Dear Francois, Marian
Thank you very much indeed for the advice Marian; I think Francois may have identified the problem I was having specifically though. I hope things are going well with you at the moment - I'll perhaps see you at the 1st year talks next month.
Francois thank you very much also for the exceptionally detailed answer. This has cleared numerous things up for me though I am slightly confused about one of the things you said:
> Error here: a CT1 atom name should not be present in the unique column
> of atom name of the PDB file format...
Why is this? I realise this is a stupid question, so apologies, but I'm not entirely sure I understand.
With best wishes and many thanks
Ben
On 6 Jan 2012, at 08:52, FyD wrote:
> Dear Ben,
>
> - In a given residue in a force field library two atoms cannot shares
> the same name (while chemical elements can be used as atom names when
> one only wants to display a PDB file in a graphical program).
> - In a force field library two equivalent atoms should bear the same
> charge value.
> - A P2N file is a PDB file with two columns of atom names; the first
> one reflecting chemical equivalencing (yellow column in the example
> below) and the second one (red column in the example below) containing
> the real atom names of the force field library to be buit. See an
> example with 2 conformations .
> http://q4md-forcefieldtools.org/Tutorial/Tutorial-1.php#EXAMPLE-P2N-FILE
>
> - Ante_R.E.D. generates P2N file(s) from PDB file(s), and R.E.D. uses
> P2N file(s) to generate mol2 file(s) (the force field library file
> format generated by R.E.D.).
>
> - Ante_R.E.D. 1.x from the R.E.D. tools at
> http://q4md-forcefieldtools.org/RED/ just uses a PDB file as input and
> generates a P2N file, but does NOT check for errors (duplicated atom
> names for instance) in the PDB input file. Ante_R.E.D. 1.x does not
> know how to deal with chemically equivalent atoms in different group
> of atoms. Thus, a user has to manually check the P2N file generated by
> Ante_R.E.D. 1.x. This is quite painful when the molecule is large...
> Ante_R.E.D. 2.0 interfaced by R.E.D. Server performs many checking
> and its goal is to correct all the errors in an input PDB file when
> one targets to build a force field library. Thus, Ante_R.E.D. 2.0
> corrects duplicated atom names and deals with chemical equivalencing.
> See http://q4md-forcefieldtools.org/REDS/news.php
> http://q4md-forcefieldtools.org/REDS/news.php#2
>
> You can also see a complex example of chemical equivalencing in the
> R.E.D. Server tutorial .
> http://q4md-forcefieldtools.org/Tutorial/Tutorial-3.php.
> (thanks to Alan Wilter for providing this molecule; P2N file generated
> by Ante_R.E.D. 2.0). See the Java applet:
> http://q4md-forcefieldtools.org/Tutorial/P2N/Complex-equiv/javaappletpdb-1.html
>
> My guess is that if you have a problem of atom names in your force
> field library you used Ante_R.E.D. 1.x. Try using Ante_R.E.D. 2.0 to
> generate your P2N file, or send us your PDB input file, P2N file and
> mol2 file.
>
> Several comments below in your email... I hope this helps.
>
> regards, Francois
>
> PS: You do not need to transform a mol2 file into OFF .lib file as
> both file formats have the same 'role' in LEaP: a force field library
> file format; you might also be interested by the .mol3 file format we
> are developing in LEaP;
> see http://q4md-forcefieldtools.org/Tutorial/leap-mol3.php
>
>
> Quoting Ben Ahmady <uccabha.ucl.ac.uk>:
>
>> My question - as per the subject title - relates in part to the R.E.D.
>> Server; however, as I am ultimately going to want to attempt to work
>> with AMBER 11/AmberTools 1.5 (specifically: Antechamber), I am posting
>> my question here rather than on the R.E.D. Server's list.
>>
>> Firstly, what I want to do is to take the RESP-derived charges I have
>> for ethyl phosphate and create a topology file for it.
>>
>> I have followed the guidelines set out by the R.E.D. team on their
>> website and have managed to obtain a mol2 file for ethyl phosphate
>> (etp.mol2). Running parmchk on this as follows ...
>>
>> parmchk -i etp.mol2 -f mol2 -o etp.frcmod
>
> - You can easily add force field atom types manually in the mol2 file
> format generated by R.E.D.
> - For ethylphosphate I do not think you need a frcmod file...
