thanks Matías, I'm not up on GROMACS development so wasn't sure of the
details. CMAPs in ff19SB are also done as corrections, though the details
shouldn't matter if they're just calculated separately from whatever is in
the dihedral list. Our atom types shouldn't matter either if they support
ff14SB, since some changed there too. For testing we made a validation
suite that calculates energies for lots of structures including all amino
acids, and I'd be happy to work with someone that knows the GROMACS end of
things.
carlos
On Tue, Nov 16, 2021 at 10:07 AM Matias Machado <mmachado.pasteur.edu.uy>
wrote:
> Dear Stephane,
>
> You're right, GROMACS can support flexible CMAP definitions as headers in
> "cmap.itp" set the specific dihedral types. Hence, ff19SB should be
> portable to GROMACS, but new (ff19SB) atom types must be introduced. The
> other thing is, porting also requires proper testing...
>
> I'm not the developer here, Carlos should better say, but despite using
> the same implementation, CHARMM and AMBER CMAPs mean slightly different
> things... While CHARMM CMAPs are corrections to the dihedral function,
> AMBER CMAPs represent the dihedral function...
>
> I'm not a big fan of automagic porting... so I suggest you to start
> implementing ff19SB from a validated ffAMBER port of ff14SB (
> https://www.gromacs.org/Downloads/User_contributions/Force_fields).
> Beware that GROMACS and AMBER use different energy units.
>
> Last but not least, note that GROMACS code performs slower with CMAPs than
> without.
>
> All the best,
>
> Matías
>
> ----- Mensaje original -----
> De: "Alan" <alanwilter.gmail.com>
> Para: "AMBER Mailing List" <amber.ambermd.org>
> Enviados: Martes, 16 de Noviembre 2021 5:28:13
> Asunto: Re: [AMBER] Can Parmed convert AMBER files to GROMACS made with
> the AMBER19SB ?
>
> Hi Stephane,
>
> GROMACS has been supporting CMÄP for different Amino acids for a long time
> > and I often use it with CHARMM36 force field. Is the CMAP approach of
> > AMBER19SB different from CHARMM36 one ? Sorry if it is a stupid
> question.
> >
>
> This is not a stupid question but it should be addressed to GROMACS.
>
> It's just a shame they stopped their mailing list. I stopped following
> their new GROMACS Forum as I'm only interested in ACPYPE questions and I
> don't have time to spare visiting their Forum, when filtered emails (like I
> have here in Amber Mailing List) do this trick simple and quick.
>
> If you get an answer there, please share it with us here.
>
> Best
> --
> Alan Silva 🚲🏊♂🏃♂🇧🇷🇫🇷🇬🇧
>
> PS: BTW, please Amber maintainers, please never stop this Mailing List.
> _______________________________________________
> 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
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Nov 16 2021 - 10:00:02 PST
