VMD-L Mailing List

From: Edward Lyman (edward.lyman_at_gmail.com)
Date: Sat Apr 30 2011 - 10:39:47 CDT

Hi all,

Jim suggested using the 2.8b1 psfgen with readmol psf. I tried:

readmol psf 50_50_10nm.xplor.psf

but it segfaults:

Info: reading file 50_50_10nm.xplor.psf using plugin psf
WARNING: Plugin-based I/O is still in development and may still have bugs
psfplugin) Detected a Charmm31 PSF file
psfplugin) Detected a Charmm31 PSF EXTEnded file
Data fields found by plugin:
bonds: 74520
psfplugin) no cross-terms defined in PSF file.
angles: 86580 dihedrals: 110700 impropers: 360 cross-terms: 0
Segmentation fault

Version 1, writing the psf from the vmd console gets the bonding topology
right, but does not write the angles and dihedrals correctly, as Axel just
explained.

So, I am looking for some new ideas to get either read a psf with the
extended format, or else convert an extended format psf into a regular
format psf. Any thought are very much appreciated, as is all the help up to
this point.

Ed

On Sat, Apr 30, 2011 at 11:16 AM, Edward Lyman <edward.lyman_at_gmail.com>wrote:

> Ah, I see. Thanks Axel, I will revert to the psfgen method then.
>
> E
>
>
>
> On Fri, Apr 29, 2011 at 4:11 PM, Axel Kohlmeyer <akohlmey_at_gmail.com>wrote:
>
>> On Fri, Apr 29, 2011 at 3:30 PM, Edward Lyman <edward.lyman_at_gmail.com>
>> wrote:
>> > Thanks for the reply. Now I have used vmd 1.9 to delete a single
>> cholesterol
>> > by
>> >
>> > (1) read in psf and pdb for the full system
>> > (2) define an atom selection that excludes the problematic cholesterol:
>> >> set write_sel [atomselect top "not same fragment as index foo"]
>> > (3) renumber residues
>> >> $write_sel set resid [$write_sel get residue]
>> > (4) write psf/pdb
>> > > $write_sel writepsf new.psf
>> > > $write_sel writepdb new.pdb
>> >
>> > This writes a psf that looks ok through the bonds section, but the
>> angles,
>> > dihedrals, impropers have problems. There are the same number of angles
>> in
>> > the new as in the original psf. (Theres hould be fewer because there is
>> one
>> > less chol) And, the angles include triplets like "1 2 2 " in the first
>> 50 or
>> > so lines, then the angle lines are all "1 1 1 1 1 1 1 1 ". same for
>> > dihedrals, etc.
>> >
>> > So, am I missing an autogenerate statement? It seems possible that
>> > autogenerate may not get the angles, dihedrals etc correct...is that a
>> real
>> > concern?
>>
>> ed,
>>
>> you are touching a real problem. i didn't think about this earlier.
>> the bond information is attached to the atom information and thus
>> will be adjusted accordingly. however, this is not true for angles and
>> dihedrals or impropers. that information is stored separately, and
>> then not subjected to any selection operations.
>>
>> i guess, that may be the reason for people to keep using psfgen
>> for these tasks. while not a VMD intrinsic, you could loading the
>> psf/pdb combo with the atoms deleted and then use the "guess" commands
>> of the topotools plugin to redo the angle, dihedral and improper
>> definitions.
>>
>>
>> http://sites.google.com/site/akohlmey/software/topotools/topotools---documentation#TOC-guess-angle-dihedral-s
>>
>> however, there is one caveat. this will generate this information from
>> the bond topology alone, and thus may not always be the same
>> as what psfgen will generate, particularly for impropers. you might
>> need to run some tests on this. i did some successful tests on
>> simple peptides, but never for a large protein.
>>
>> sorry for overlooking the angles stuff. that was my bad.
>>
>> axel.
>>
>> >
>> > Thanks,
>> > Ed
>> >
>> >
>> > On Fri, Apr 29, 2011 at 9:40 AM, Axel Kohlmeyer <akohlmey_at_gmail.com>
>> wrote:
>> >>
>> >> On Fri, Apr 29, 2011 at 9:00 AM, Edward Lyman <edward.lyman_at_gmail.com>
>> >> wrote:
>> >> > Thanks for the work around, it's very much appreciated. But, it
>> occurs
>> >> > to me
>> >> > now...why can't I simply use $sel writepsf blah.psf and $sel writepdb
>> >> > blah.pdb from the vmd console to write out a psf and a pdb for
>> >> > selections
>> >> > which delete the unwanted molecules? ie, why bother with psfgen at
>> all
>> >> > when
>> >> > you already have in hand a psf that contains all the correct force
>> field
