Fast Approximate Search in Strings with Rearrangements

Abstract

One of the types of genovariation is the intrasequence rearrangement. If we consider nucleotide sequence as a string, then the intrasequence rearrangement may be interpreted as mutual transposition of string parts. Common instruments for the analysis of the nucleotide strings (such as BLAST (Altschul et al., 1990)) are based on the computation of edit distance. Edit-distance approach is very efficient in case of such mutations as deletion, insertion or substitution of a single nucleotide, but almost inapplicable for the sequences with rearrangements. Another approach to the approximate searching in strings under condition of sequence rearrangements is block-edit approach (Lopresti and Tomkins, 1997). However algorithms based on this approach have a very high time complexity, which leads to the tight restrictions in its bioinformatics applications. One of the modern methods of approximate searching with rearrangements is reflected in QUASAR system (Burkhardt et al., 1999). In this article an approach to the search is proposed that is similar to QUASAR. The proposed approach gives visualization of the searching results. As distinct from QUASAR the proposed approach does not involve edit distance concept which allows to obtain time complexity O(n middot ln m), where n is the length of data string and m is the length of pattern. Another peculiarity is the visualization of the results