Hi Shiji,
Yes, it seems like antechamber does accept GAMESS outputs, using the -fi
gamess option. This option is mostly undocumented, antechamber -L lists it
but the Amber manual does not.
The complete command is something like this:
antechamber -i test.dat -fi gamess -o resp.mol2 -fo mol2 -c resp -eq 2
The same input works in espgen -i test.dat -o test.esp (the format is
apparently autodetected). I've tested with ambertools=22.0 and 20.9.
Here test.dat is what GAMESS calls a "punch" file. I have some examples of
the file format, please let me know if you'd like to have a look. GAMESS
generates its own points at which to evaluate the potential, these seem
very similar to Gaussian.
I am happy I got this working, but would still really like to be able to do
the same thing for other QM software, especially the software already
bundled with Amber, and software which can be installed directly from
conda-forge :)
The tricky part is generating the points - Orca, Psi4, etc can easily
calculate potential-at-a-point, but don't have any builtin function to find
the points on Connoly-type surfaces around the provided atoms. Do you have
any example code which calculates the coordinates of the points?
Best,
Alex
On Tue, Oct 18, 2022 at 12:34 AM Shiji Zhao <shijiz.uci.edu> wrote:
> Hi Alexander,
>
> I read the source code of espgen.c in AmberTools and noticed it has added
> support to GAMESS. The espgen version is 1.1.
>
> I have never used GAMESS before. Maybe you can give it a try and let me
> know if espgen can handle GAMESS output files. This information will also
> be helpful to me since I am considering adding more features to the
> espgen code.
>
> Best,
> Shiji
>
> On Thu, Jul 7, 2022 at 5:58 PM Alexander Izvorski via AMBER <
> amber.ambermd.org> wrote:
>
>> Hello Amber folks,
>>
>> I'd like to use antechamber and espgen/respgen to fit RESP charges without
>> using Gaussian.
>>
>> I'd be potentially interested in any other QM software: QUICK (which is
>> nice because it's now bundled with ambertools), Psi4, CP2K, GAMESS, Orca
>> etc. It just seems a bit limiting to have Gaussian be the only choice for
>> this type of calculation.
>>
>> The latest Amber22 manual seems to assume Gaussian-only in a few places:
>>
>> 15.3.5. espgen
>> > Espgen reads in a gaussian (92,94,98,03) output file and extracts the
>> ESP
>> > information
>> >
>> > 15.1.1. antechamber
>> > -ge gaussian esp file generated by iop(6/50=1), default is g09.gesp
>> >
>>
>> Has anyone successfully used non-Gaussian software for RESP? Are there
>> any
>> supported options?
>>
>> If not, could someone describe the expected output file format from
>> Gaussian? I can probably try to write a format converter, since most of
>> the tools mentioned above can calculate and save electrostatic potential.
>> For example Psi4 has the GRID_ESP
>> <
>> https://urldefense.com/v3/__https://psicode.org/psi4manual/master/oeprop.html__;!!CzAuKJ42GuquVTTmVmPViYEvSg!O17LEeXQ1OBXwzTktJhx4KPTeYr_tNx1yFOLRLOc5cdl2zCjZKuSgxK1Mn6bAUVS0-pDwyFA1PkLTw$
>> > option. I just don't
>> have any idea what the espgen or respgen-compatible format should look
>> like.
>>
>> Best,
>> Alex
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>>
>> https://urldefense.com/v3/__http://lists.ambermd.org/mailman/listinfo/amber__;!!CzAuKJ42GuquVTTmVmPViYEvSg!O17LEeXQ1OBXwzTktJhx4KPTeYr_tNx1yFOLRLOc5cdl2zCjZKuSgxK1Mn6bAUVS0-pDwyHUuc7TsA$
>>
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sat Oct 29 2022 - 13:30:02 PDT
