Re: Re: some recommendations for parametrizing new molecules requiredž

From: Giacomo Fiorin (gfiorin_at_seas.upenn.edu)
Date: Fri Jan 23 2009 - 18:41:14 CST

Hi Vlad, CHARMM charges are derived in one way, and AMBER charges in
another. Either set of charges is not perfect.

Each chemical group carries a small deviation from the reality. In
some cases, such deviations are also smaller than the differences
between one type of chemical group and another (you don't need to be
too accurate to see that a carbonyl is different from a methyl).

But in many other cases, chemical groups are more similar and there is
in general no solution other than to accept that these errors count.
By being consistent, at least you ensure that identical chemical
groups will also have identical properties.

Instead, if you are not consistent and start mixing different force
fields, you will create an artificial imbalance between (otherwise
equivalent) chemical groups, and you could mistake spurious effects
for actual phenomena.

---- -----
 Giacomo Fiorin
   Center for Molecular Modeling at
     University of Pennsylvania
     231 S 34th Street, Philadelphia, PA 19104-6323
 phone: (+1)-215-573-4773
 fax: (+1)-215-573-6233
 mobile: (+1)-267-324-7676
 mail: giacomo.fiorin_<at>_gmail.com
 web: http://www.cmm.upenn.edu/
---- ----

2009/1/23 CHINDEA Vlad <vchindea_at_hotmail.com>:
> Hi everybody
>
> Many thanks for all this valuable information.
> Certainly I did not expect the parametrization of a new molecule to be a
> ride in the park ! That is why I thought it appropriate to ask first the
> opinon of wise men instead of rushing boldly in the wrong direction....and
> hopefully I'll limit the potential damage to myself.
>
> Now it's clear to me that the splitting of the larger cycle (80 atoms) of
> the molecule is unavoidable and that I have to 'squeez' more computer power
> from my university. Probably I will do the first learning trials on the
> 'piunny little' machine mentioned and than I will try to snatch some time on
> the Linux cluster which is foreseen to be installed at our site.
>
> But still it is not clear for me if simulations with NAMD using CHARMM22
> would be more appropriate with CHARMM or AMBER charges. In Paratool
> documentation it is explain how to derive them but not if one or the other
> would be more appropriate in a certain context.
>
> Best wishes
> Vlad
>
>
> ________________________________
> See all the ways you can stay connected to friends and family

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:50:25 CST