Hi Ananya
This will require code changes in the Gaussian interface. It is not hard to do but requires changes in AmberTools/src/sander/qm2_extern_gau_module.F90 to enable reading from a provided checkpoint file.
Currently, sander removes the checkpoint file at the beginning of a simulation to avoid potential issues with restarts. The checkpoint file from step 1 is then used from step 2 onwards.
It would be easy to change this behavior or add a new keyword to the &gau namelist to provide an initial guess and have Gaussian read from this in the first step. I don’t have bandwidth to look at this currently, I will try to revisit later, but perhaps someone else can help in the meantime?
All the best,
Andy
—
Dr. Andreas W. Goetz
Associate Research Scientist
San Diego Supercomputer Center
Tel: +1-858-822-4771
Email: agoetz.sdsc.edu
Web: www.awgoetz.de
On Oct 14, 2025, at 12:15 AM, Ananya Nayak via AMBER <amber.ambermd.org> wrote:
Hi PUJAN,
Thanks for your reply. As of now, we have not started the cDFT
calculations. We were planning to use QChem or Gaussian. The plan of
action was to generate a guess with all high spins(I have more than one
metal center) and then use that guess for a physiologically reportated
state, As the physiologically reported state is not converging. I'm using
Sanders to do the patching between the MM and QM region. The main thing I'm
concerned about is it's not reading the .chk files that I'm providing
manually. So, in order to do what we have planned I need to find a way to
make sander use the already provided chk file, instead of writing it's own.
I'm quite new to this and I don't know how to go forward. Please let me
know if there is anything I can do.
Thanks
Ananya
On Tue, Oct 14, 2025 at 12:15 PM PUJAN AJMERA <pujanajmera1.g.ucla.edu<mailto:pujanajmera1.g.ucla.edu>>
wrote:
Hi Ananya,
It really depends on which software you used to generate the constrained
DFT wavefunction. Can you provide the software and/or input files from the
cDFT run? Regardless, you would want to reconverge the wavefunction in
Gaussian, or use the Gaussian fragment guess option to reconstruct
antiferromagnetic coupling, which generally works well. Further, your cDFT
guess may not stay restrained to the desired electronic state if you are
switching softwares that can't provide that.
Best,
Pujan Ajmera
On Mon, Oct 13, 2025 at 11:35 PM Ananya Nayak via AMBER <amber.ambermd.org>
wrote:
Dear Users,
We are trying to do a qmmm calculation for a metalloprotein. The
metalloprotein has some anti ferromagnetic coupling. And we wanted to use
a
guess for the wave function obtained from previous cDFT calculations and
do
the calculations there after. I'm not able to find a way to do that.
Please let me know if it's possible and how to do that. Any insight is
helpful.
Thank you
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
https://urldefense.com/v3/__http://lists.ambermd.org/mailman/listinfo/amber__;!!Mih3wA!Gbyb8xRX3ODs7wavAs3vdnvIr3Zsz6RREY8QxKVjVYQe9V3PmkIK_RGkUi82ulbrD1u79eRA6LzI-g$
--
*4th year, Alexandrova Lab*
*UCLA Chemistry PhD Program, Theoretical and Computational Chemistry*
_______________________________________________
AMBER mailing list
AMBER.ambermd.org<mailto:
AMBER.ambermd.org>
https://urldefense.com/v3/__http://lists.ambermd.org/mailman/listinfo/amber__;!!Mih3wA!Gbyb8xRX3ODs7wavAs3vdnvIr3Zsz6RREY8QxKVjVYQe9V3PmkIK_RGkUi82ulbrD1u79eRA6LzI-g$
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Oct 14 2025 - 19:30:02 PDT
