Abstract
We present a new robust solution to short peptide searching, tested on a relational platform, with a set of biological queries. Our algorithm is appropriate for large scale scientific data analysis, and has been tested with 1.4 GB of amino-acids. Protein sequences are indexed as short overlapping string windows, and stored in a relation. To find approximate matches, we use a neighbourhood generation algorithm. The words in the neighbourhood are then fetched and stored in a relation. We measure execution time and compare the matches found to those delivered by BLAST. We report some performance gains in exact matching and searching within edit distance 1, and very significant quality improvements over heuristics, as we guarantee to deliver all relevant matches. © Springer-Verlag Berlin Heidelberg 2007.
Cite
CITATION STYLE
Hunt, E. (2007). Exhaustive peptide searching using relations. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4587 LNCS, pp. 13–24). Springer Verlag. https://doi.org/10.1007/978-3-540-73390-4_3
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
