Thank you very much for your suggestions.
I will check it out.
On Mon, Jul 10, 2023 at 5:17 PM David A Case <david.case.rutgers.edu> wrote:
> On Fri, Jul 07, 2023, Siddhique via AMBER wrote:
> >
> >I have used a slightly modified nucleotide (isolated nucleoside
> >monophosphate). The parameters were generated using antechamber (RESP
> >method). TIP3P was used for water and ions.
> >Most hydrogen bonds were more than 2.8 A distance, but a few were shorter,
> >which is causing the problem.
> >
> >A part of the hydrogen bond output (solute-solute) from cpptraj:
> >(command used: hbond intra :1-last out hbond_onlyinter.dat avgout
> > hbond_onlyinter-avg.dat )
> >
> >#Acceptor DonorH Donor Frames Frac
> > AvgDist AvgAng
> >GMP_11.O8 GMP_14.H5 GMP_14.O2 1000 1.0000
> >2.4945 168.1674
>
> All of your H-bonds are between two different GMP molecules. Only one is
> less than 2.5 Ang., so there is no real evidence of any error. I
> occasionally see short H-bonds to waters, and this may be similar.
>
> Basically, force fields are tuned to thermodynamic properties more than
> geometries of H-bonds. This can indeed cause problems if you need to to
> quantum calculations. You might consider a quantum minimization, or
> some very fast quantum-like optimization such as xtb, prior to other
> analysis.
>
> If you really need MM results with no short H-bonds, you will probably be
> faced with generating a new force field, probably with non-zero van der
> Waals radii on hydrogens bonded to oxygen. You might also look at the
> contributed parameters database, and modify what is there for GTP into GMP.
> That might be more like other nucleotides than you would get by using
> antechamber.
>
> ...good luck...dac
>
>
--
Siddhique
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Jul 12 2023 - 01:30:02 PDT
