[gmx-users] Cu-O distance too long

[nicolis at guideo.fr]

Hello again,
I try to model copper/drug complexes and as a test to begin with, I tried 
just Cu2+ in a water box.

The problem is I obtain distances of 2.25-2.35 Angstroms between 
copper and the nearest water oxygens, instead of the 1.95-2.00 that is the 
correct value.

What did I wrong?
the forcefield? (I used ffG43a1.itp)
the water? (I used spc.itp)
the box size? (I used a 7nm dodecahedron, big enough I believe)
the long range electrostatics? (I just took cut-off, but even it is better to use 
PME I don't expect it to have an incidence on nearest neighbours)

Any suggestions please? If I don't get correct Cu-water distances I can't 
go forward to the drugs.

Thank you,

Ioannis