Those clarifications are certainly true... particularly they did some tweaking... so the statement "without any further optimization" should be retreated... still covering almost a five million library of molecules by just ~300 parameters at similar accuracy of GAFF but avoiding some bond issues of the later force field, seems "outperforming" to me, that's the reason for quotations in the initial thread...
Best,
Matías
------------------------------------
PhD.
Researcher at Biomolecular Simulations Lab.
Institut Pasteur de Montevideo | Uruguay
[
http://pasteur.uy/en/labs/biomolecular-simulations-laboratory]
[
http://www.sirahff.com]
----- Mensaje original -----
De: "Niel Henriksen" <shireham.gmail.com>
Para: "AMBER Mailing List" <amber.ambermd.org>
Enviados: Viernes, 7 de Junio 2019 16:01:49
Asunto: Re: [AMBER] Most recent GAFF version to use?
Some clarification ....
> The recently published SMIRNOFF force field seems to "outperforms" GAFF(2)
I don't think that is accurate. The paper says "In summary, then,
SMIRNOFF99Frosst v1.0.7 provides a level of accuracy on these tests that is
comparable
to that of its sibling GAFF"
by just using a subset of GAFF2 parameters (without any further
> optimization)
No it does not just use a subset of GAFF2. From the paper:
"SMIRNOFF99Frosst, derived by converting and adapting AMBER parm99 and
Merck’s parm.Frosst". In the SI, it states that they adjusted some
parameters "to improve generality, and also imported several parameters
from GAFF2.
Best,
--Niel
_______________________________________________
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 Fri Jun 07 2019 - 13:30:03 PDT
