Simple (wrong) models for complex trees: a case from retroviridae

Molecular systematics of the Philippine forest skinks (Squamata: Scincidae: Sphenomorphus): testing morphological hypotheses of interspecific relationships

Skinks of the genus Sphenomorphus are the most diverse clade of squamates in the Philippine Archipelago. Morphological examination of these species has defined six phenotypic groups that are commonly used in characterizations of taxonomic hypotheses. We used a molecular phylogeny based on four mitochondrial and two nuclear genes to assess the group's biogeographical history in the archipelago and examine the phylogenetic validity of the currently recognized Philippine species groups. We re‐examined traditional characters used to define species groups and used multivariate statistics to quantitatively evaluate group structure in morphometric space. Clustering analyses of phenotypic similarity indicate that some (but not all) members of previously defined species groups are phenotypically most similar to other members of the same group. However, when species group membership was mapped on our partitioned Bayesian phylogenetic hypothesis, only one species group corresponds to a clade; all other species group arrangements are strongly rejected by our phylogeny. Our results demonstrate that (1) previously recognized species group relationships were misled by phenotypic convergence; (2) Sphenomorphus is widely paraphyletic; and (3) multiple lineages have independently invaded the Philippines. Based on this new perspective on the phylogenetic relationships of Philippine Sphenomorphus, we revise the archipelago's diverse assemblage of species at the generic level, and resurrect and/or expand four previously recognized genera, and describe two new genera to accommodate the diversity of Philippine skinks of the Sphenomorphus group.

© 2011 The Linnean Society of London, Zoological Journal of the Linnean Society, 2011, 163, 1217–1243.

Using MODELTEST and PAUP* to select a model of nucleotide substitution

Models of nucleotide substitution are commonly used in the analysis of DNA sequences. This unit describes the use of the program MODELTEST (coupled with PAUP*) to find the best-fit model of substitution for the sequence alignment at hand. An example data file is analyzed and the interpretation of the results is discussed. Some background theory on model selection and a discussion of the relevance of models is included at the end of the unit.

Phylogenomic analysis of the L1 retrotransposons in Deuterostomia

L1 retrotransposons constitute the largest single component of mammalian genomes. In contrast to the single remaining lineage of L1 retrotransposons in mammalian genomes, some teleost fishes contain a highly diverse L1 retrotransposon repertoire. Major evolutionary changes in L1 retrotransposon repertoires have therefore taken place in the land vertebrates (Tetrapoda). The lack of sequence data for L1 retrotransposons in the basal living Tetrapoda lineages prompted an investigation of their distribution and evolution in the genomes of the key tetrapod lineages, amphibians and reptiles, and in lungfishes. In this study, we combined genome database searches with PCR analysis to demonstrate that L1 retrotransposons are present in the genomes of lungfishes, amphibians, and lepidosaurs. Phylogenomic analysis shows that the genomes of Deuterostomia possess three highly divergent groups of L1 retrotransposons, with distinct distribution patterns. The analysis of L1 diversity shows the presence of a very large number of diverse L1 families, each with very low copy numbers, at the time of the origin of tetrapods. During the evolution of synapsids, all but one L1 lineage have been lost. This study establishes that the loss of L1 diversity and explosion in copy numbers occurred in the synapsid ancestors of mammals, and was most probably caused by severe population bottlenecks.

Phylogenetic analyses of parasites in the new millennium

Phylogenetic analysis has changed greatly in the last decade, and the most important themes in that change are reviewed here. Sequence data have become the most common source of phylogenetic information. This means that explicit models for evolutionary processes have been developed in a likelihood context, which allow more realistic data analyses. These models are becoming increasingly complex, both for nucleotides and for amino acid sequences, and so all such models need to be quantitatively assessed for each data set, to find the most appropriate one for use in any particular tree-building analysis. Bayesian analysis has been developed for tree-building and is greatly increasing in popularity. This is because a good heuristic strategy exists, which allows large data sets to be analyzed with complex evolutionary models in a practical time. Perhaps the most disappointing aspect of tree interpretation is the ongoing confusion between rooted and unrooted trees, while the effect of taxon and character sampling is often overlooked when constructing a phylogeny (especially in parasitology). The review finishes with a detailed consideration of the analysis of a multi-gene data set for several dozen taxa of Cryptosporidium (Apicomplexa), illustrating many of the theoretical and practical points highlighted in the review.

