[gmx-users] Re: SEP and HO atom type

[Graham Smith]

> But, the problem is not in the SEP definition. If I run grompp using
> just the protein everything goes fine. But, when the ligand is
> included, which contains the HO atom, grompp says Atomtype 'HO' not
> found!  The ligand was included in the .top file after pdb2gmx for
> the protein, since the ligand topology was made outside gromacs.  in
> the ligand.itp: 

Isn't the problem that you are mixing atom types from ffgmx and ffG43? 
e.g HO is defined only in ffgmx but is equivalent to H in ffG43 (same
mass, same vdw interactions [i.e. none]).  So I think all you need to
do is change atom type HO to H in your GAD file. Well, you may also
need to change some of the others - I haven't checked. You could do
that by finding some atom of the type you are interested in in
ffgmx.rtp, then look at what atom type is used for the same atom in
the corresponding residue in ffG43XX.rtp. 

Graham