The combinatorial extension (CE) method is like the DALI method in that it breaks each structure into a series of fragments and attempts to reassemble them into a complete alignment. Pairwise combinations of fragments, named aligned fragment pairs, or AFPs, are used to make a similarity matrix where an optimal path is generated to identify the final alignment. To reduce the amount of searching and increase efficiency, only AFPs that meet a criteria for local similarity are used in the matrix. The first CE method only included structural superpositions and inter-residue distances but has been grown to include local environmental properties like secondary structure, solvent exposure, hydrogen-bonding patterns, and dihedral angles. An alignment path is the optimal path that is calculated through a similarity matrix by linear progression through the sequences and extending the alignment with a possible high-scoring AFP pair. The initial AFP pair that nucleates the alignment can occur at any place in the sequence matrix. This is then extended with AFPs that meet the given distance criteria which restricts the alignment to low gap sizes. The size of each AFP and maximum gap size are required input parameters that are usually set to the empirically determined values of 8 and 30, respectively. An all-to-all fold classification database from known protein structures in the PDB has been constructed using CE.