[gmx-users] Gibbs Energy Calculation and charges

[Christopher Neale]

Dear Dallas:

Seems like you could test Michael's idea by removing all 1-4 NB interactions from your topology. It won't produce any biologically useful results, but might be a worthwhile check to see if indeed this is the issue.

To do this, I figure you would set gen-pairs to "no" in the [ defaults ] directive of forcefield.itp, remove the [ pairtypes ] section from ffnonbonded.itp, and remove the [ pairs ] section from your molecular .itp file. (You can quickly check that the 1-4 energy is zero in all states to ensure that this works).

If that gives you the result that you expect, then you could go on to explicitely state the 1-4 interactions for the A and B states (I presume that this is possible). Of course, you should be able to jump directly to this second test, but the first test might be useful because it rules out the possibility that you make a typo somewhere.

Chris.

-- original message --

I think the grammar got a little garbled there, so I'm not sure quite
what you are claiming.

One important thing to remember; 1-4 interactions are treated as
bonded interactions right now FOR COUPLE intramol (not for lambda
dependence of the potential energy function), so whether
couple-intramol is set to yes or no does not affect these interactions
at all.  It only affects the nonbondeds with distances greater than
1-5.  At least to me, this is nonintuitive (and we're coming up with a
better scheme for 5.0), but might that explain what you are getting?

On Tue, Oct 29, 2013 at 9:44 PM, Dallas Warren <Dallas.Warren at monash.edu> wrote:
> Just want this to make another pass, just in case those in the know missed it.
>
> Using couple-intrmol = yes the resulting dH/dl plot actually looks like that at lamba = 1 it is actually equal to couple-intramol = no with lambda = 0.
>
> Should that be the case?
>
> Catch ya,
>
> Dr. Dallas Warren