Missing data and the design of phylogenetic analyses

Concerns about the deleterious effects of missing data may often determine which characters and taxa are included in phylogenetic analyses. For example, researchers may exclude taxa lacking data for some genes or exclude a gene lacking data in some taxa. Yet, there may be very little evidence to support these decisions. In this paper, I review the effects of missing data on phylogenetic analyses. Recent simulations suggest that highly incomplete taxa can be accurately placed in phylogenies, as long as many characters have been sampled overall. Furthermore, adding incomplete taxa can dramatically improve results in some cases by subdividing misleading long branches. Adding characters with missing data can also improve accuracy, although there is a risk of long-branch attraction in some cases. Consideration of how missing data does (or does not) affect phylogenetic analyses may allow researchers to design studies that can reconstruct large phylogenies quickly, economically, and accurately.

Origin and Evolution of the Chloroplast trnK (matK) Intron: A Model for Evolution of Group II Intron RNA Structures

The trnK intron of plants encodes the matK open reading frame (ORF), which has been used extensively as a phylogenetic marker for classification of plants. Here we examined the evolution of the trnK intron itself as a model for group II intron evolution in plants. Representative trnK intron sequences were compiled from species spanning algae to angiosperms, and four introns were newly sequenced. Phylogenetic analyses showed that the matK ORFs belong to the ML (mitochondrial-like) subclass of group II intron ORFs, indicating that they were derived from a mobile group II intron of the class. RNA structures of the introns were folded and analyzed, which revealed progressive RNA structural deviations and degenerations throughout plant evolution. The data support a model in which plant organellar group II introns were derived from bacterial-like introns that had "standard" RNA structures and were competent for self-splicing and mobility and that subsequently the ribozyme structures degenerated to ultimately become dependent upon host-splicing factors. We propose that the patterns of RNA structure evolution seen for the trnK intron will apply to the other group II introns in plants.

Divergent histories of rDNA group I introns in the lichen family Physciaceae

The wide but sporadic distribution of group I introns in protists, plants, and fungi, as well as in eubacteria, likely resulted from extensive lateral transfer followed by differential loss. The extent of horizontal transfer of group I introns can potentially be determined by examining closely related species or genera. We used a phylogenetic approach with a large data set (including 62 novel large subunit [LSU] rRNA group I introns) to study intron movement within the monophyletic lichen family Physciaceae. Our results show five cases of horizontal transfer into homologous sites between species but do not support transposition into ectopic sites. This is in contrast to previous work with Physciaceae small subunit (SSU) rDNA group I introns where strong support was found for multiple ectopic transpositions. This difference in the apparent number of ectopic intron movements between SSU and LSU rDNA genes may in part be explained by a larger number of positions in the SSU rRNA, which can support the insertion and/or retention of group I introns. In contrast, we suggest that the LSU rRNA may have fewer acceptable positions and therefore intron spread is limited in this gene.

Model selection and model averaging in phylogenetics: advantages of akaike information criterion and bayesian approaches over likelihood ratio tests

Model selection is a topic of special relevance in molecular phylogenetics that affects many, if not all, stages of phylogenetic inference. Here we discuss some fundamental concepts and techniques of model selection in the context of phylogenetics. We start by reviewing different aspects of the selection of substitution models in phylogenetics from a theoretical, philosophical and practical point of view, and summarize this comparison in table format. We argue that the most commonly implemented model selection approach, the hierarchical likelihood ratio test, is not the optimal strategy for model selection in phylogenetics, and that approaches like the Akaike Information Criterion (AIC) and Bayesian methods offer important advantages. In particular, the latter two methods are able to simultaneously compare multiple nested or nonnested models, assess model selection uncertainty, and allow for the estimation of phylogenies and model parameters using all available models (model-averaged inference or multimodel inference). We also describe how the relative importance of the different parameters included in substitution models can be depicted. To illustrate some of these points, we have applied AIC-based model averaging to 37 mitochondrial DNA sequences from the subgenus Ohomopterus(genus Carabus) ground beetles described by Sota and Vogler (2001).

