Re: [h5md-user] Additional properties of particles

[Jakub Krajniak]

W dniu 01.12.2014 o 16:50, Felix Höfling pisze:
> Am 01.12.2014, 15:04 Uhr, schrieb Pierre de Buyl
> <address@hidden>:
>
> > I didn't think that we would need a full proposal for charges as it would be
> > just an additional field. To remove the ambiguity about the kind of charges, I
> > have put a proposal :-)
> >
> > It is not yet generated on the official homepage but github renders the markdown
> > at https://github.com/h5md/h5md/blob/master/proposals/0102_particles_charge.mdwn
> >
> > Cheers,
> >
> > Pierre
>
> Hi Pierre,
>
> Thanks for putting a proposal on this! I fully support it, but have some
> minor questions/comments:
>
> - Why do you use "real" for the formal charge? We could simply use
> "formal" (and say that this kind of charge is meant if nothing else is
> specified?)
>
> - In the case of force fields, is the charge unique to the particle or to
> the bond? Perhaps, it would be best to specify force-field bonds as a
> whole (outside of particles/), including all parameters (partial charges,
> bond length, bond stiffness, ...). If I understand the concept of force
> fields correctly, there is only a finite (albeit large) number of such
> parameter sets, and the bonds depends on the atoms involved and their
> position within the protein.

Great, thanks Pierre for the proposal :-)

AFAIK the popular force field formats attaches the charge as an attribute of the 
atom type not a bond type. So looking in e.g. GROMACS FF definitions there is a 
multiple set of carbon atoms with different properties depending also on the 
bond type.

Best,
Jakub


--
Jakub Krajniak, Computer Science Department @ KU Leuven
mob. +32 477 688 404
room: 200a 01.41

Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm