Dear Francesco,
> Thanks for your kind offer. Attached please find the pdbs (ending NXT
> or not) and the mol2 from the NXT. I used antechamber 1.2 fully
> patched and gcc compiled on amd64.
I have loaded your mol2 file in xLEaP without any problem (64 bits
Linux Redhat based operating systems).
I would upgrade to the last version of the AmberTools.
Your molecule is quite simple: You could try modifying your mol2 file
to understand what LEaP does not like in this file.
- Use babel or Ante_R.E.D. 1.x to generate another set of atom connectivities
- Simplify the atom names (& avoid NXT & HXT atom names)
- Use a basic mol2 file such as:
http://q4md-forcefieldtools.org/REDDB/project/W-1/tripos1.mol2
to look what could be wrong/you could change...
replace ????
25 25 1 0 0
by
25 25 1 0 1
replace ???
1 PHA 1 TEMP 0 **** **** 0 ROOT
by
1 PHA 1 **** 0 **** ****
- look at character at the end of each line, check that your file ends
with a return line etc...
regards, Francois
> On Wed, Oct 6, 2010 at 11:29 AM, FyD <fyd.q4md-forcefieldtools.org> wrote:
>> Francesco,
>>
>>> I forgot: also, even the mol2 file generated by antechamber is
>>> accepted by Chimera, extrapolating what I said below.
>>
>> So ?
>>
>> This is not because Antechamber generates a mol2 file that is
>> recognized by Chimera, (VMD, insightII... whatever) that this mol2
>> file will indeed be recognized by LEaP... There are bugs everywhere
>> here - in all the programs (for instance VMD (64 bits) used to not
>> handle mol2 files composed by a single atom).
>>
>> Can you post your mol2 file so that we can try to "finish" this problem ?
>>
>> regards, Francois
>>
>>> ---------- Forwarded message ----------
>>> From: Francesco Pietra <chiendarret.gmail.com>
>>> Date: Wed, Oct 6, 2010 at 10:21 AM
>>> Subject: Re: [AMBER] Incorrect handling of phenylalanine amide by
>>> antechamber
>>> To: AMBER Mailing List <amber.ambermd.org>
>>>
>>>
>>> I agree, however the starting pdb has no problems with even Chimera.
>>> In my experience, any but slight deviation from standard will not be
>>> accepted by Chimera, which follows a cliche of standard for amino
>>> acids, peptides and proteins (while VMD accepts nearly every deviation
>>> from standard). Also, I have taken the information for the CONH2 from
>>> amino acids that have such functionality in the side chain. What
>>> remains, is the standard for phenylalanine. That to explain why at the
>>> moment I have no better ideas.
>>>
>>> My further attempts to generate prmtop/inpcrd file is that I have my
>>> proven program for simulated annealing (i.e., conformational search)
>>> which only accepts such an input. Otherwise I had already turned to
>>> other approaches to the conformational problem of the peptide that
>>> contains Fa.
>>>
>>> thanks
>>> francesco
>>>
>>> On Wed, Oct 6, 2010 at 8:19 AM, FyD <fyd.q4md-forcefieldtools.org> wrote:
>>>> Dear Francesco,
>>>>
>>>>> By using exactly the same procedure below, I started by renaming atoms
>>>>> of NH2 (of CONH2) NXT HXT1 HXT2. Same error in prmtop/inpcrd: ring and
>>>>> NH2 of CONH2 highly deformed. Notice that, again, residue PHA in the
>>>>> pdb file had to be renumbered from 4 to 1, otherwise xleap crashes
>>>>> when attempting to load the mol2 file obtained with antechamber.
>>>>
>>>> When Antechamber and the R.E.D. Tools have to build a FF library a
>>>> common problem is ... the input file, itself, provided by the user -
>>>> and even by a novice user. In a FF lib, the topology (usually defined
>>>> from a correct input geometry), atom & residue names (two atom names
>>>> in a residue cannot share the same name) and atom equivalencing (used
>>>> in charge equivalencing) have to be correctly defined. As a new user
>>>> does not have any idea about these rules, the PDB file she/he provides
>>>> as input is often "broken" or does not follow a standard format. Thus,
>>>> Antechamber and the R.E.D. Tools have to deals with these problems and
>>>> the number of problems to handle is quite a challenge.
>>>>
>>>> To solve these numerous problems one has two options:
>>>> (1) One tries to make the program compatible with the numerous inputs
>>>> provided and one generates the FF library following the information
>>>> available in the input. I think this is the approach followed by
>>>> Antechamber. A problem here is that a FF library does not always have
>>>> a strict format. Consequently, the file might be recognized by a
>>>> program such as VMD and not by another one such as LEaP...
>>>> (2) One corrects the input provided so that one generates a FF library
>>>> in a strict format which in principle is compatible with all programs.
>>>> This is the approach followed in Ante_R.E.D. 2.0/R.E.D. IV. This will
>>>> be available in R.E.D. Server 2.0. We are late and it will be released
>>>> soon.
>>>>
>>>> regards, Francois
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Oct 06 2010 - 05:00:03 PDT
