Re: [AMBER] Amber force-filed parameters for amino-acid side chain analogues

From: David Cerutti <dscerutti.gmail.com>
Date: Thu, 31 Jan 2019 08:34:13 -0500

Yes, this business of resolving integer charge values by adjusting CB and
its two or three associated hydrogens on side chain analogues is standard.
Not seeing the rest of your original email, if you are doing this, then you
are drawing conclusions about more foundational things than the ff14SB
force field--really I'd say you're looking all the way back to ff95. The
changes in ff99 were van-der Waals parameters on hydrogens, and for ff99SB
there were backbone torsional changes. ff14SB introduced backbone torsion
edits as well as side chain Chi1 / Chi2 edits, but you see where I'm
going--the later force fields didn't change all that much in the actual
side chains, even though they did make changes to improve side chain
rotamer propensities.

Dave


On Thu, Jan 31, 2019 at 7:46 AM David A Case <david.case.rutgers.edu> wrote:

> On Thu, Jan 31, 2019, mish wrote:
>
> >I thought it makes sense to
> >keep the side-chain part exactly same as in AMBER14 force-field and just
> to
> >change it on Cb. This is going to affect some properties but the
> >interaction of side-chain part with my small molecules may not be affected
> >much and I can still keep parameters closer to native AMBER force-field.
>
> I agree with the above. Among other things, there are lots of
> calculations out there that have used exactly this scheme. (I think
> Michael Shirts was the first to do this, but I'm not sure.) Certainly,
> if you plan to draw some conclusions about the Amber14SB force field,
> this seems like the best way to proceed. (At least to me.) Since there
> are only 19 amino acid sidechains, it wouldn't take very long to just do
> it.
>
> OR, check the supplementary materials in the literature. People that
> have carried out these sorts of calculations are quite diligent in
> providing complete details of what they have done. So there might be
> some files out there. The relevent authors are probably not following
> the Amber mailing list. Try an email to Michael or David Mobely.
>
> ...good luck...dac
>
>
> _______________________________________________
> 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 Thu Jan 31 2019 - 06:00:02 PST
Custom Search