Re: [AMBER] Missing atom types in parmchk.dat

From: Robert Heße via AMBER <amber.ambermd.org>
Date: Thu, 25 May 2023 13:37:50 +0200

Am 2023-05-23 19:06, schrieb David A Case:
> On Tue, May 23, 2023, Robert Heße via AMBER wrote:
>> Am 2023-05-22 18:18, schrieb Rodrigo Galindo-Murillo:
>
>>> As far as I can see, you do not need an frcmod file, I ran your
>>> molecule through tleap and it generated the topology and coordinates
>>> just fine.
>>
>> thank you for the fast response. As far as i know it should be an
>> already defined residue as i am saving it directly from the OL15 force
>> field via tleap. For the future i wanted to incoperate modified
>> nucleobases based on the DG of OL15, which i most likely have to
>> parametrize. As such i wanted to compare them with an already
>> established residue but i fear that the parametrization might not be
>> possible without defining the missing atom types somehow.
>
> I would still wait to see what parameters are missing in your modified
> bases. If something is missing (without running parmchk), you may be
> able
> to hand-create a frcmod file based on analogies to what is in
> gaff2.dat.
>
> ....dac
Hello,

it seems like multiple parameters are missing. For example when
introducing an imidazoyl ring a total of 39 bounds, angles and torsion
parameters are missing. Here are some examples from the leap.log:
"Error: Could not find bond parameter for atom types: CB - CM
         for atom C7 at position -7.517909, 4.302219, -0.380126
         and atom C10 at position -7.745268, 3.845895, -1.666567.
Error: ** No torsion terms for atom types: CB-CM-CR-NA
         for atom C7 at position -7.517909, 4.302219, -0.380126,
             atom C10 at position -7.745268, 3.845895, -1.666567,
             atom C11 at position -8.896810, 4.202512, -2.503798,
         and atom N6 at position -8.999518, 3.718721, -3.761957."
I also tried to to run it with gaff2 atom types, for which i generated a
frcmod of the nucleobase. This reduced the missing parameters to 11 but
these parameters are mixed between OL15 for the backbone and gaff 2 for
the nucleobase:
"Error: ** No torsion terms for atom types: CT-CT-na-cc
         for atom C2' at position -4.042350, 5.022000, 2.260350,
             atom C1' at position -5.273350, 4.140000, 2.447350,
             atom N1 at position -6.311605, 4.370884, 1.453440,
         and atom C6 at position -7.450184, 5.152021, 1.549649."
Would it be reasonable to replace the missing parameters with those from
a pure gaff2 frcmod or should i try to find a way to parametrize it with
the missing atom types i mentioned in my first mail? Ideally i want to
be able to parametrize and implement more residues in the future, so
being able to reliable generate parameters for these mixed interactions
would be quite helpful.

With best regards,
Robert Hesse

_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu May 25 2023 - 05:00:05 PDT
Custom Search