[gmx-users] mix berger ff with opls

[chris.neale at utoronto.ca]

Dear Bin,

I think that you are getting confused here between different  
publications. It was the Lindahl, E., and O. Edholm. 2000. paper that  
you quote below, but recall that they had no protein in that work so  
manipulations of fudgeQQ would have been available to them.

The HEDP method is pretty clear about how one gets different factors.  
I suggest taking a look at the non-peer reviewed .pdf file that I  
posted earlier as it contains more details about the method.

If you still have difficulty, perhaps start from the beginning when  
writing your question.

If you are simply concerned if the method works as advertised, then  
you can search the mailing list to see that I have had this type of  
discussion before -- the method works. Also, you can look at the plot  
in that .pdf file that shows some evidence that it is working alright.

Chris.


-- original message --

Dear Chris:

Thanks a lot for your reply. That's really helpful. Now I have a
further question:
> "1,4 electrostatic inter-actions were reduced a factor of 2 and 1,4   
> Lennard?Jones interactions a factor of 8."

I can understand that the LJ1-4 can be reduced by simply scale epsilon
in [pairtypes], how about 1,4 electrostatics? Since fudgeQQ is 1, how
is the factor of 2 achieved?

Thanks,
Bin

p.s. your paper looks really good, congrats!