[gmx-users] False C6/C12 parameter for DMSO and chloroform in GROMOS 45a3 force field (ffG45a3)

[Roman Affentranger]

Dear gmx-users (and developers ;-)

I believe there are false LJ pair-parameters in gromacs' implementation of
the GROMOS 45a3 force field for interactions between the atoms of DMSO, and
between the atoms of chloroform. The C6/C12 pair paramteters for
CDMSO-SDMSO, ODMSO-SDMSO, ODMSO-CDMSO, and CLCHL-CCHL, HCHL-CCHL, HCHL-CLCHL
in ffG45a3nb.itp are simply the (root of the) product of the individual
atomic parameters, whereas separate parameters - with no relation to the
atomic ones - are defined in the original GROMOS force field. The errors are
there in the force field files distributed with gromacs version 4.0.5 all
the way down to at least version 3.2.1 (it's as far as I've looked), and
also in the ffG45a3.tar.gz file to be found at
http://www.gromacs.org/index.php?title=Download_%26_Installation/User_contributions/Force_fields

For example:
The lines in ffG45a3nb.itp read:
   CDMSO    SDMSO  1  0.009777435  2.1627421e-05
...
   ODMSO    SDMSO  1  0.004897196  4.0192831e-06
   ODMSO    CDMSO  1  0.0045335636  4.0434685e-06

However, they should be
   CDMSO    SDMSO  1  0.0097827  2.16523e-05
...
   ODMSO    SDMSO  1  0.0052442  4.6094e-06
   ODMSO    CDMSO  1  0.0049187  4.7597e-06
as they are in the GROMOS 43a1 force field.
A similar problem holds for chloroform (atoms CCHL, CLCHL and HCHL).

Find attached a corrected version of ffG45a3nb.itp.

Cheers,

Roman



-- 
=======================================
Roman Affentranger, Ph.D.
Group of Computational Biology and Proteomics
Institute of Biotechnology and Biomedicine
Universitat Autònoma de Barcelona
E-08193 Bellaterra (Barcelona)

Phone: (+34)93 581 2807
Fax: (+34)93 581 2011
E-mail: roman.affentranger at bioinf.uab.es
=======================================
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20090702/cbd05c03/attachment.html>