PSF file generation

Charging the system

Update: March, 2018

A better version of creating the psf file has been discussed here. Please avoid this page and follow that process.

VegaZZ

Pdb files contain only the atom coordinates (and some citation references). However, to apply the bonded interactions in MD simulation, we need to have the information regarding bonds, angles, dihedrals and impropers. This information is contained in PSF files, they also contain the charges to apply the electrostatic interactions. We will generate psf file using Vega ZZ (another open source software, we don't wanna pay for trifle things, right?).

Wide Pdb

A side note here, though it might not be required for our case, we will create a wide pdb file. When we convert pdb to psf, vegazz will use a distance criteria to determine which atom is connected to which other atom. This means that if packmol packed two ethanol molecules too closely (we specified a tolerance of 2 A, so we are safe) then VegaZZ will connect the two molecules and portray it as one. Computers are egoistic and don't care about what to do, they just do what they can. It can sometimes cause error. To avoid this, people create another pdb file with really large box dimenions. This ensures that no two molecules are too close. As such there won't be any "Bond Order too high" error in VegaZZ. (Be very cautious every time you see this warning in VegaZZ). Follow these steps to create the wide pdb file:

This is the packmol script file, save it as mixture_wide.txt
Notice how it is exactly the same as the previous file, except that the box dimensions have changed
Run using packmol as before, a final_wide.pdb will be created

Generating PSF

Follow these steps:

Open the final_wide.pdb in Vega ZZ, make sure it doesn't show any error like 'Bond Order too high',etc.
Go to File-> Save As, save this file as .psf
This will create a final_wide.psf file

Adding Charges

Have a look at the psf file (open it in Notepad++). Do you notice the plethora of question marks in the file? Actually, they are meant to specify the charge of the respective atoms. We obtained the charges from Gaussian calculations (in case you remember), we need to copy those charges in a separate file (in case you haven't till now).

Charges from Gaussian09: (copy ESP charges if you specified the pop=chelpg keyword!)

charges

Next we need to insert these charges into the psf file. Also, we will also give an atomtype to each of these atoms. Why? We are trying to think a step ahead. Our next step will be creating the param file or the force field file. At times it occurs that many different atoms are quiet similar to each other and as such they can be specified the same atomtype. This will save computational time and energy. As such we will give the same atomtypes to atoms that have the same kind of parameters. How do we know if two atoms have the same parameters? Experience helps. Also, you can check the forcefield file first, then edit this file and run it again. For now, just follow these steps:

Create a new folder 'charges'
Copy the final_wide.pdb and final_wide.psf files to this folder
This is the main tcl script, save it as charge.tcl
Create a new text file and save it as 'ethaAtomType.dat'; this file contains the atomtype names
Create another text file and save it as 'ethaCharge.dat'; copy the charges into this file
Open vmd and execute the tcl script in tk console (you know how to do it, cmon, reformatting pdb, right?)

A file named final.psf will be created. This is your final psf file. Viola!!! Again, if you are tired, well then its the appropriate time to quit this field of research, and instead start preparing for Reliance, Jindal or even IOCL. In case you aren't, move on to the next section.

Previous Page Main Page Next Page