A Greedy Algorithm for Aligning DNA Sequences
Zheng Zhang, Scott Schwartz, Lukas Wagner, Webb Miller
Abstract
For aligning DNA sequences that differ only by sequencing errors, or by equivalent errors from other sources, a greedy algorithm can be much faster than traditional dynamic programming approaches and yet produce an alignment that is guaranteed to be theoretically optimal. We introduce a new greedy alignment algorithm with particularly good performance and show that it computes the same alignment as does a certain dynamic programming algorithm, while executing over 10 times faster on appropriate data. An implementation of this algorithm is currently used in a program that assembles the UniGene database at the National Center for Biotechnology Information.
Topics & Concepts
Greedy algorithmDynamic programmingComputer scienceAlgorithmUniGeneBiologyGenomeGeneticsExpressed sequence tagGeneGenomics and Phylogenetic StudiesAlgorithms and Data CompressionRNA and protein synthesis mechanisms