[gmx-users] RE: RE: [nonbond_params] section , choice of parameters

[Ehud Schreiber]

Hi again,

 

Yes, you can simply specify sigma and epsilon, but indeed in your case
you do not want to do so but rather to specify C_6 and C_{12} with the
former being zero to avoid attraction.

Your .top file should include a [ defaults ] section, perhaps through an
#include of a .itp file. Perhaps yours is in ffamber99.itp (I don't have
it installed so I can't tell). In this section there are 5 parameters;
the second is a "combination rule" which must equal 1 for the two
numbers to represent indeed C_6 and C_{12} (see table 5.3, eq. 5.1-5.4,
and in page 112).

In short, check whether the second parameter in the [ defaults ] section
of ffamber99.itp is 1, if not that's probably the problem. Of course, if
you change this you'll screw up all the other parameters, so it seems
you could either change the force field or create your copy of the files
with all the relevant LJ parameters changed to the "combination rule"=1
convention.

 

 

From: Ehud Schreiber 
Sent: Wednesday, July 28, 2010 3:01 PM
To: 'gmx-users at gromacs.org'
Subject: RE: [nonbond_params] section , choice of parameters

 

Hi Amir,

 

Are you sure you're giving the parameters indeed as C_6 and C_{12}? 

According to the combination rule in the [ defaults ] section, you may
be giving the two parameters directly as sigma and epsilon; sigma=0 (as
well as epsilon=0) would give an identically zero potential . See
sections 5.3.3 and 5.7.1 in the manual (version 4.0).

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20100728/0d3122f2/attachment.html>