The Brookhaven Protein Data Bank stores atomic coordinate information for protein structures in a column based format. This is designed to be read easily read by FORTRAN programs. Indeed, if you get the format description (from anonymous ftp to ftp.pdb.bnl.gov, the file /pub/format.desc.ps) they show the single input line needed to read each record type. However, I am a C/C++ programmer in the Unix environment. It is a easier for me to deal with field based input than column based ones. If the fields are white space delimited I can easily use awk and perl to manipulate the coordinate information. So I needed some way to convert the ATOM and HETATM records of PDB files from the standard column based format to a field based one and back again. It needed to denote missing fields if they exist. That converter is `pdbcat'.