Many Thanks!
Now I'm working on accurate optimization of starting geometry, but it seems
that the main origin of error is a choice of QM region. I try to include in
QM part both ligand and fragment of N-terminal serine (that is a key residue
in catalysis). So I cut serine in following manner:
H
|
HO-CH2-CA- linkatom - CO-protein
|
NH2
Both OH- and neutral NH2- groups are crucial in catalysis. I know that it is
a bad choice to cut peptide bond, but cutting CA-C bond seems to be
dangerous too.
Is there any way to resolve the problem? And another question for the
future: is it reasonable to include in QM part residues, that act in
catalysis via backbone (because I have no idea how to cut QM part from the
rest of the protein in this case)?
On Wed, Mar 31, 2010 at 1:10 AM, Ross Walker <ross.rosswalker.co.uk> wrote:
> Hi Dmitry,
>
> > QMMM: WARNING!
> > QMMM: Unable to achieve self consistency to the tolerances specified
> > QMMM: No convergence in SCF after 1000 steps.
> > QMMM: Job will continue with unconverged SCF. Warning energies
> > QMMM: and forces for this step will not be accurate.
> > QMMM: E = -0.8502E+06 DeltaE = 0.1024E-07 DeltaP = 0.2569E-08
> > QMMM: Smallest DeltaE = 0.1013E-07 DeltaP = 0.2645E-08 Step =
>
> > Protein-ligand complex is based on crystallographic structure, thus, i
> > believe that starting conformation is good.
>
> If only that were true. Often crystal structures are very poor resolution,
> have errors and or guesses in certain parts of the structures and lack
> protons. Net result is that even a 'good' crystal structure can represent a
> bad starting geometry. I would start by followings Bud's advice. Minimize
> and run some MD first classically to let the system relax. Then try turning
> on the QM and see if the problem persists.
>
> > I've tried cut off 8, 9, 10; AM1 and PM3; SHAKE and NOSHAKE
> > minimization. Example of input file is attached.
>
> It is also possible that your QM region is a poor choice, perhaps there are
> link atoms across inappropriate bonds, cutting a peptide backbone for
> example. You may also have the charge for your ligand set incorrectly
> forcing it into some high energy state.
>
> Note, though that SCF got close to convergence:
>
> > QMMM: Smallest DeltaE = 0.1013E-07 DeltaP = 0.2645E-08 Step =
>
> The default limit is 1.0 x 10^-8 on energy and it got to 1.013 x 10^-8 so
> it
> was REAL close. You could try running the initial part of your heating /
> equilibration with scfconv=1.0e-7. This will not give as good forces but
> may
> be enough to get you away from a high energy starting structure and then
> you
> can tighten the convergence back up again later.
>
> Good luck,
> Ross
>
>
> /\
> \/
> |\oss Walker
>
> | Assistant Research Professor |
> | San Diego Supercomputer Center |
> | Tel: +1 858 822 0854 | EMail:- ross.rosswalker.co.uk |
> | http://www.rosswalker.co.uk | http://www.wmd-lab.org/ |
>
> Note: Electronic Mail is not secure, has no guarantee of delivery, may not
> be read every day, and should not be used for urgent or sensitive issues.
>
>
>
>
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Dmitry Nilov,
Lomonosov Moscow State University
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Mar 31 2010 - 01:30:04 PDT
