How do any of the conventional MD codes deal with Jahn-Teller effects in copper(II)? Answer: they don't. So, if you're really interested in the local environment around the metal, you'll need something special like Ligand Field Molecular Mechanics which I have been developing (see Deeth, R. J. Inorg. Chem. 2007, 46, 4492-4503 for an application involving Type I blue copper centres). If you're not so fussed about the detailed coordination chemistry, it's not too hard to add a few parameters to keep the metal from flying out of the active site.

I have no experience with Desmond but Gromacs that is also a free program to run molecular dynamics will work. I performed a MDS with a metalloprotease (Zn) and it worked perfectly. I used Charmm36 as all atom force field.

Defenitively yes... I think that the CHARMM force field contains some specific parametrization for Histidine coordinating metals.

Just as an example: http://schererlab.uchicago.edu/pubs/42_bpjlu.pdf

This is a quite old paper but i'm sure that you can find more recent studies about that.

Just for the record: if you need advice about something "I've encountered problems" is a little bit vague.

I'm not familiar with the specific program Desmond, but in general it is possible to do MD simulations on metalloproteins. There are some issues to consider though:

I have found that copper (we were looking at copper(I) using an AMBER forcefield, so you system may be a little different) doesn't stay bound if you simply treat it as a sphere with a charge, it needs to have explicit bonds to its ligands. Other metals such as Mg, which has a larger charge and generally more ligands does stay bound in this way. We got parameters for the Cu-ligand bonds from ab initio 6-31g** / LANL2DZ calculations.

You need to consider the coordination arrangement - if it's not octahedral (or tetrahedral for small ions) then simple point charges will not give you the right arrangement. If you have explicit bonds, you can also define bond angles to maintain the correct coordination arrangment but some geometries end up with multiple allowed values or the ligand-metal-ligand angle and so may lead to problems with some MM formalisms (e.g. AMBER). e.g. square-planar arrangements have lig-met-lig angles of 90 and 180. We were fortunate as our system was a linear bidentate system with 2 ligands and an angle of 180 only.

You also need to take the charges and electrostatic model into account - full positive charge equal to oxidation state may be appropriate for a metal that interacts largely through coulombic interactions (e.g. Mg) but charge should probably be reduced if you have explicit bonding.

Whatever forcefield you use, it probably won't be parameterised for your metal ions so you have to add your own parameters and monitor carefully what the system does and compare with experiment wherever possible.

How do any of the conventional MD codes deal with Jahn-Teller effects in copper(II)? Answer: they don't. So, if you're really interested in the local environment around the metal, you'll need something special like Ligand Field Molecular Mechanics which I have been developing (see Deeth, R. J. Inorg. Chem. 2007, 46, 4492-4503 for an application involving Type I blue copper centres). If you're not so fussed about the detailed coordination chemistry, it's not too hard to add a few parameters to keep the metal from flying out of the active site.

I can add my opinion to that already given. Surely it's possible to simulate a metalloprotein by MD, and the force fields suggested (AMBER, CHARMM and also GROMACS) are very good for describing the metal center and its ligands. In particular, the presence of copper as metal ion can give better chances than other one. Furthermore the ligands as histidine, probably present in the selected metalloprotein, should not have problems. But if the parameters for the copper bound to imidazole ring of histidines are not present in a force field, is always possible to overcme the problem by obtaining them by previous theoretical calculations also at DFT level, using a good quality basis set . In this way is sufficient to take into account the charges arising from a NBO analysis, some people use also Mullikan ones, and put them in the molecular data file of that particular force field used. In any case in literature is present a lot of works that can clarify this point.

I have used CHARMM27 force field in Gromacs to do MD simulation of Metalloprotein which contains ZN. you can search CHARMM bbs. someone might have done this and you can use their parameters. at least they works well for zincfinger

You can also use the NAMD MD program with the CHARMM27 force field and harmonic Tcl forces to keep your Cu in place.

I want only to add that you should pay attention because copper ligand it is a difficult issue to modelize with MD. Another way should be mixed MD/QM simulation.

These links will give you some idea about force field parameters of metal containing systems, as all available force fields are devoid of metal-ligand potentials. On the hand, developing force field is a cumbersome task or you should go for QM/MM MD simulations.

http://ambermd.org/tutorials/advanced/tutorial1_orig/

http://ambermd.org/tutorials/advanced/tutorial1_adv/

I agree with Tiziana Marino. On my side I am using either XPLOR-NIH or NAMD with a force field modified of CHARMM. the best is to find all paramters needed for MD in using DFT and NBO for atomic charges. For Cu+, Cu2+, Mn2+, Mn3+, have a look to the paragraph "5.2. Possible cation exchange at chelate cage sites" of Charlet et al, CCR 2012, Neurodegenerative diseases and exposure to the environmental metals Mn, Pb, and Hg . Paper is free in my list.

Yes...!!! You can certainly go ahead with the Desmond. For Mettaloproteases in Desmond before going for the energy minimization its always suggested to go for the Hetero state generation (Generate Het states) in the preparation step (Which is a part of maestro's Protein Preparation wizard). This step will enables the formation of suitable het state or co-ordinate covalent state in complex with protein, so that your force field like OPLS or CHARMM can easily recognize the Copper atom. The only problem for performing this step is that it requires a licence from Schrodinger software vendors for using this specific optimization and minimization method.