Re: [AMBER] What if MM Energy > QM Energy Even Vn=0

From: David Cerutti <>
Date: Mon, 13 Dec 2021 11:06:46 -0500

Sorry I need to check my [AMBER] forum box more frequently. From the looks
of this, the MM energy has a lot of parameters in it and should track the
QM energy, if only in a mean sense, perhaps wildly vacillating around the
QM curve, if that's feasible for the parameters. I concur with Dave Case,
that some sort of clashes are probably throwing off the overall MM energy.
If you run the mdgx parameter fitting with the "AccReport" keyword in the
&param namelist (run mdgx -PARAM to see the keywords and their
descriptions) you will get some Matlab / Octave parseable files that will
help illustrate the energies you put in, the fitted energies, and how the
fitted parameters contribute to them. That will help you track this down.


On Sat, Dec 4, 2021 at 11:40 AM David A Case <> wrote:

> On Sat, Dec 04, 2021, Erdem Yeler wrote:
> >I am trying to parameterize a metal dihedral. I tried to parameterize this
> >dihedral by using mdgx and/or paramfit but I couldn't achieve it. Then I
> >wrote a script for this but also this script failed (even if it worked
> >perfectly for other dihedrals), I changed the molecule design (used
> similar
> >molecules) and tried again and again but I got nothing. Here is the
> >problem: Me is a metal atom and for "os-Me-ca-ca" dihedral MM relative
> >energy is greater than QM energy even if I put in frcmod file this:
> >os-Me-ca-ca 1 0.000 0.000 0.000
> >NOTE: bond and angle parameters are almost OK. There was no imaginer peak
> >in FTIR but The metal is transition metal and VFFDT calculate some
> >Me-os bonds as 0.000kcal/mol.
> >Here is the chart when I use Vn=0:
> >How can I solve this problem?
> I'm guessing that for the MM curve, you are seeing the change in non-bonded
> interactions (and not the change in dihedral energy) as you rotate about
> the
> Me-ca bond. (This assumes that your frcmod file is being correctly read,
> etc.) If so, look at the a configuration near the top of the MM curve, and
> see if you can spot clashes or other interactions that would lead to a
> higher energy than QM is predicting.
> Others might have more relevant experience or suggestions here....
> ....dac
> _______________________________________________
> AMBER mailing list
AMBER mailing list
Received on Mon Dec 13 2021 - 08:30:02 PST
Custom Search