Pair-wise sequence alignments fall into one of the following categories:

• optimal global alignment -- the best alignment along the entire lengths of the two sequences (or the entire length of the shorter sequence). The most common algorithm is Needleman-Wunsch-Sellers.
• optimal local alignment -- the best aligned segment of two sequences. The most common algorithm is Smith-Waterman.
• suboptimal local alignments -- multiple aligned segments are reported. Common implementations are dotplots and LFASTA.

It is important to realize that the optimal alignment algorithms report only one alignment -- the one with the best mathematical score given the values of the parameters used by the algorithm. (If two alignments have the same score, only one is reported.) This is not necessarily the best biological alignment. There may be other significant alignments between the two sequences. To see these, change the values of the parameters used in the alignment or use a method that reports suboptimal alignments.

Some parameters that affect alignment scores:

• scoring matrix -- assigns scores for matches and mismatches
• gap (gap opening) penalty -- penalty assessed when opening a gap in the alignment
• gap length (gap extension) penalty -- penalty assessed for each residue when extending the length of a gap in the alignment

Pair-wise alignments may also be performed between a DNA sequence and a protein sequence. These are done by translating the DNA sequence in all reading frames, aligning the translations to the protein sequence, and converting the translated sequence back to DNA for the final report. Some of these programs take frame shifts into account and thus are useful in comparing DNA sequences that may contain sequencing errors with known protein sequences.

This website will look much better in a browser that supports web standards, but it has been designed so that it is still usable and accessible to any browser or web-enabled device.