[gmx-users] restraining Ca2+ ion

[Marc F. Lensink]

On Wed, Jan 07, 2009 at 09:43:25AM +1100, Mark Abraham wrote:
> jayant james wrote:
> >Hi!
> >I have been performing MD simulations of a protein which has 3 ca2+ 
> >ions. I find that after 3ns of simulations the ca2+ ions start to drift 
> >away. So what would be a good strategy to have the ca2+ in place during 
> >the simulations. I would  prefer not to use position restraining as this 
> >would pin the ca2+ in space hindering the spacial movement of that 
> >particular area!! but I am wondering if distance restraining of the ca2+ 
> >to the adjacent residues would be fine? or is there any other/better 
> >strategy? Awaiting your suggestions.
> 
> This is one of the purposes of distance restraints. Better would be a 
> proper model for protein-bound calcium ions. Also, a poor-quality 
> electrostatic treatment could be the source of the problem (e.g. a 
> cut-off shorter than the distances to nearby negative (partial) charges).

you need to modify the charges of the residues binding the calcium, as
they are heavily polarized:

Shiratori and Nakagawa, J.Comp.Chem. 12 (1991), 717-730
"Parametrization of the calcium binding site in proteins and molecular
dynamics simulation of phospholipase A2"

if you can't get this article, have a look at table 6.5 in my thesis:
http://irs.ub.rug.nl/ppn/237863316

cheers,
marc