Hi Francesco,
I have read back through all of your entries regarding this issue, and
unless I'm missing something, it appears you either want to model
model an isolated phenylalanine amide (e.g., +H3N-Phe-NH2) or a
peptide containing a C-terminal phenylalanine amide (e.g., rest of
peptide-Phe-NH2). If either of these is correct, then there is no
need to be messing around with antechamber, GAFF, or the like.
(Granted you may still want to learn about building new residues for
AMBER, but you do *not* need to build a new residue for this
particular application.) There are two straightforward ways to proceed:
(1) If you already have a PDB file for the isolated peptide, just make
sure that the atoms of the NH2 group have a different residue number
-- specifically one number greater than that of the phenylalanine --
and that these atoms belong to a residue called NHE. (See the
attached PDB files for an example of a 10 residue peptide with a
valine amide -- pneumadin.pdb -- and an isolated phenylalanine amide
-- pha.pdb. In both cases, you'll see an "extra" residue called NHE
for the NH2 group.) Then you should be able to load your sequence
straight into leap as follows:
source leaprc.ff99SB
pep = loadpdb peptide.pdb
... do what you want here and save out files ...
quit
(2) If you are building this residue from scratch (i.e., you don't
have a specific starting structure), then just input the sequence
manually and build it. I have attached two leap setup scripts that
show this sort of implementation (and generated the PDB files for
option 1):
source leaprc.ff99SB
pep = sequence { NPHE NHE }
... do what you want here and save out files ...
quit
Hopefully this is what you're after, but if not, apologies for the
diversion...
Best,
Paul
P.S. Because the structures in the PDB files haven't been minimized,
there are some close contacts and thus VMD might render some atoms as
being bonded when they aren't supposed to be.
On Oct 7, 2010, at 9:14 AM, Paul S. Nerenberg wrote:
> Francesco,
>
> Actually, I'm still a bit confused about what you want to do here. Is
> this the same C-terminal phenylalanine amide residue that you were
> looking at with another program several weeks ago? If so, then there
> is no reason to be jumping through any of these (increasingly
> elaborate) hoops. The AMBER ff99SB force field already includes a
> definition for the amidated C-terminus (NHE). You just need your
> sequence (either in a PDB file or manually in leap) to end with PHE
> NHE -- this will build phenylalanine as the last residue with an NH2
> "residue" attached to its C-terminus.
>
> Best,
>
> Paul
>
>
> On Oct 7, 2010, at 8:08 AM, Francesco Pietra wrote:
>
>> Dear Francois:
>> I forgot the main question following your discovery that ff99SB
>> should
>> not be loaded to leap in my case. That is, do you believe that the
>> gaff-only route is a viable route to prepare the new residue
>> phenylalanine amide? I plan to add the new residue to the standard
>> residues in an Amber ff (ideally ff99SB) so that my peptide
>> containing
>> the new residue is recognized by leap.
>>
>> thanks
>> francesco
>>
>>
>> ---------- Forwarded message ----------
>> From: Francesco Pietra <chiendarret.gmail.com>
>> Date: Thu, Oct 7, 2010 at 4:57 PM
>> Subject: Re: [AMBER] Incorrect handling of phenylalanine amide by
>> antechamber
>> To: AMBER Mailing List <amber.ambermd.org>
>>
>>
>> Dear Francois:
>> Thanks a lot. Plase see below.
>>
>> On Thu, Oct 7, 2010 at 11:56 AM, FyD <fyd.q4md-forcefieldtools.org>
>> wrote:
>>> Dear Francesco,
>>>
>>> I tested your two sets of data & this is exactly what I told you in
>>> my
>>> first email: Antechamber authorizes non-strict mol2 file format. I
>>> think this is potentially buggy if one wants to use the outputs of
>>> Antechamber in other programs.
>>
>> Perhaps worth knowing that when I said that this is the first time
>> (out of very many) that antechamber outputs are not perfect for leap,
>> the "very many" were always for organic natural products, or their
>> modification. This is the first time I used Antechamber for a
>> modified
>> amino acids. See below comments about not loading ff99SB.
>>>
>>> When your mol2 file format has a residue number = 4, LEaP crashes
>>> (because a patch was not applied; I think you did not applied the
>>> patch suggested by Dr Case).
>>
>> As I wrote, I applied all current bugfixes (1-35) and recompiled
>> ambertools/gcc and amber/gfortran (previously I had only applied
>> bugfixes 1-5 and compiled all intel/mkl. Still. leap crashed when
>> PHA#4.
>>>
>>> When your mol2 file has a residue number = 1, LEaP does not crash. I
>>> only loaded gaff and the geometry of PHA is correct (correct in
>>> LEaP,
>>> in VMD using the prmtop/inpcrd generated, correct when using
>>> ambpdb &
>>> the prmtop/inpcrd generated).
>>>
>>> I did not load amber99SB & then gaff - if you do so and if the
>>> geometry looks broken this is just another bug or an incompatibility
>>> between the leaprc of amber99SB and gaff.
>>
>> I had suggested in one in my mails if it is worth trying to load gaff
>> alone. As I had no reply, I thought it was a bad idea (it is common
>> knowledge that gaff was made to be compatible with amber ffs) and did
>> not try.
>>>
>>> regards, Francois
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
- application/octet-stream attachment: setup_pha
Received on Thu Oct 07 2010 - 10:30:03 PDT
