Phylogeny Reconstruction Based on Molecular Property Patterns

Summary

An essentially new method to relate taxa of any kind by means of a predefined set of dichotomous properties (i.e., either present or absent) will be presented. The main feature of the analysis is the usage of a sophisticated distance measure to describe the pairwise dissimilarities quantitatively in dependence on the individual properties. The distance measure implies in a natural way the derivation of a tree structure by successive joining of taxa with minimum distance (neighbor joining). The distances are based on the jointly compatible properties and, consequently, they are referred to as the consensus taxa with respect to the properties. Further, the branch lengths of the resulting tree will be calculated simultaneously to the branching structure and is no longer the result of a second procedure. Moreover, the distances can be interpreted by means of a (nonprobabilistic) information concept and negative or other noninterpretable distances never occur. Generally, the proposed tree reconstruction method is not based on stochastic or other mathematical models of the underlying evolutionary processes and can be interpreted best in terms of discrete information theory. Theproposed method is highly flexible and can be applied to data of various types. In the present chapter this is shown by reanalyzing published data and comparison of the trees derived with the property pattern method and with other established methods: amino acid sequences, restriction sites data, gene frequencies, and sensitivities of bacterial protein synthesis to different antibiotics. In all cases we found that the tree derived with the property pattern method is in good accordance with the biological expectation.