From: Kenno Vanommeslaeghe (kvanomme_at_rx.umaryland.edu)
Date: Wed Apr 02 2014 - 14:05:49 CDT
On 04/02/2014 02:38 PM, Rawan Al Nsour wrote:
> Dear All,
> I generated full set of parameters for a linear polymer, the backbone
> dihedral includes ten terms (i.e., n=1,2,...10).
> I started my simulations and I got this message
> "Multiple dihedral with multiplicity of 7 greater than max of 6"
> Any recommendation is appreciated.
My recommendation is to redo the parametrization with only terms that
reflect the symmetry of the system.
http://mackerell.umaryland.edu/~kenno/cgenff/faq.html#multi
What you did is total overfitting madness.
https://en.wikipedia.org/wiki/Overfitting
There's a reason terms higher than 6 are not implemented in most MD
programs.(*)
(*) The MM implementation in Gaussian even doesn't go higher than 4,
though that's also wrong; for example, in trifluorotoluene, 6 is the only
"symmetry-allowed" multiplicity, and is indispensable.
This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:22:16 CST