CCL:G: ONIOM Calculations (Protein-Ligand)

 Sent to CCL by: "Fenley, Andrew" [afenley*_*ucsd.edu]
 Hello,
 I am trying to setup 3-layer ONIOM calculations for a protein-ligand binding
 site within Gaussian G09 RevE. The high layer contains the ligand and any whole
 residue within 3 Angstroms of the ligand. The medium layer is any whole residue
 within 5 Angstroms of the ligand that has not been included in the high layer.
 And the low layer consists of everything else. The layer models are: B3LYP
 (6-31G*):PM6:AMBER. One issue I keep running into with regards to optimization
 is that if I specify the ligand to be a ‘frozen’ block (by setting
 -2 in the appropriate field per atom), then I get an error saying the default
 optimizer supports only 1 rigid structure.
 Ideally, what I want to do is, let all hydrogens of the protein (in any layer)
 optimize, along with the ligand optimizing its location within the binding site
 as a rigid block (so just translation / rotation effectively of the whole
 molecule). And the rest of the atomic coordinates of the system stay frozen
 (fixed to the original input coordinates). I thought the way to do this would be
 to specify ‘0’ for all Hydrogens, ‘-2’ for the
 ligand, and ‘-1’ for everything else per input line describing
 each atom right after the element-type-charge column. Attempts to use other
 options to the optimizer like Opt=QuadMacro did not help.
 Additional details:
 The atoms in the input file are represented using cartesian coordinates.
 Geom=connectivity is used, and all the bonding information is provided.
 Additional vdW parameters for the ligand are also provided in the input file
 (taken from GAFF). Partial charges are provided per atom with EmbedCharge set in
 the header of the input file.
 Thank you for your help!
 Andrew
 afenley*o*vt.edu