To Junmei and anyone else working on antechamber . . .
I've been following this thread because I'm working
on similar systems: protonated and unprotonated 2-amino-benzimidazoles.
Here are some molecules of interest and the problems antechamber has:
(by the way I did a hf/6-31* optimization to get starting structures)
2-amino-benzimidazole (attached mol2)
- I think antechamber gets atom types right, except the 2-amino should be pyramidal
and I think it will produce planar. So it might need different torsion parameters . . .
- if I try to produce charges antechamber crashes (see output). command:
$AMBERHOME/bin/antechamber -i 2-amino-benzimidazole.mol2 -fi mol2 -o initial.mol2 -fo mol2 -c bcc -nc 0
2-amino-1-methyl-benzimidazole (attached mol2)
- same atom typing problem as above. This time I can produces charges (using bcc).
2-amino-3H-1-methyl-benzimidazolium (attached mol2)
- antechamber gets C2 type wrong and also N2. (I'm numbering based on
conventional naming, not by the mol2 atom names.) Somehow antechamber gives
C2 atom type "cz" which isn't in gaff.dat. The resulting structure makes C2
pyramidal (ie 2-amino is out of plane with ring). Also, N2 is now sp3 rather sp2.
-Antechamber will produce charges without crashing.
I've hacked together some parameters that make it work, but these examples
might be of interest for improving antechamber.
--Niel
________________________________________
From: amber-bounces.ambermd.org [amber-bounces.ambermd.org] On Behalf Of Gabriel Rocklin [gabriel.rocklin.ucsf.edu]
Sent: Friday, February 19, 2010 4:51 PM
To: amber.ambermd.org
Subject: Re: [AMBER] Re: Antechamber: imidazole=aromatic, imidazolium=not aromatic?
Here's the file Junmei sent me. You'd have to ask him how it's going to be
distributed.
Gabriel
On Fri, Feb 19, 2010 at 3:48 PM, Gabriel Rocklin <grocklin.gmail.com> wrote:
> Hi there,
>
> I may have skipped part of this conversation, but where is the attached
> file? or at least the added lines. And if this fix will be available as a
> patch in AT 1.3 soon or later.
>
> Thanks,
>
> Alan
>
> On Thu, Feb 18, 2010 at 23:21, Gabriel Rocklin <gabriel.rocklin.ucsf.edu
> >wrote:
>
> *> Hi Junmei, *
> *> *
> *> This seems to have fixed the problem. Thanks! *
> *> *
> *> Gabriel *
> *> *
> *> On Thu, Feb 18, 2010 at 1:48 PM, Junmei Wang < *
> *> Junmei.Wang.utsouthwestern.edu *
> *> > wrote: *
> *> *
> *> > Hi, Gabriel *
> *> > I am sorry for not providing a quick solution for your problem. I
> found *
> *> the *
> *> > best solution is not to do coding, but to modify the atom type
> definition *
> *> > file. I simply added two more lines to ATOMTYPE_GFF.DEF. Please see
> the *
> *> > attached file. Please let me know if the definition file works for
> you. *
> *> > *
> *> > Best *
> *> > *
> *> > Junmei *
> *> > *
> *> > *
> *> > *
> *> > *
> *> > *
> *> > >>> Gabriel Rocklin <gabriel.rocklin.ucsf.edu> 02/15/10 5:22 PM >>> *
> *> > Hi Junmei, *
> *> > *
> *> > Just wondering how the Antechamber fix is going, and if you've
> learned *
> *> > anything now about the scope of the problem. *
> *> > *
> *> > Thanks, *
> *> > Gabriel *
> *> > *
> *> > On Wed, Feb 10, 2010 at 9:24 AM, Gabriel Rocklin *
> *> > <gabriel.rocklin.ucsf.edu>wrote: *
> *> > *
> *> > > Thanks Junmei. The parameters do more than maintain the planar *
> *> > > conformation of the ring - they keep the ring substituents in
> plane *
> *> with *
> *> > the *
> *> > > rest of the ring. I'm not actually interested in imidazole and *
> *> > imidazolium *
> *> > > themselves, but in their substituted versions, and binding free *
