When dealing with the amino acid sequence for a given CATH domain (or PDB chain), it is important to differentiate between the amino acid sequence as defined by the ATOM records and the sequence as defined by the SEQRES records in the PDB file (they are not necessarily the same). We resolve this issue by using an internal NW algorithm to align the SEQRES and ATOM sequences to provide the 'full' amino acid sequence (called COMBS).