A greedy algorithm for aligning DNA sequences

Z Zhang, S Schwartz, L Wagner… - Journal of Computational …, 2000 - liebertpub.com
Z Zhang, S Schwartz, L Wagner, W Miller
Journal of Computational biology, 2000liebertpub.com
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 …
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.
Mary Ann Liebert