$OPTRST group           (optional, for RUNTYP=OPTFMO)
 
Restart data for FMO geometry optimizations. The data
inside vary for each optimization method, and are supposed
to be taken from a previous run (from the punch file).
 
==========================================================
 
$PDB group                                  (optional)
 
Put here a complete PDB file. It is used currently only for
two purposes.
1. Define segments in the partition analysis (PA), if
INDATP(1)<0. The item of primary importance
in the PDB is the residue number column for ATOM and HETATM
entrees. This number is used as the segment ID (note that
it is permitted to omit some IDs in the PDB; for example,
it is permitted to start numbers from any value). In this
case the ID column is used to differentiate segments, which
will be renumbered starting from 1. If you want to define
a functional group in a ligand as a separate segment, then
simply set the ID column to a new number for atoms in the
functional group you want to split (while keeping the old
ID for the rest of the ligand; repeat ad libitum/nauseam).
Note that currently, you cannot resume a segment.
For example, it is not possible to use
HETATM  305  C11 NCF    21       0.896   2.696  -3.512  1.00  0.00           C
HETATM  306  C7  NCF    22       1.846   1.989  -5.589  1.00  0.00           C
HETATM  307  C13 NCF    22       2.153   2.786  -2.952  1.00  0.00           C
HETATM  308  C4  NCF    21       3.270   2.477  -3.718  1.00  0.00           C
because segment 21 is intervened with 22 and then resumed again.
In this case either change the order of atoms as in
HETATM  305  C11 NCF    21       0.896   2.696  -3.512  1.00  0.00           C
HETATM  306  C4  NCF    21       3.270   2.477  -3.718  1.00  0.00           C
HETATM  307  C7  NCF    22       1.846   1.989  -5.589  1.00  0.00           C
HETATM  308  C13 NCF    22       2.153   2.786  -2.952  1.00  0.00           C
or, instead of $PDB, use INDATP to define segments.
Possibly, this restriction will be removed in future.
 
2. For RUNTYP=OPTIMIZE if you provided a PDB in the input,
then a new PDB will be written to the PUNCH file with the
optimized coordinates.
 
In addition to that, atom and residue names are also read,
and used as labels (without affecting the results).
Entrees other than ATOM/HETATM are ignored.
A PDB is never used to define fragments in FMO.
NOTE: atomic coordinates in the PDB are ALWAYS ignored;
$FMOXYZ is always used to define coordinates.