A new method called the neighbor-joining method is proposed for reconstructing phylogenetic trees from evolutionary distance data. The principle of this method is to find pairs of operational taxonomic units (OTUs [= neighbors]) that minimize the total branch length at each stage of clustering of OTUs starting with a starlike tree. The branch lengths as well as the topology of a parsimonious tree can quickly be obtained by using this method. Using computer simulation, we studied the efficiency of this method in obtaining the correct unrooted tree in comparison with that of five other tree-making methods: the unweighted pair group method of analysis, Farris's method, Sattath and Tversky's method, Li's method, and Tateno et al.'s modified Farris method. The new, neighbor-joining method and Sattath and Tversky's method are shown to be generally better than the other methods.

/pdf/the-neighbor-joining-method-a-new-method-for-reconstructing-j5gk4xv4lj.pdf

The neighbor-joining method: a new method for reconstructing phylogenetic trees.

/pdf/molecular-evolutionary-genetics-analysis-1uvpbl4lte.pdf

Molecular Evolutionary Genetics Analysis

A direct method for calculating expected data from an evolutionary model for two state characters is described. The method uses four vectors p, q, r and s. p and q are the probabilities of a character change on the 2n - 3 edges of a tree T (n is the number of taxa). r and s are properties of the data, are independent of any tree and have 2n-1 entries. For a given tree T, and with specified probabilities (p or q), we determine r, then s, the expected probabilities of each of the 2n-1 possible partitions of taxa. For any tree T the relationship can be inverted. This allows the probabilities of change on the tree, p and q, to be estimated directly from observed data (r or s). These relationships have been used to analyse the behaviour of tree building algorithms under conditions when there are sufficient data. (This is when the tree does not change as more data are collected, i.e., convergence to a single tree.) With equal rates of evolution (i.e., with a molecular clock), we show that for n = 4 taxa, parsimony will always converge to the correct tree, but we give examples with n = 5 where parsimony will converge on an incorrect tree, even for equal rates of evolution. A further example with n = 6 shows convergence to an incorrect tree with equal but arbitrarily small rates of change. We interpret a basic difficulty with parsimony as 'long edges attract.' If there are additional taxa that intersect long edges on the tree, then this effect can be reduced. Some distance methods may also converge to an incorrect tree. (Evolu- tionary trees, evolution, parsimony.) Reconstructing evolutionary trees from events that occurred many millions of years ago is a major intellectual challenge. Two recurring problems have been: the choice of optimality criteria and the estimation of

A Framework for the Quantitative Study of Evolutionary Trees

Lake's evolutionary parsimony (EP) method of constructing a phylogenetic tree is primarily applied to four DNA sequences. In this method, three quantities--X, Y, and Z--that correspond to three possible unrooted trees are computed, and an invariance property of these quantities is used for choosing the best tree. However, Lake's method depends on a number of unrealistic assumptions. We therefore examined the theoretical basis of his method and reached the following conclusions: (1) When the rates of two transversional changes from a nucleotide are unequal, his invariance property breaks down. (2) Even if the rates of two transversional changes are equal, the invariance property requires some additional conditions. (3) When Kimura's two-parameter model of nucleotide substitution applies and the rate of nucleotide substitution varies greatly with branch, the EP method is generally better than the standard maximum-parsimony (MP) method in recovering the correct tree but is inferior to the neighbor-joining (NJ) and a few other distance matrix methods. (4) When the rate of nucleotide substitution is the same or nearly the same for all branches, the EP method is inferior to the MP method even if the proportion of transitional changes is high. (5) When Lake's assumptions fail, his chi2 test may identify an erroneous tree as the correct tree. This happens because the test is not for comparing different trees. (6) As long as a proper distance measure is used, the NJ method is better than the EP and MP methods whether there is a transition/transversion bias or whether there is variation in substitution rate among different nucleotide sites.

Limitations of the evolutionary parsimony method of phylogenetic analysis.

The relative efficiencies of several tree-making methods for obtaining the correct phylogenetic tree were studied by using computer simulation. The methods examined were the Fitch-Margoliash (FM), maximum-parsimony (MP), maximumlikelihood (ML), minimum-evolution (ME), and neighbor-joining (NJ) methods. We simulated the evolutionary changes of six DNA sequences each with a length of either 300 or 600 nucleotides. Both constant and varying rates of nucleotide substitution were considered. The DNA sequences generated were used to reconstruct phylogenetic trees by applying the five tree-making methods, and the trees obtained were compared with the model (correct) tree. This process was repeated 50 times for each case, and the following results were obtained: ( 1) The efficiency of obtaining the correct tree for the FM method was considerably lower than those for the other methods. (2) The NJ and ME methods showed a high performance in obtaining the correct tree, and their relative efficiencies were similar to each other. ( 3) For distance methods (NJ, FM, and ME), the results obtained by using corrected nucleotide substitutions were much better than those obtained by using nucleotide differences when the rate of substitution varied greatly among different branches. (4) The ML method was slightly inferior to the NJ and ME methods when a constant rate of nucleotide substitution was assumed, but it was slightly better than the latter two methods when the evolutionary rate varied drastically among branches. If one considers the computational time involved, the NJ method seems to be a method of choice.

/pdf/relative-efficiencies-of-the-fitch-margoliash-maximum-31x7sngm09.pdf

Relative Efficiencies of the Fitch-Margoliash, Maximum-Parsimony, Maximum-Likelihood, Minimum-Evolution, and Neighbor-joining Methods of Phylogenetic Tree Construction in Obtaining the Correct Tree

The relative merits of four different tree-making methods in obtaining the correct topology were studied by using computer simulation. The methods studied were the unweighted pair-group method with arithmetic mean (UPGMA), Fitch and Margoliash's (FM) method, thd distance Wagner (DW) method, and Tateno et al.'s modified Farris (MF) method. An ancestral DNA sequence was assumed to evolve into eight sequences following a given model tree. Both constant and varying rates of nucleotide substitution were considered. Once the DNA sequences for the eight extant species were obtained, phylogenetic trees were constructed by using corrected (d) and uncorrected (p) nucleotide substitutions per site. The topologies of the trees obtained were then compared with that of the model tree. The results obtained can be summarized as follows: (1) The probability of obtaining the correct rooted or unrooted tree is low unless a large number of nucleotide differences exists between different sequences. (2) When the number of nucleotide substitutions per sequence is small or moderately large, the FM, DW, and MF methods show a better performance than UPGMA in recovering the correct topology. The former group of methods is particularly good for obtaining the correct unrooted tree. (3) When the number of substitutions per sequence is large, UPGMA is at least as good as the other methods, particularly for obtaining the correct rooted tree. (4) When the rate of nucleotide substitution varies with evolutionary lineage, the FM, DW, and MF methods show a better performance in obtaining the correct topology than UPGMA, except when a rooted tree is to be produced from data with a large number of nucleotide substitutions per sequence.(ABSTRACT TRUNCATED AT 250 WORDS)

/pdf/accuracy-of-phylogenetic-trees-estimated-from-dna-sequence-6k8w989kkk.pdf

J Sourdis

Papers

Accuracy of phylogenetic trees estimated from DNA sequence data.