Re: Need help with psfgen : psfgen can't read file.pdb

From: Joao Ribeiro (jribeiro_at_ks.uiuc.edu)
Date: Tue May 07 2019 - 10:53:30 CDT

Dear Priyanka,

 

What are the errors that psfgen prints out when you try to build your system? Does psfgen recognize the residue “GRA” as you are naming it? Psfgen needs a topology for the system that is building, which the file that you are providing does not contain.

 

Unless you have any protein or lipid in your system, you should not use psfgen to prepare your structure.

 

Best

 

João

 

……………………………………………………...
João Vieira Ribeiro
Theoretical and Computational Biophysics Group
Beckman Institute, University of Illinois

http://www.ks.uiuc.edu/~jribeiro/

jribeiro_at_illinois.edu

+1 (217) 3005851

 

From: <owner-namd-l_at_ks.uiuc.edu> on behalf of Brian Radak <brian.radak_at_gmail.com>
Reply-To: <namd-l_at_ks.uiuc.edu>, Brian Radak <brian.radak_at_gmail.com>
Date: Tuesday, May 7, 2019 at 10:48 AM
To: Priyanka Mondal <pmondal_at_udel.edu>
Cc: namd-l <namd-l_at_ks.uiuc.edu>
Subject: Re: namd-l: Need help with psfgen : psfgen can't read file.pdb

 

As stated in my last email, this is not the correct psfgen syntax for reading a PSF with PDB coordinates -- the PDB should be included as an argument to the readpsf command (this is covered in the user guide):

 

readpsf zgnr_f2.psf pdb zgnr_f2.pdb ;# you can also read NAMD binary coordinates with the "namdbin" flag

 

Returning to my first question, I don't understand what you hope to accomplish by reading a PSF/PDB into psfgen and then writing them out again. Your psfgen script does not modify the input PSF/PDB in any way so there is no obvious value in doing this.

 

On Tue, May 7, 2019 at 11:05 AM Priyanka Mondal <pmondal_at_udel.edu> wrote:

Hello,

Thank you for your reply. I have file.xyz and I am trying to get file.pdb and file.psf from that and add water and ions using solvate and autoionize. Since file.xyz has info about coordinates but not bonds, I used topotools. So far when I loaded the output file into vmd window, everything looked right (atleast visually). Now when I tried adding water using "solvate" which uses psfgen with command

 

solvate file.psf file.pdb -minmax {{-20 -20 -20} {20 20 20}} -o file_sol

 

 I see that file_sol.pdb coordinates are all zero. To be sure that it is psfgen that's unable to read input files and hence causing the error, I wrote the small attached script.

 

package require psfgen
resetpsf
topology top_all27_prot_lipid.rtf
readpsf zgnr_f2.psf
coordpdb zgnr_f2.pdb
writepsf zgnr_f3.psf
writepdb zgnr_f3.pdb

 

which confirmed that psfgen is not being able to read file.pdb. Your help is much appreciated.

 

Priyanka Mondal,

Department of Physics and Astronomy,

University of Delaware,

USA

 

 

On Tue, May 7, 2019 at 9:16 AM Brian Radak <brian.radak_at_gmail.com> wrote:

I am not a topotools expert (in fact, not even a user), but I know psfgen pretty well -- what is this section of the script intended to accomplish? It doesn't do anything useful that I'm aware of.

 

package require psfgen
resetpsf
topology top_all27_prot_lipid.rtf
readpsf zgnr_f2.psf
coordpdb zgnr_f2.pdb
writepsf zgnr_f3.psf
writepdb zgnr_f3.pdb

 

Note that coordpdb is meant to read in coordinates for a segment and has a segment argument that is that segment name (or defaults to the last new segment, which is null here). I think what you want is instead:

 

readpsf zgnr_f2.psf pdb zgnr_f2.pdb

 

but again, I'm not sure why you would want/need to do this.

 

HTH,

BKR

 

 

On Mon, May 6, 2019 at 9:55 PM Priyanka Mondal <pmondal_at_udel.edu> wrote:

Hello all,

I am trying to create zgnr ribbon with dangling bonds passivated by H-atoms. I created ZGNR using nanotube builder and imported zgnr.xyz to quantumatk to add passivated H-atom and then imported final structure back to vmd as zgnr_H.xyz. Now I created zgnr_f.pdb and zgnr_f.psf using following command :

 

% **********************************************************************

mol delete all
mol load xyz zgnr_H.xyz
package require topotools
set sel [atomselect top all]
topo guessangles
topo guessdihedrals
topo guessimpropers
topo guessbonds
$sel writepsf zgnr_f.psf
$sel writepdb zgnr_f.pdb

mol delete all
mol load psf zgnr_f.psf pdb zgnr_f.pdb
% **********************************************************************

 

The files zgnr_f.psf and zgnr_f.pdb (just created) have fixed residue number 1 for all res GRA (which is different from residue number increasing in ascending order when I created zgnr.psf without H-atom directly from nanotube builder (i.e. no quantumatk) followed by a successful MD run).

 

% **********************************************************************

mol delete all
mol load psf zgnr_f.psf pdb zgnr_f.pdb
set sel [atomselect top all]
$sel set segid SHT
$sel set resname GRA
animate write psf zgnr_f2.psf waitfor all
animate write pdb zgnr_f2.pdb waitfor all

mol delete all
package require psfgen
resetpsf
topology top_all27_prot_lipid.rtf
readpsf zgnr_f2.psf
coordpdb zgnr_f2.pdb
writepsf zgnr_f3.psf
writepdb zgnr_f3.pdb

% **********************************************************************

 

Next when I try to use psfgen to read it using above commands, the output zgnr_f3.pdb sets all coordinates as zero. I looked up some similar questions and tried setting segname and resname as mentioned above but nothing worked. I am not sure how to fix it. I really hope that someone helps me solve this issue. I am not sure what to do. Thank you.

Priyanka Mondal,

Department of Physics and Astronomy,

University of Delaware,

USA

This archive was generated by hypermail 2.1.6 : Wed Dec 04 2019 - 23:20:46 CST