Note that the READ SEQUence PDB command has several options that allow you to fine tune what is to be read from the file. Excerpt from io.doc:
The SEGI <segid> option allows the sequence to be read only for the residues
belonging to the corresponding segid in the coordinate file.
For the PDB option resids are always read from the resSeq (resid) field.
This is useful when one wants to specify residue names (rather than use
the number representation). No other information is read from the coordinate
file during this process. To read the sequence for a specific chain in a PDB
file the CHAIn <char> option can be used; <char> is the one letter PDB chain
id in position 22 of ATOM/HETATM records. If the SEGI <segid> option is used
the sequence is read for the atoms that have the corresponding segid in
columns 73-76 of the PDB file. NCHAin <int> starts reading from chain number
<int> as defined by TER separator records. Variables SQNRES and SQRESID are
set to the number of residues read and the SEGId used.

There are simple examples in the Script Archive forum of how to set up a waterbox.

Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden