James S. Farris

TNT, a free program for phylogenetic analysis

The main features of the phylogeny program TNT are discussed. Windows versions have a menu interface, while Macintosh and Linux versions are command‐driven. The program can analyze data sets with discrete (additive, non‐additive, step‐matrix) as well as continuous characters (evaluated with Farris optimization). Effective analysis of large data sets can be carried out in reasonable times, and a number of methods to help identifying wildcard taxa in the case of ambiguous data sets are implemented. A variety of methods for diagnosing trees and exploring character evolution is available in TNT, and publication‐quality tree‐diagrams can be saved as metafiles. Through the use of a number of native commands and a simple but powerful scripting language, TNT allows the user an enormous flexibility in phylogenetic analyses or simulations.

PARSIMONY JACKKNIFING OUTPERFORMS NEIGHBOR-JOINING

Abstract— Because they are designed to produced just one tree, neighbor‐joining programs can obscure ambiguities in data. Ambiguities can be uncovered by resampling, but existing neighbor‐joining programs may give misleading bootstrap frequencies because they do not suppress zero‐length branches and/or are sensitive to the order of terminals in the data. A new procedure, parsimony jackknifing, overcomes these problems while running hundreds of times faster than existing programs for neighbor‐joining bootstrapping. For analysis of large matrices, parsimony jackknifing is hundreds of thousands of times faster than extensive branch‐swapping, yet is better able to screen out poorly‐supported groups.

A Successive Approximations Approach to Character Weighting

Abstract Fa~ris, J. S. (Dept. Biol. Sci., State Uniu., Stony Brook, New York 11790) 1969. A successive approximations approach to charaaer weighting. Syst. Zool., 18:374385.-Characters that are reliable for cladistic inference are those that are consistent with the true phyletic relationships, that is, those that have little homoplasy. A set of cladistically reliable characters are correlated with each other in a particular non-linear fashion here referred to as hierarchic correlation. Cladistically unreliable characters can be hierarchically correlated only by chance. A technique that infers cladistic relationships by successively weighting characters according to apparent cladistic reliability is suggested, and computer simulation tests of the technique are described. Results indicate that the successive weighting procedure can be highly successful, even when cladistically reliable characters are heavily outnumbered by unreliable ones. [Evolutionary taxonomy. Cladistics. Char-acter weighting.]

Improvements to resampling measures of group support

Several aspects of current resampling methods to assess group support are reviewed. When the characters have different prior weights or some state transformation costs are different, the frequencies under either bootstrapping or jackknifing can be distorted, producing either under‐ or overestimations of the actual group support. This is avoided by symmetric resampling, where the probability p of increasing the weight of a character equals the probability of decreasing it. Problems with interpreting absolute group frequencies as a measure of the support are discussed; group support does not necessarily vary with the frequency itself, since in some cases groups with positive support may have much lower frequencies than groups with no support at all. Three possible solutions for this problem are suggested. The first is measuring the support as the difference in frequency between the group and its most frequent contradictory group. The second is calculating frequencies for values of p below the threshold under which the frequency ranks the groups in the right order of support (this threshold may vary from data set to data set). The third is estimating the support by using the slope of the frequency as a function of different (low) values of p; when p is low, groups with actual support have negative slopes (closer to 0 when the support is higher), and groups with no support have positive slopes (larger when evidence for and against the group is more abundant).

SKEWNESS AND PERMUTATION

Abstract— The skewness criterion of phylogenetic structure in data is too sensitive to character state frequencies, is not sensitive enough to number of characters (degree of corroboration) and relies on counts of arbitrarily‐resolved bifurcating trees. For these reasons it can give misleading results. Permutation tests lack those drawbacks and can be performed quickly by using approximate parsimony calculations, but the test based on minimal tree length can imply strong structure in ambiguous data. A more satisfactory test is obtained by using a support measure which takes multiple trees into account.

