Dear Stano,
I can’t spot anything that seems obviously wrong with your procedure and input file. The temperature at the beginning of the adaptive qm/mm trajectory should be the same as in the restart file. Differences due to changing the potential from mm to qm in a small region should then be rather small. In any case the thermostat should push it to your target temperature.
Have you tried with fixed number of molecules in the transition region - do you get a different behavior? Have you tested this with the built-in semiempirical methods? These are quick and convenient for figuring out reasonable numerical settings. PM3-MAIS is a pretty good potential for water but it should work with any qm method. If you send me your input files (off the mailing list), I can have a closer look.
All the best,
Andy
—
Dr. Andreas W. Goetz
Assistant Project Scientist
San Diego Supercomputer Center
Tel: +1-858-822-4771
Email: agoetz.sdsc.edu
Web: www.awgoetz.de
> On May 25, 2016, at 5:58 AM, Stanislav Simko <s.simko.uu.nl> wrote:
>
> Dear all,
> we're running the adaptive qm/mm NVT equilibration with the target
> temperature 300K (default temperature). However, we're reaching
> temperatures up to 280K.
> Our QM/MM region is fairly small - up to 4.5A sphere in the 40A box,
> with several transition regions - 0.3, 0.5, 0.75, 1.0 and 1.25
> Angstrom, so there's up to ~10 H2O molecules in the QM/MM. Therefore,
> the effect on the temperature should be negligible.
> We're getting the same effect as show on the the attached files for the
> 0.75A case for all of the simulations.
>
> At first we ran full MM NPT and then NVT equilibration and used the
> final restart as input for each of the QM/MM calculations. Very strange
> is also that the temperature drops by 100K already after the first
> step.
> Please, do you have any suggestions what could be going wrong?
> Attached are our input file for the 0.75A case and its temperature plot
> and the temperature plot for the NVT equilibration
> Thank you very much.
> Regards
> stanislav.
> <0.75A.in><nvt_fullMM.eps><temperature0.75.eps>_______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed May 25 2016 - 17:00:03 PDT
