Multiple maxima of likelihood in phylogenetic trees: an analytic approach

Maximum likelihood (ML) is a widely used criterion for selecting optimal evolutionary trees. However, little is known on the nature of the likelihood surface for trees, especially as to the frequency of multiple optima. We initiate an analytic study for identifying sequences that generate multiple optima. We report a new approach to calculating ML directly, which we have used to find large families of sequences that have multiple optima, including sequences with a continuum of optimal points. Such datasets are best supported by different (two or more) phylogenies that vary significantly in their timings of evolutionary events Some standard biological processes can lead to data with multiple optima and consequently the field needs further investigation. Our results imply that hill climbing techniques, as currently implemented in various software packages, cannot guarantee to find the global ML point, even if it is unique.

[1]  D. Penny,et al.  Spectral analysis of phylogenetic data , 1993 .

[2]  Asami,et al.  Towards Resolving the Interordinal Relationships of Placental Mammals , 2001 .

[3]  Dan Pelleg,et al.  Constructing Phylogenies from Quartets: Elucidation of Eutherian Superordinal Relationships , 1998, J. Comput. Biol..

[4]  J. Neyman.  MOLECULAR STUDIES OF EVOLUTION: A SOURCE OF NOVEL STATISTICAL PROBLEMS* , 1971 .

[5]  Mike Steel,et al.  The Maximum Likelihood Point for a Phylogenetic Tree is Not Unique , 1994 .

[6]  M. Kimura,et al.  The neutral theory of molecular evolution. , 1983, Scientific American.

[7]  D Penny,et al.  Parsimony, likelihood, and the role of models in molecular phylogenetics. , 2000, Molecular biology and evolution.

[8]  D. Penny,et al.  Spectral Analysis, Systematic Bias, and the Evolution of Chloroplasts , 1999 .

[9]  Michael D. Hendy,et al.  A combinatorial description of the closest tree algorithm for finding evolutionary trees , 1991, Discret. Math..

[10]  Tandy J. Warnow,et al.  A few logs suffice to build (almost) all trees (I) , 1999, Random Struct. Algorithms.

[11]  J. Felsenstein,et al.  Invariants of phylogenies in a simple case with discrete states , 1987 .

[12]  Mike Steel,et al.  Links between maximum likelihood and maximum parsimony under a simple model of site substitution , 1997 .

[13]  Joseph T. Chang,et al.  Full reconstruction of Markov models on evolutionary trees: identifiability and consistency. , 1996, Mathematical biosciences.

[14]  M. Tristem.  Molecular Evolution — A Phylogenetic Approach. , 2000, Heredity.

[15]  J. S. Rogers,et al.  On the consistency of maximum likelihood estimation of phylogenetic trees from nucleotide sequences. , 1997, Systematic biology.

[16]  Z. Yang,et al.  Complexity of the simplest phylogenetic estimation problem , 2000, Proceedings of the Royal Society of London. Series B: Biological Sciences.

[17]  T. Jukes.  CHAPTER 24 – Evolution of Protein Molecules , 1969 .

[18]  G. Lothian,et al.  Spectral Analysis , 1971, Nature.

[19]  Neil J. A. Sloane,et al.  The theory of error-correcting codes (north-holland , 1977 .

[20]  Nick Goldman,et al.  MAXIMUM LIKELIHOOD INFERENCE OF PHYLOGENETIC TREES, WITH SPECIAL REFERENCE TO A POISSON PROCESS MODEL OF DNA SUBSTITUTION AND TO PARSIMONY ANALYSES , 1990 .

[21]  A. Dress,et al.  Reconstructing the shape of a tree from observed dissimilarity data , 1986 .

[22]  N. Saitou,et al.  Maximum likelihood methods. , 1990, Methods in enzymology.

[23]  S J Willson.  Measuring inconsistency in phylogenetic trees. , 1998, Journal of theoretical biology.

[24]  K. Strimmer,et al.  Quartet Puzzling: A Quartet Maximum-Likelihood Method for Reconstructing Tree Topologies , 1996 .

[25]  M.A. CHARLESTON,et al.  Toward a Characterization of Landscapes of Combinatorial Optimization Problems, with Special Attention to the Phylogeny Problem , 1995, J. Comput. Biol..

[26]  Daniel Barry,et al.  [Statistical Analysis of Hominoid Molecular Evolution]: Rejoinder , 1987 .

[27]  P. Erdös,et al.  A few logs suffice to build (almost) all trees (l): part I , 1997 .

[28]  M. Kimura.  The Neutral Theory of Molecular Evolution: From Lamarck to population genetics , 1983 .

[29]  J. S. Rogers,et al.  Multiple local maxima for likelihoods of phylogenetic trees: a simulation study. , 1999, Molecular biology and evolution.

[30]  Ziheng Yang.  Statistical Properties of the Maximum Likelihood Method of Phylogenetic Estimation and Comparison With Distance Matrix Methods , 1994 .

[31]  D Penny,et al.  A discrete Fourier analysis for evolutionary trees. , 1994, Proceedings of the National Academy of Sciences of the United States of America.