Biomedical Engineering Reference
In-Depth Information
positions in the sequences are identified. A typical window-stringency combination is 15/10 for
nucleotide sequences and much narrower combinations for polypeptide sequences, such as 1/1 or
3/2.
Figure 8-4. Dot Matrix Pairwise Alignment of Sequences (G—top) and
(H—side). The filter, with a window of 2 and stringency of 1, emphasizes
contiguous aligned sequence pairs.
Dot matrix analysis is especially useful in identifying repeats—repeating characters or short
sequences—within a sequence, as in mapping the repeated regions of whole chromosomes. Repeats
of the same character, as in sequence (P), create alignments with artificially high scores and make
sequence alignment more difficult. Dot matrix methods are most applicable to single pairwise
alignment problems, especially those with relative high degrees of similarity. Sequences with lower
degrees of similarity as well as multiple sequence alignment require methods that are more powerful.
Although window-stringency values are often established heuristically, they may also be based on
dynamic averages, scores based on the occurrence of matches in aligned protein families, or on
various methods of scoring the similarity of amino acids. For example, scoring matrices provide
scores for alignment based on their statistical occurrence in aligned protein families. Using these
matrices, described in the following section, a sliding window feature can be implemented, in which
only dots above a certain average score can appear in the matrix.