>> >> > terms, etc, and you just want to delete a few molecules?
>> >>
>> >> you can do this. the functionality to store and write out the complete
>> >> topology
>> >> information from and to .psf files has not been around that long and
>> >> before that
>> >> using psfgen was the only option. not everybody is aware of this new
>> >> functionality
>> >> in VMD and various plugins and scripts are only slowly being migrated
>> to
>> >> make
>> >> use of these new features (why break something that works?).
>> >>
>> >> cheers,
>> >> axel.
>> >>
>> >> >
>> >> > Thx,
>> >> > Ed
>> >> >
>> >> >
>> >> >
>> >> >
>> >> > On Tue, Apr 26, 2011 at 3:21 PM, Jim Phillips <jim_at_ks.uiuc.edu>
>> wrote:
>> >> >>
>> >> >> If you can't get CHARMM-GUI to write a non-EXT, non-CHEQ format PSF
>> (I
>> >> >> don't think you need either of those options) then try reading the
>> file
>> >> >> into
>> >> >> VMD, writing it out again, and using that file with psfgen.
>> >> >>
>> >> >> If you use the psfgen binary from NAMD 2.8b1 you can probably read
>> the
>> >> >> file with "readmol psf 50_50_10nm.xplor.psf" since that uses the VMD
>> >> >> plugins
>> >> >> (I don't think VMD's psfgen module has plugins enabled yet).
>> >> >>
>> >> >> -Jim
>> >> >>
>> >> >>
>> >> >> On Tue, 26 Apr 2011, Edward Lyman wrote:
>> >> >>
>> >> >>> attached is the psf.
>> >> >>>
>> >> >>> Any suggestions for a workaround? I want to delete a couple lipid
>> >> >>> molecules
>> >> >>> as described in the psfgen docs...if there is some othe way to
>> >> >>> accomplish
>> >> >>> this, I am all ears.
>> >> >>>
>> >> >>> E
>> >> >>>
>> >> >>>
>> >> >>> On Tue, Apr 26, 2011 at 11:53 AM, John Stone <johns_at_ks.uiuc.edu>
>> >> >>> wrote:
>> >> >>>
>> >> >>>> Hi,
>> >> >>>> It seems that psfgen doesn't like the X-PLOR format PSF file
>> you've
>> >> >>>> got,
>> >> >>>> although the VMD PSF plugin seems to read it without trouble.
>> >> >>>>
>> >> >>>> Can you send me a gzipped copy of the PSF file that is causing you
>> >> >>>> trouble?
>> >> >>>>
>> >> >>>> This is most likely a limitation of the PSF reader currently in
>> >> >>>> psfgen,
>> >> >>>> and is something we are looking at improving.
>> >> >>>>
>> >> >>>> Cheers,
>> >> >>>> John Stone
>> >> >>>> vmd_at_ks.uiuc.edu
>> >> >>>>
>> >> >>>> On Fri, Apr 22, 2011 at 11:03:30AM -0400, Edward Lyman wrote:
>> >> >>>>>
>> >> >>>>> Hi all,
>> >> >>>>>
>> >> >>>>> I am trying to use psfgen to delete a couple of lipids and
>> write a
>> >> >>>>> new
>> >> >>>>> psf/pdb pair. The toplogy files read in fine, but when I read
>> teh
>> >> >>>>> psf,
>> >> >>>>
>> >> >>>> I
>> >> >>>>>
>> >> >>>>> get:
>> >> >>>>>
>> >> >>>>> reading structure from psf file 50_50_10nm.xplor.psf
>> >> >>>>> Error processing bonds
>> >> >>>>>
>> >> >>>>> MOLECULE DESTROYED BY FATAL ERROR! Use resetpsf to start over.
>> >> >>>>> psfplugin) Detected a Charmm31 PSF file
>> >> >>>>> psfplugin) Detected a Charmm31 PSF EXTEnded file
>> >> >>>>> Info) Using plugin psf for structure file 50_50_10nm.xplor.psf
>> >> >>>>> psfplugin) no cross-terms defined in PSF file.
>> >> >>>>> Info) Analyzing structure ...
>> >> >>>>>
>> >> >>>>> I am running vmd 1.9, psfgen 1.5. The psf file is an xplor psf
>> >> >>>>> (runs
>> >> >>>>
>> >> >>>> in
>> >> >>>>>
>> >> >>>>> namd) that was written by the charmm gui. The psf/pdb pair work
>> >> >>>>> just
>> >> >>>>
>> >> >>>> fine
>> >> >>>>>
>> >> >>>>> for reading into vmd.
>> >> >>>>>
>> >> >>>>> Any help is appreciated.
>> >> >>>>> Ed
>> >> >>>>
>> >> >>>> --
>> >> >>>> NIH Resource for Macromolecular Modeling and Bioinformatics
>> >> >>>> Beckman Institute for Advanced Science and Technology
>> >> >>>> University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
>> >> >>>> http://www.ks.uiuc.edu/~johns/ Phone: 217-244-3349
>> >> >>>> http://www.ks.uiuc.edu/Research/vmd/ Fax: 217-244-6078
>> >> >>>>
>> >> >>>
>> >> >
>> >> >
>> >>
>> >>
>> >>
>> >> --
>> >> Dr. Axel Kohlmeyer
>> >> akohlmey_at_gmail.com http://goo.gl/1wk0
>> >>
>> >> Institute for Computational Molecular Science
>> >> Temple University, Philadelphia PA, USA.
>> >
>> >
>>
>>
>>
>> --
>> Dr. Axel Kohlmeyer
>> akohlmey_at_gmail.com http://goo.gl/1wk0
>>
>> Institute for Computational Molecular Science
>> Temple University, Philadelphia PA, USA.
>>
>
>