Rotamerically Induced Perturbations

Applying RIP to the Zinc-Finger Mimic

As a first example, we apply RIP to a very small protein, the Zinc-Finger mimic 1FME, an artificial protein that folds into the topology of the zinc finger without the need for zinc.

Setting up the RIP Config File

This is found in the 'examples/zf-mimic' directory. In that directory is a CONFIG file, called 'zf-mimic.config'. It is a config file for a systematic perturbation of the zinc-finger mimic. It is essentially a PYTHON dictionary that will be read by the script PERTURB.PY, and consists of the following:

{ 'raw_pdb':            '1fme.pdb',
  'sim_dir':            '.', 
  'force_field':        'AMBER',
  'constraint':         '',
  'n_step_per_pulse':   100,
  'residues':           [],
  'heating_cases':      [('rip', 10, 26, 10000), ('rip', 300, 300, 10000)],
}

The parameters in the file are:

Running the RIP Simulations

To run, go to the 'pdbtool' directory:

python perturb.py ../examples/zf-mimic/zf-mimic.config

PERTURB.PY will process the RIP protocol encoded in zf-mimic.config, and attempt to run them using AMBER. This simulation took 11 hours on an iMac with a 2.4 GHz Intel processor. After the simulation is done, the .config file will be renamed by adding a '.done' at the end of the filename. The results are stored in

examples/zf-mimic/300k-rip-300k.

In this directory, there will be a series of directories numbered i=1 to 28, each representing the trajectory of a RIP perturbation on residue i in the protein.

I like to group all the files related to a single trajectory with the same basename but different extensions. Thus, all files associated with the trajectory start with 'md'. Analysis scripts can then just refer to 'md' and know where all the associated files are.

Analysis of Flexibility

Of course, you may have your own AMBER trajectory analysis scripts. But if you don't we have several analysis tools provided in the RIP scripts.

The main analysis tool is FLEXIBILITY.PY, if you're in the 'pdbtool' directory, to analyze the 10th (last) picosecond:

python flexibility.py ../examples/zf-mimic/300k-rip-300k 10

It expects the directory structure of the results produced by PERTURB.PY, and will calculate Cα-RMSD of each trajectory, and store these in a file called 'ca_rmsd' for each atom, and in 'ca_rmsd.ave' for each residue. These will be collated in a directory:

examples/zf-mimic/300k-rip-300k/ca_rmsd.10ps

In the analysis, a significant deviation is defined as a Cα-RMSD deviation greater than 6 Å. In this directory, you will find :

Analysis of Pathways

The main analysis tool for the coupling analysis is PATHWAY.PY, if you're in the 'pdbtool' directory, to analyze the 5th (last) picosecond of the slower simulations:

python flexibility.py ../examples/zf-mimic/10k-rip-26k 5

It expects the directory structure of the results produced by PERTURB.PY, and will calculate the kinetic energy 'kin' of each trajectory, and store these in a file called 'kin.ave' for each residue. These will be collated in a directory:

examples/zf-mimic/10k-rip-26k/kin.5ps

In this directory, you will find:

Examining the MD trajectories and PDB files

In order to look at trajectories and PDB files, I've developed two convenience scripts that are wrappers around PYMOL.

The first loads trajectories into PYMOL using only a single parameter that refers to the base name of the results of an AMBER trajectory. Furthermore, if a number is given is a second parameter, the script will highlight that residue for better display of the simulation. In this case, we have chosen the perturbation on Phe-12, which produces a very large conformational change:

 python pytraj.py ../examples/zf-mimic/300k-rip-300k/12/md 12

This will open a PYMOL window, where the trajectory can be played by clicking on the play button in the bottom right hand corner:

The second script uses PYMOL to display proteins with better default rendering. In the following example, we show the last snapshot of the RIP perturbation to Phe-12, with the residue highlighted in green "-h 12", using a white background "-g white",and displaying the rest of the protein in putty mode where the thickness and color of the backbone indicates the B-factor of the PDB file "-p". Incidentally the B-factor is set to the Cα-RMSD from the initial conformation of the trajectory.k:

 python showpdb.py -p -g white -h 12 ../examples/zf-mimic/300k-
 rip-300k/ca_rmsd.10ps/12.pdb

Both these helper scripts contain various display options that can be viewed by running the scripts with no command-line parameters. Remember to set the location of PYMOL within these scripts as described in the Pre-requisites section.

Generating a Visual Summary

If you've installed all the pre-requisite modules (PYMOL, MATPLOTLIB, MAKO), then you can run the analysis with the '-h' option that tells the program to generate images of the above files, and an HTML file that collates these images

python flexibility.py -h ../examples/zf-mimic/300k-rip-300k 10

This can be viewed through the HTML file examples/zf-mimic/300k-rip-300k/ca_rmsd.10ps/

which should look like this.

Similarly, for the pathway analysis:

python pathway.py -h ../examples/zf-mimic/10k-rip-26k 5

This can be viewed through the HTML file 

examples/zf-mimic/10k-rip-26k/kin.5ps/