RECOVERING EVOLUTIONARY TREES UNDER A MORE REALISTIC MODEL OF SEQUENCE EVOLUTION

We report a new transformation, the LogDet, that is consistent for seq uences with differing nucleotide composition and that have arisen unde r simple but asymmetric stochastic models of evolution. This transform ation is required because existing methods tend to group sequences on the basis of their nucleotide composition, irrespective of their evolu tionary history. This effect of differing nucleotide frequencies is il lustrated by using a tree-selection criterion on a simple distance mea sure defined solely on the basis of base composition, independent of t he actual sequences. The new LogDet transformation uses determinants o f the observed divergence matrices and works because multiplication of determinants (real numbers) is commutative, whereas multiplication of matrices is not, except in special symmetric cases. The use of determ inants thus allows more general models of evolution with asymmetric ra tes of nucleotide change. The transformation is illustrated on a theor etical data set (where existing methods select the wrong tree) and wit h three biological data sets: chloroplasts, birds/mammals (nuclear), a nd honeybees (mitochondrial). The LogDet transformation reinforces the logical distinction between transformations on the data and tree-sele ction criteria. The overall conclusions from this study are that irreg ular A,C,G,T compositions are an important and possible general cause of patterns that can mislead tree-reconstruction methods, even when hi gh bootstrap values are obtained. Consequently, many published studies may need to be reexamined.