>
>> ... suggests (I looked at the frcmod file) that there are missing
>> parameters for 'MASS', 'BOND', 'ANGLE', 'DIHE', 'IMPROPER', and 'NONBON'.
>>
>> Using Ross Walker's tutorial
>> (http://ambermd.org/tutorials/basic/tutorial4b/) as my framework I ran
>> the following:
>>
>> tleap -f leaprc.ff99SB
>> source leaprc.gaff
>> ETP = loadmol2 etp.mol2
>>
>> ... so far so good. Moving on:
>>
>> check ETP
>>
>> ... this suggests that "there are missing parameters", which makes sense
>> given the output of parmchk and the contents of Ross Walker's tutorial.
>> Fine.
>
> Missing force field parameters for ethylphosphate? this is strange: I
> might be wrong but it looks like to me this is very unlikely...
>
> Did you load of force field parameter file such as parm99.dat in LEaP?
>
> Did you add the force field atom types in the mol2 file generated by R.E.D.?
> See http://q4md-forcefieldtools.org/REDDB/projects/F-90/
> http://q4md-forcefieldtools.org/REDDB/projects/F-90/script1.ff
>
> set P2M name "P2M"
> ...
> set P2M.1 restype undefined
> set P2M.1 name "P2M"
> # set FF atom types
> set P2M.1.OR type OS
> set P2M.1.P type P
> set P2M.1.OP1 type O2
> set P2M.1.OP2 type O2
> set P2M.1.OP3 type O2
> -> you need to define the FF atom types for the ethyl group...
> CT, HC, HC, HC, CT, H1, H1
>
>> Moving on:
>>
>> loadamberparams etp.frcmod
>> saveoff ETP etp.lib
>
> saveoff is useless because OFF and mol2 file formats are both a force
> field library file format.
>
>> saveamberparm ETP etp.prmtop etp.inpcrd
>>
>> Nothing obviously wrong so far. So then I went out of tleap and went
>> back in:
>>
>> tleap -f leaprc.ff99SB
>> source leaprc.gaff
>> loadamberparams etp.frcmod
>> loadoff etp.lib
>>
>> Then, I tried to run:
>>
>> pdb = loadpdb etp.pdb
>>
>> But was alerted to the fact that some of my atom names weren't unique in
>> 'etp.pdb', so I set about changing that according to convention (making
>> sure not to disturb any of the spacing); however, these didn't match
>> those in the P2N file which was generated by R.E.D since it is apparent
>> to me that the P2N accepts duplicate atom names on the basis of atom
>> equivalence and I am given to believe that the PDB does not (at least in
>> this context).
>
> See explanations above
>
>> In any case, after fixing the above I then ran all of the
>> above commands again leading up to the last one, which I modified
>> slightly (calling it 'ethylphosphate' as opposed to 'pdb'):
>>
>> ethylphosphate = loadpdb etp.pdb
>>
>> At first it seemed to be working and was giving output along the lines
>> of ...
>>
>> Created a new atom named: CT1 within residue: .R<ETP 0>
>
> Error here: a CT1 atom name should not be present in the unique column
> of atom name of the PDB file format...
>
>> Created a new atom named: O2 within residue: .R<ETP 0>
>> Created a new atom named: P3 within residue: .R<ETP 0>
>> Created a new atom named: O4 within residue: .R<ETP 0>
>>
>> ... but then it suddenly gave the following error message:
>>
>> Bond: Maximum coordination exceeded on .R<ETP 0>.A<H1 14>
>> -- setting atoms pert=true overrides default limits
>> ATOMS NOT BONDED: .R<ETP 0>.A<CT1 15> .R<ETP 0>.A<H1 14>
>> !FATAL ERROR----------------------------------------
>> !FATAL: In file [atom.c], line 444
>> !FATAL: Message: bondAtomProblem found
>> !
>> !ABORTING.
>>
>> ... I realise that other AMBER users have had similar sorts of problems
>> before (as evidenced by some e-mails in the mailing list archive) but
>> having read the responses I haven't been able to find anything that
>> actually helps me fix this problem, and as a new computational chemistry
>> PhD I don't have enough of an idea about how to implement 'atoms
>> pert=true'. As a result any help would be greatly appreciated.
>
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Jan 06 2012 - 12:30:03 PST
