A Greedy Algorithm for Aligning DNA Sequences
JOURNAL OF COMPUTATIONAL BIOLOGY Volume 7, Numbers 1/2, 2000
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.
leave a comment