> *> energies *
> *> > > appear to change significantly when the ring substituents are
> allowed *
> *> to *
> *> > > bend when in protein complexes. *
> *> > > *
> *> > > Hope to hear from you by Monday. *
> *> > > *
> *> > > Gabriel *
> *> > > *
> *> > > *
> *> > > On Wed, Feb 10, 2010 at 8:07 AM, Junmei Wang < *
> *> > > Junmei.Wang.utsouthwestern.edu> wrote: *
> *> > > *
> *> > >> Dear Gabriel, *
> *> > >> I will fix this no later than next Monday. Did you observe
> bizarre *
> *> > >> behavior of gaff for imidazolium? It is true that the torsional
> angle *
> *> > >> parameter of X-c2-na-X (2.5) is weaker than X-cc-na-X (6.8), but
> those *
> *> > >> parameters are simply used to maintain the planar conformation of
> the *
> *> > ring. *
> *> > >> The minimization or MD results of the two set of parameters should
> be *
> *> > very *
> *> > >> similar anyway. *
> *> > >> *
> *> > >> Best *
> *> > >> *
> *> > >> Junmei *
> *> > >> *
> *> > >> >>> Gabriel Rocklin <gabriel.rocklin.ucsf.edu> 02/10/10 12:56 AM
> >>> *
> *> > >> Dear Junmei, *
> *> > >> *
> *> > >> Do you have an estimate on how long that will take? I'd like to
> know *
> *> > >> whether I should try to develop a workaround, or whether I can
> wait *
> *> for *
> *> > >> you. *
> *> > >> Also, it would be good to know what the scope of the problem is
> as *
> *> soon *
> *> > >> as *
> *> > >> you've identified it, even if you haven't yet designed a fix. *
> *> > >> *
> *> > >> Thanks, *
> *> > >> Gabriel *
> *> > >> *
> *> > >> *
> *> > >> *
> *> > >> On Tue, Feb 9, 2010 at 9:07 PM, Junmei Wang *
> *> > >> <Junmei.Wang.utsouthwestern.edu>wrote: *
> *> > >> *
> *> > >> > Dear Gabriel *
> *> > >> > I got the same result as yours. It is not a correct antechamber
> *
> *> > >> behavior. I *
> *> > >> > will fix this bug as soon as possible. Thank you very much for *
> *> > reporting *
> *> > >> > this problem *
> *> > >> > *
> *> > >> > I really appreciate your great help *
> *> > >> > *
> *> > >> > Junmei *
> *> > >> > *
> *> > >> > *
> *> > >> > >>> Gabriel Rocklin <gabriel.rocklin.ucsf.edu> 02/09/10 1:33 AM
> >>> *
> *> > >> > Dear Junmei and others, *
> *> > >> > *
> *> > >> > I am getting strange results when trying to parameterize some *
> *> > >> heterocyclic *
> *> > >> > cations using antechamber. Using a mol2 file for imidazole, when
> I *
> *> > do- *
> *> > >> > *
> *> > >> > antechamber -i imidazole.mol2 -o imidazole_out.prepi -fi mol2
> -fo *
> *> > prepi *
> *> > >> > *
> *> > >> > I get a prep file for imidazole that has ring atom types cc, cd,
> and *
> *> > na. *
> *> > >> > But when I do the same thing for imidazolium, I get atom types
> c2 *
> *> and *
> *> > >> na. *
> *> > >> > Because GAFF c2 is not aromatic, the dihedral parameters about
> the *
> *> > ring *
> *> > >> > become significantly weaker. The protonation should not affect
> the *
> *> > >> > aromaticity like this, correct? I see the same thing *
> *> > >> > for constitutively charged, n-substituted imidazoliums, as well
> as *
> *> for *
> *> > >> > thiazoles vs thiazoliums. *
> *> > >> > *
> *> > >> > Is this the correct antechamber / GAFF behavior? If so, why? *
> *> > >> > *
> *> > >> > I've attached the input and output files. *
> *> > >> > *
> *> > >> > Thanks, *
> *> > >> > Gabriel *
> *> > >> > *
> *> > >> > Gabriel Rocklin *
> *> > >> > UCSF Shoichet/Dill Labs *
> *> > >> > *
> *> > >> > *
> *> > >> *
> *> > >> *
> *> > > *
> *> > *
> *> > *
> *> _______________________________________________ *
> *> 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 Wed Feb 24 2010 - 18:00:02 PST
