[gmx-users] More dihedrals generated than provided in the force field & "No default Proper Dih. types" as a result

[Timofey Tyugashev]

I'm building a protein with a cubic [Fe4S4] cluster using parameters 
from a published paper.
This parameters provide only two dihedral angles types (S-FE-SG-CB and 
FE-S-FE-SG, where S and FE belong to the cluster and SG and CB belong to 
the Cys residue).
GROMACS automatically generates far more dihedrals with types like 
S-FE-S-FE (looks like all of them are inta-cluster) which are omitted in 
the source paper and are explicitly stated to be redundant for MD. 
However  grommp generates errors and stops:

ERROR 49 [file topol.top, line 32916]:
   No default Proper Dih. types

How to solve this problem?