Repeatability of clades as a criterion of reliability: a case study for molecular phylogeny of Acanthomorpha (Teleostei) with larger number of taxa

Although much progress has been made recently in teleostean phylogeny, relationships among the main lineages of the higher teleosts (Acanthomorpha), containing more than 60% of all fish species, remain poorly defined. This study represents the most extensive taxonomic sampling effort to date to collect new molecular characters for phylogenetic analysis of acanthomorph fishes. We compiled and analyzed three independent data sets, including: (i) mitochondrial ribosomal fragments from 12S and 16s (814bp for 97 taxa); (ii) nuclear ribosomal 28S sequences (847bp for 74 taxa); and (iii) a nuclear protein-coding gene, rhodopsin (759bp for 86 taxa). Detailed analyses were conducted on each data set separately and the principle of taxonomic congruence without consensus trees was used to assess confidence in the results as follows. Repeatability of clades from separate analyses was considered the primary criterion to establish reliability, rather than bootstrap proportions from a single combined (total evidence) data matrix. The new and reliable clades emerging from this study of the acanthomorph radiation were: Gadiformes (cods) with Zeioids (dories); Beloniformes (needlefishes) with Atheriniformes (silversides); blenioids (blennies) with Gobiesocoidei (clingfishes); Channoidei (snakeheads) with Anabantoidei (climbing gouramies); Mastacembeloidei (spiny eels) with Synbranchioidei (swamp-eels); the last two pairs of taxa grouping together, Syngnathoidei (aulostomids, macroramphosids) with Dactylopteridae (flying gurnards); Scombroidei (mackerels) plus Stromatoidei plus Chiasmodontidae; Ammodytidae (sand lances) with Cheimarrhichthyidae (torrentfish); Zoarcoidei (eelpouts) with Cottoidei; Percidae (perches) with Notothenioidei (Antarctic fishes); and a clade grouping Carangidae (jacks), Echeneidae (remoras), Sphyraenidae (barracudas), Menidae (moonfish), Polynemidae (threadfins), Centropomidae (snooks), and Pleuronectiformes (flatfishes).

A novel use of equilibrium frequencies in models of sequence evolution

Current mathematical models of amino acid sequence evolution are often applied in variants that match their expected amino acid frequencies to those observed in a data set under analysis. This has been achieved by setting the instantaneous rate of replacement of a residue i by another residue j proportional to the observed frequency of the resulting residue j. We describe a more general method that maintains the match between expected and observed frequencies but permits replacement rates to be proportional to the frequencies of both the replaced and resulting residues, raised to powers other than 1. Analysis of a database of amino acid alignments shows that the description of the evolutionary process in a majority (approximately 70% of 182 alignments) is significantly improved by use of the new method, and a variety of analyses indicate that parameter estimation with the new method is well-behaved. Improved evolutionary models increase our understanding of the process of molecular evolution and are often expected to lead to improved phylogenetic inferences, and so it seems justified to consider our new variants of existing standard models when performing evolutionary analyses of amino acid sequences. Similar methods can be used with nucleotide substitution models, but we have not found these to give corresponding significant improvements to our ability to describe the processes of nucleotide sequence evolution.

A reanalysis of the ancient mitochondrial DNA sequences recovered from Neandertal bones

Recent reports analyzing mitochondrial DNA sequences from Neandertal bones have claimed that Neandertals and modern humans are different species. The phylogenetic analyses carried out in these articles did not take into account the high substitution rate variation among sites observed in the human mitochondrial D-loop region and also lack an estimation of the parameters of the nucleotide substitution model. The separate phylogenetic position of Neandertals is not supported when these factors are considered. Our analysis shows that Neandertal-Human and Human-Human pairwise distance distributions overlap more than what previous studies suggested. We also show that the most ancient Neandertal HVI region is the most divergent when compared with modern human sequences. However, the opposite would be expected if the sequence had not been modified since the death of the specimen. Such incongruence is discussed in the light of diagenetic modifications in ancient Neandertal DNA sequences.

Selecting models of nucleotide substitution: an application to human immunodeficiency virus 1 (HIV-1)

The blind use of models of nucleotide substitution in evolutionary analyses is a common practice in the viral community. Typically, a simple model of evolution like the Kimura two-parameter model is used for estimating genetic distances and phylogenies, either because other authors have used it or because it is the default in various phylogenetic packages. Using two statistical approaches to model fitting, hierarchical likelihood ratio tests and the Akaike information criterion, we show that different viral data sets are better explained by different models of evolution. We demonstrate our results with the analysis of HIV-1 sequences from a hierarchy of samples; sequences within individuals, individuals within subtypes, and subtypes within groups. We also examine results for three different gene regions: gag, pol, and env. The Kimura two-parameter model was not selected as the best-fit model for any of these data sets, despite its widespread use in phylogenetic analyses of HIV-1 sequences. Furthermore, the model complexity increased with increasing sequence divergence. Finally, the molecular-clock hypothesis was rejected in most of the data sets analyzed, throwing into question clock-based estimates of divergence times for HIV-1. The importance of models in evolutionary analyses and their repercussions on the derived conclusions are discussed.