[gmx-users] Non-bonded Copper

[Mark Abraham]

Matthew Roode wrote:
> Greetings all,
> 
> Let me preface this by stating I'm new at MD simulation.  I am trying to 
> perform a simulation which involves non-bonded copper atoms, amongst 
> other things.  I'm making use of the Lennard-Jones potential to describe 
> the non-bonded interactions between them, and I've input the parameters 
> in the ff*nb.itp file after ensuring Cu was defined in the ffgmx.atp 
> file.  I use x2top to generate the topology file, but the program stops 
> on a fatal error telling me "No forcefield type for atom Cu (841) with 0 
> bonds".  Can any of you out there offer any advice on the cause of said 
> error?  Thus far my troubleshooting has failed.

As a general rule, don't use ffgmx. Like it says in the manual and 
pdb2gmx, its usage is deprecated.

What you need is a forcefield that models non-bonded copper. That's 
easier said than found. Most forcefields in common use with GROMACS are 
for biochemical simulations, so you may not find anything suitable 
already in GROMACS. Telling us your whole simulation system would have 
been a good idea. Likely, you'll need to search literature for a 
suitable model, which GROMACS is likely to be able to implement. That 
might require a thorough knowledge of chapter 5 of the manual.

Mark