[gmx-users] Regarding conversion of C6/C12 parameters to sigma/epsilon

[Chris Neale]

I assume that you are talking about the [ nonbons_params ] section from lipid.itp.

To answer this question requires an understanding of what information is contained in the [ nonbond_params ] section (A) generally, 
and (B) in lipid.itp

The answer to (A) is in the manual and the mailing list

For the answer to (B) look at the lipid.itp file. First there is a section that describes lipid-lipid parameters.
Feel free to convert them and include them if you want but you can take care of that simply by using the default comb-rule 3 from oplsaa.
To check that out for yourself check some of the combinations in that section vs the rule for comb-rule 3.

Next there is a section that begins with the comment ";; paramaters for lipid-GROMOS interactions". 
You obviously don't want any of those from the lipid-GROMOS section if you are using oplsaa 
(in fact it would be a disaster to include them.) 

Next there is a section that begins with the comment ";; lipid-SPC/SPCE interactions".
This section is treated similarly.

Therefore the simple answer is: Ignore them.

I have seen your (many) posts of this question but I have been too busy to reply until now.
Chris.

--- original message ---

              I have one doubt regarding conversion of
C6/C12 parameters to sigma /epsilon. 
I added atomtypes, pairtypes and dihedraltypes from
lipid.itp to ffoplsaanb.itp---after changing C6/C12
paramters to sigma/epsilon as given in mailing list
(http://www.gromacs.org/pipermail/gmx-users/2006-May/021416.html
). 
 
My questions are
(1) Is it also required to do conversion of C6/C12
parameters of “nonbond_params“to sigma/epsilon and
adding these parameters from lipid.itp to
ffoplsaanb.itp.
 
(2) If it is not required to convert” nonbond_params”
parameters, is it required to call these parameters
without conversion.
 
               Any suggestion is appreciated, thank
you in advance.