%0 Journal Article
%T Fast estimation of the difference between two PAM/JTT evolutionary distances in triplets of homologous sequences
%A Christophe Dessimoz
%A Manuel Gil
%A Adrian Schneider
%A Gaston H Gonnet
%J BMC Bioinformatics
%D 2006
%I BioMed Central
%R 10.1186/1471-2105-7-529
%X We show that an alternative estimator, based on pairwise estimates and therefore much faster to compute, has almost the same statistical power as the maximum likelihood estimator. We also provide a numerical approximation for its variance, which could otherwise only be estimated through an expensive re-sampling approach such as bootstrapping. An extensive simulation demonstrates that the approximation delivers precise confidence intervals. To illustrate the possible applications of these results, we show how they improve the detection of asymmetric evolution, and the identification of the closest relative to a given sequence in a group of homologs.The results presented in this paper constitute a basis for large-scale protein cross-comparisons of pairwise evolutionary distances.The estimation of evolutionary distances between biological sequences is at the basis of many bioinformatics problems: it plays a particularly important role in phylogenetic tree inference [1,2] and in an increasing number of comparative genomics analyses over large sets of genes or proteins (e.g. [3-5]). The most accurate way of estimating evolutionary distances is currently maximum likelihood, but the procedure is so time-consuming that is hardly practical when dealing with large datasets. In such cases, complexity is often tackled by working on the basis of individual pairs, such as in distance tree methods or in the "all-against-all" at the beginning of many comparative genomics analyses. However, by estimating an evolutionary distance for each pair individually, no knowledge about the covariance of distance estimates with common evolution can be directly obtained. Thus, when comparing pairwise distances among related sequences, for instance to infer which of two homologs is closer to a third one, confidence intervals cannot be derived directly from the pairwise estimates.The present article investigates this fundamental problem of estimating the difference between two distances in a tripl
%U http://www.biomedcentral.com/1471-2105/7/529