No CrossRef data available.
Published online by Cambridge University Press: 10 February 2011
Alignment algorithms are commonly used to detect and quantify similarities between DNA sequences. We study these algorithms in the framework of a recent theory viewing similarity detection as a geometrical critical phenomenon of directed random walks. We show that the roughness of these random walks governs the fidelity of an alignment, i.e., its ability to capture the correlations between the sequences compared. Criteria for the optimization of alignment algorithms emerge from this theory.