSI modified by multiplication by 0.25, 0.5, 1, 2, 4. This corresponds to output lines enemul.* with three columns : the restrained phi, psi, and the energy in Hartree.

1 Hartree=27.2116 eV=627.509 kcal/mol

In this way you will obtain energy profiles joining the two minima

Could you expain an idea how to setup a nudged elastic band simulation to study the reaction pathway from A to the B point?
you@eulerX exercise_2.3$ gnuplot
gnuplot> load "mod_ff.gnu"
How will the line profile change? Why?