TOPAL 2.0: improved detection of mosaic sequences within multiple alignments

Motivation: The Dss statistic Mms proposed by McGuire et al. (Mol. Biol. Evol., 14, 1125-1131, 1997) for scanning data sets for the presence of recombination, an important step in some phylogenetic analyses. The statistic, however; could not distinguish well between among-site rate variation and recombination, and had no statistical test for significant values. This paper addresses these shortfalls. Results: A modification to the Dss statistic is proposed which accounts for rate variation to a large extent. A statistical test, based on parametric bootstrapping, is also suggested.