A Numerical Approach to Phylogenetic Systematics

Abstract Farris, J. S. (Biol. Sci., State Unia., Stony Brook, New York, 11790), Kluge, A. G., and Eckardt, Jl. J. (Zool., Univ. Alichigan, Ann Arbor 48104) 1970. A Numerical approach to phylogenetic systematics. Syst. Zool., 19:172-191.-Principles abstracted from Hennig (1966) are used as axioms to form a quantitative analog of phylogenetic systematics. A close connection is demonstrated betneen phylogenetics and most parsinlonious trees. The compatibility of some existing clustering methods nith the principles is discussed, and a nen clustering technique, the Weighted Invariant Step Strategy ( WISS) is described. Generalization of the axioms to the case \there direction of evolution is not assumed is examined, and it is shown that the Wagner Method for estimating evolutionary trees is consistent nith the generalized phylogenetic axioms. The taxonomic philosophy propounded in Hennigs ( 1966) Plz ylogenetic Systematics

Methods for Quick Consensus Estimation

A method that allows estimating consensus trees without exhaustive searches is described. The method consists of comparing the results of different independent superficial searches. The results of the searches are then summarized through a majority rule, consensed with the strict consensus tree of the best trees found overall. This assumes that to the extent that a group is recovered by most searches, it is more likely to be actually supported by the data. The effect of different parameters on the accuracy and reliability of the results is discussed. Increasing the cutoff frequency decreases the number of spurious groups, although it also decreases the number of correct nodes recovered. Collapsing trees during swapping reduces the number of spurious groups without significantly decreasing the number of correct nodes recovered. A way to collapse branches considering suboptimal trees is described, which can be extended as a measure of relative support for groups; the relative support is based on the Bremer support, but takes into account relative amounts of favorable and contradictory evidence. More exhaustive searches increase the number of correct nodes recovered, but leave unaffected (or increase) the number of spurious groups. Within some limits, the number of replications does not strongly affect the accuracy of the results, so that using relatively small numbers of replications normally suffices to produce a reliable estimation.

Gunnerales are sister to other core eudicots: implications for the evolution of pentamery

Phylogenetic relationships among many lineages of angiosperms have been clarified via the analysis of large molecular data sets. However, with a data set of three genes (18S rDNA, rbcL, and atpB), relationships among lineages of core eudicots (Berberidopsidales, Caryophyllales, Gunnerales, Santalales, Saxifragales, asterids, rosids) remain essentially unresolved. We added 26S rDNA sequences to a three-gene matrix for 201 eudicots (8430 base pair aligned nucleotides per taxon). Parsimony analyses provided moderate (84%) jackknife support for Gunnerales, which comprise the two enigmatic families Gunneraceae and Myrothamnaceae, as sister to all other core eudicots. This position of Gunnerales has important implications for floral evolution. A dimerous or trimerous perianth is frequently encountered in early-diverging eudicots (e.g., Buxaceae, Proteales, Ranunculales, Trochodendraceae), whereas in core eudicots, pentamery predominates. Significantly, dimery is found in Gunneraceae and perhaps Myrothamnaceae (the merosity of the latter has also been interpreted as labile). Parsimony reconstructions of perianth merosity demonstrate lability among early-diverging eudicots and further indicate that a dimerous perianth could be the immediate precursor to the pentamerous condition characteristic of core eudicots. Thus, the developmental canalization that yielded the pentamerous condition of core eudicots occurred after the node leading to Gunnerales.

Homoplasy Increases Phylogenetic Structure

According to currently accepted theories, rapidly evolving nucleotide sites are phylogenetically less informative than more slowly evolving ones, especially for recognizing more ancient groupings. For this reason third codon positions are often regarded as less reliable than first and second positions as indicators of phylogeny. Analysis of the largest nucleotide matrix treated to date—2538 rbc L sequences covering all major lineages of green plants—shows the opposite: although rapidly evolving and highly homoplastic, third positions contain most of the phylogenetic structure in the data. Frequency of change should thus be used with caution as a criterion for weighting or selecting characters.

Phylogenetic analysis of 73 060 taxa corroborates major eukaryotic groups

Obtaining a well supported schema of phylogenetic relationships among the major groups of living organisms requires considering as much taxonomic diversity as possible, but the computational cost of calculating large phylogenies has so far been a major obstacle. We show here that the parsimony algorithms implemented in TNT can successfully process the largest phylogenetic data set ever analysed, consisting of molecular sequences and morphology for 73 060 eukaryotic taxa. The trees resulting from molecules alone display a high degree of congruence with the major taxonomic groups, with a small proportion of misplaced species; the combined data set retrieves these groups with even higher congruence. This shows that tree‐calculation algorithms effectively retrieve phylogenetic history for very large data sets, and at the same time provides strong corroboration for the major eukaryotic lineages long recognized by taxonomists.