[gmx-users] Free energy perturbation of Cu(I) binding

[Mark Abraham]

abhigna polavarapu wrote:
> Hey Mark,
>  I am just not getting the theory behind doing position restraints and
> applying them to equilibrate.

Sure... so that's why I recommended chapter 5 (which has an example of 
position restraints in section 5.7.1, which you can understand with 
references to the table in that section) and some tutorial material 
(which may deal with some issues in a real context).

> So let me frame my question well. Cu(I)
> is binding to three amino acids in the protein. Now If I remove Cu(I)
> the -1 charge of the protein changes to 0.

The net charge of the system will decrease, surely.

> So the protein has two
> specific charge states which it should equilibrate. So coming to
> position restraint should I restrain the total protein when I remove
> Cu from the structure or just the copper binding core and how does
> restraining help the protein from destabilizing due to huge
> electrostatic interactions.

It won't. Like I've said previously in this thread 
(http://www.gromacs.org/pipermail/gmx-users/2009-February/039443.html), 
you cannot expect the same MM model to work in both states. Position 
restraints are useful to keep something close to the same structure 
while something else relaxes, but they are not a magic bullet.

Anyway, I've repeated myself enough times today :-)

Mark