Another common problem with coordinate file format conversion involves atom types. TNT defines atom types with four letter names. PDB format uses the one or two letter standard element abbreviation. This means that a PDB file cannot define the ionic state of an atom, which is important when calculating the X-ray scattering from that atom.
The first two columns in the atom name field are reserved for the elemental
type in PDB format. Knowing this convention usually allows
Convert to at least identify the element correctly. However, the PDB
does not follow its own rules. The standard atom labels chosen for
a number of cofactors do not begin with the element symbol for that
atom. For instance, the C 
atom in the ribose ring in ATP is
labeled AC4*. Convert will misidentify this atom, and almost all
the other atoms in an ATP, as having the elemental type ``A''.
To correct the problem you will need to INCLUDE the file
$tntdata/pdb_fixup.dat between the PDB and PUNCH commands.
Diamond and DSN2 formats use integer codes to identify atom types. Unfortunately the definition of the formats do not specify an atom type code for any atom types other than carbon, nitrogen, oxygen, and sulfur. Beyond those atom types your have to create your own standard. Convert allows you to specify the atom type code for each atom type.
For the file formats Diamond and DSN2 you can specify, on the end of the PUNCH command, a table defining the integer code for each atom type found in your model. When reading back the Diamond file you must include this same table of conversion values on the DIAMOND statement. Convert hides this information in the DSN2 format so you are not required to reenter the table when reading back a DSN2 file.