From: Massimiliano Porrini (M.Porrini_at_ed.ac.uk)
Date: Fri Mar 30 2012 - 08:20:33 CDT
I sent only the PGN and input PDB file to let you run psfgen and create the PSF,
in order to check if I am doing something wrong.
By the way, those pdbalias commands are needed exactly for the reasons
you explained:
the residue AGL does not exist in carbohydrate CHARMM FF and I have to map
it to the one I need, i.e. AGLCNA.
The same holds true for all the atom names in my input pdb file: I had to map
every single atom to those of carbohydrates CHARMM FF, just because in
the original
pdb the nomenclature was different.
However, the issue is about the residue name length; the procedure to
map and to create
the atoms names works fine.
Thanks again.
Max
Il 30 marzo 2012 13:32, Vincent Leroux <vincent.leroux_at_loria.fr> ha scritto:
> Well, where is the PSF file?
>
> There are a lot of pdbalias commands in your script - if you use a
> forcefield topology file that contains your fragment of interest you should
> not need to use those... Those commands are meant for allowing psfgen to
> understand PDB segments with residue names different from the forcefield
> definitions (e.g. histidines are 'HIS' in Brookhaven PDB format and CHARMM
> wants 'HSD'/'HSE'/'HSP' - for waters it is usually 'WAT' but CHARMM wants
> 'TIP3', etc.)
>
> So you are probably complicating matters uselessly.
>
> You can also try "regenerate angles dihedrals" before "guesscoord". That
> will not hurt.
>
>
>
>
> On 30/03/2012 14:06, Massimiliano Porrini wrote:
>>
>> Sorry, here are the files.
>>
>> Cheers,
>>
>>
>> Il 30 marzo 2012 13:03, Massimiliano Porrini<M.Porrini_at_ed.ac.uk> ha
>> scritto:
>>>
>>> Dear Vincent,
>>>
>>> I have tried with
>>>
>>> PSF NAMD
>>>
>>> and
>>>
>>> PSF NAMD CMAP
>>>
>>> and both did not work.
>>>
>>> I've also tried to load it from VMD command line and no luck again.
>>>
>>>
>>> Yes, with regard to the procedure, I upload the psf file first and
>>> then I upload the pdb file in it
>>> (without calling a New Molecule), but it does not work.
>>>
>>> If you are willing to have a go, I enclose my very small sugar
>>> molecule I am working with
>>> and the very small pgn file I am using, so that I can understand if I
>>> am making any error
>>> in the procedure (especially with the commands in the pgn file, even
>>> though I doubt,
>>> as the psfgen execution is smooth).
>>>
>>> Regards,
>>>
>>>
>>> Il 30 marzo 2012 12:40, Vincent Leroux<vincent.leroux_at_loria.fr> ha
>>> scritto:
>>>>
>>>> Regarding the "long atom types tag" the PSF 1st line should look like
>>>>
>>>> PSF NAMD
>>>>
>>>> without the ""
>>>>
>>>> and
>>>>
>>>> PSF NAMD CMAP
>>>>
>>>> if you use the CMAP version of CHARMM22
>>>>
>>>> If you use the VMD GUI you have to load the PSF as "New molecule" then
>>>> inject PDB data into it (not load the PDB as "New molecule" too). The
>>>> 1st
>>>> option of the "Molecule File Browser" is "Load files for:" - this should
>>>> be
>>>> set to the previously-loaded PSF when you load the PDB
>>>>
>>>> You can also try using the command from the VMD terminal:
>>>>
>>>> mol load psf yourpsf.psf pdb yourpdb.pdb
>>>>
>>>> If that does not work inspect your files carefully, because there is
>>>> probably something wrong with them...
>>>>
>>>> Regards
>>>> VL
>>>>
>>>>
>>>>
>>>> On 29/03/2012 21:05, Massimiliano Porrini wrote:
>>>>>
>>>>>
>>>>> Hi all,
>>>>>
>>>>> Thanks a lot for all the replies to my email.
>>>>> I have tried to upload the psf first and then the pdb, but no luck yet:
>>>>> still the structure is "messy".
>>>>>
>>>>>
>>>>> Dear Vincent,
>>>>>
>>>>> The names in the psf file have not been truncated and I have just
>>>>> downloaded and installed VMD-1.9.1, read the release notes and then I
>>>>> simply added
>>>>> the "NAMD" keyword in the first line, leaving a blank space just after
>>>>> the PSF keyword, like this:
>>>>>
>>>>> PSF "NAMD"
>>>>>
>>>>> But no luck again.
>>>>>
>>>>> I enclose also my input files (pdb and pgn), so that you might help me
>>>>> to see if I am doing something wrong here.
>>>>> To be more precise, I am using the psfgen version come out
>>>>> with NAMD_2.8_Linux-x86_64-multicore .
>>>>>
>>>>> Thanks a lot in advance.
>>>>>
>>>>> Regards,
>>>>>
>>>>>
>>>>> Il 29 marzo 2012 14:26, Vincent Leroux<vincent.leroux_at_loria.fr> ha
>>>>> scritto:
>>>>>>
>>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> Check your PSF file to see if the atom and residue names have not been
>>>>>> truncated by psfgen. If so, use VMD v1.9.1.
>>>>>>
>>>>>> If the PSF file appears ok, look for the "NAMD" keyword that should be
>>>>>> set
>>>>>> on the 1st line.
>>>>>>
>>>>>> See release notes here:
>>>>>> http://www.ks.uiuc.edu/Research/vmd/current/devel.html
>>>>>>
>>>>>> And of course, make sure you load the PSF file first in VMD and then
>>>>>> the
>>>>>> PDB
>>>>>> file on top of it. Loading PDB files directly that contain
>>>>>> non-peptidic
>>>>>> residue names and no CONECT records (psfgen do not generate those
>>>>>> explicit
>>>>>> bond definitions since they are in the PSF file already) is a sure way
>>>>>> to
>>>>>> get a lot of mess as a result.
>>>>>>
>>>>>> Regards
>>>>>> VL
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> On 29/03/2012 13:03, Massimiliano Porrini wrote:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Dear all,
>>>>>>>
>>>>>>> I am experiencing some problems with the CHARMM ff carbohydrates
>>>>>>> residues names in the PDB file created via psfgen.
>>>>>>>
>>>>>>> I defined some alias names in the pgn file in order
>>>>>>> to map the carbohydrates residue names to those used by CHARMM.
>>>>>>>
>>>>>>> The pdb (and psf) files are created smoothly so far (with no errors
>>>>>>> and warnings),
>>>>>>> but when I try to display the pdb with VMD I do not get the proper
>>>>>>> structure (VMD just shows an ensemble
>>>>>>> of scattered atoms). As far as I could understand this is because the
>>>>>>> name
>>>>>>> of the sugar units exceeds the field reserved to the residues name,
>>>>>>> which should be 3 characters
>>>>>>> long according to the PDB format, whilst some sugar names in CHARMM
>>>>>>> are
>>>>>>> longer (like for instance AGLCNA or AIDOA and so on).
>>>>>>>
>>>>>>> How could this issue get sorted out?
>>>>>>>
>>>>>>> As long as the oligosaccharide is relatively small I can do it
>>>>>>> manually (and I did it
>>>>>>> for a test case leaving just the first 3 characters of the residue
>>>>>>> names and VMD displayed
>>>>>>> properly the structure), but with a very long (and branched) sugar
>>>>>>> chain, it becomes harder.
>>>>>>>
>>>>>>> Any hint would be very appreciated.
>>>>>>>
>>>>>>> Regards,
>>>>>>>
>>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>
>>>
>>>
>>> --
>>> Dr Massimiliano Porrini
>>> Institute for Condensed Matter and Complex Systems
>>> School of Physics& Astronomy
>>>
>>> The University of Edinburgh
>>> James Clerk Maxwell Building
>>> The King's Buildings
>>> Mayfield Road
>>> Edinburgh EH9 3JZ
>>>
>>> Tel +44-(0)131-650-5229
>>>
>>> E-mails : M.Porrini_at_ed.ac.uk
>>> mozz76_at_gmail.com
>>> maxp_at_iesl.forth.gr
>>
>>
>>
>>
>
-- Dr Massimiliano Porrini Institute for Condensed Matter and Complex Systems School of Physics & Astronomy The University of Edinburgh James Clerk Maxwell Building The King's Buildings Mayfield Road Edinburgh EH9 3JZ Tel +44-(0)131-650-5229 E-mails : M.Porrini_at_ed.ac.uk mozz76_at_gmail.com maxp_at_iesl.forth.gr
This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:21:49 CST