On Mon, Sep 12, 2016 at 11:44 AM, Elisa Pieri <elisa.pieri90.gmail.com>
wrote:
> OK, then I have another question: why the charges given by the atominfo
> feature in cpptraj and the ones given by the describe flag in cpinutil.py
> do not match for AS4 - state 0 - (while they do for other residues like
> tyrosine)?
>
This is a good question. The default charges of AS4 were set to match the
charges of ASP. However, ASP and ASH (the protonated form) have different
charges on the backbone atoms (the original RESP derivations of the
standard amino acids in Amber constrained the backbone atoms of all neutral
amino acids to be the same as each other, all positively charged amino
acids to be the same as each other, and all negatively charged amino acids
to be the same as each other, but different charge states were not
constrained to have the same backbone charges).
For CpHMD, backbone charges must be equivalent between protonation states
when actually titrating, because otherwise the reference energy would be
sequence-dependent (since the 1-4 scaling factors would differ based on
whether neighboring residues were positive, neutral, or negative). So the
choice was made to make the backbone charges equivalent to the protonated
form for every charge state. The side-chain atoms were set to the charge
states of their corresponding "non-titrating" residues. But since the
backbone charges differed for the deprotonated form, there was some "extra"
charge left over -- this was simply added to the CB atom. There is a note
of this in the original paper by Mongan, et al. if I recall correctly.
From personal experience, early in my PhD work, I completely reset all of
the charges to match the corresponding non-titratable residues, re-derived
all reference energies to give the correct pKas for the model compound, and
applied it to HEWL as a test. The results were absurd, for this very
reason.
If you look at the charges of ASP, ASH, and the charge states described by
cpinutil.py for AS4, you will see this behavior. However, since the CpHMD
code *always* sets the charges for the titrating residues to be equal to
what's in the cpin file, the most sensible choice of charges for AS4 to
store in the prmtop file is the same as ASP (the most common protonation
state). That way, if you do *not* titrate AS4, it will behave as much like
ASP as possible.
HTH,
Jason
> Elisa
>
> Il 12/set/2016 17:15, "Jason Swails" <jason.swails.gmail.com> ha scritto:
>
> The charges are the same for every solvent model. I added the --describe
> flag because I found it annoying to have to keep going back to source code
> files to look up charges and wanted a quick way to dump a nicely formatted
> summary :).
>
> All the best,
> Jason
>
> On Mon, Sep 12, 2016 at 10:47 AM, Elisa Pieri <elisa.pieri90.gmail.com>
> wrote:
>
> > Hi everybody,
> >
> > I need to know what kind of charges are to one printed by the command:
> > cpinutil.py --describe AS4
> >
> > I have the feeling that they are the one for "implicit solvent
> simulations,
> > i.e. where the aspartates and glutamates atomic radius are not modified.
> >
> > Do you know how to get the information of the charges in different
> > protonation states in this last case (let's say "explicit solvent
> > simulations")?
> >
> > Thank you in advance for your help!
> > Elisa
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
>
>
>
> --
> Jason M. Swails
> _______________________________________________
> 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
>
--
Jason M. Swails
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Sep 12 2016 - 13:00:02 PDT
