A review of long-branch attraction

The history of long‐branch attraction, and in particular methods suggested to detect and avoid the artifact to date, is reviewed. Methods suggested to avoid LBA‐artifacts include excluding long‐branch taxa, excluding faster evolving third codon positions, using inference methods less sensitive to LBA such as likelihood, the Aguinaldo et al. approach, sampling more taxa to break up long branches and sampling more characters especially of another kind, and the pros and cons of these are discussed. Methods suggested to detect LBA are numerous and include methodological disconcordance, RASA, separate partition analyses, parametric simulation, random outgroup sequences, long‐branch extraction, split decomposition and spectral analysis. Less than 10 years ago it was doubted if LBA occurred in real datasets. Today, examples are numerous in the literature and it is argued that the development of methods to deal with the problem is warranted. A 16 kbp dataset of placental mammals and a morphological and molecular combined dataset of gall waSPS are used to illustrate the particularly common problem of LBA of problematic ingroup taxa to outgroups. The preferred methods of separate partition analysis, methodological disconcordance, and long branch extraction are used to demonstrate detection methods. It is argued that since outgroup taxa almost always represent long branches and are as such a hazard towards misplacing long branched ingroup taxa, phylogenetic analyses should always be run with and without the outgroups included. This will detect whether only the outgroup roots the ingroup or if it simultaneously alters the ingroup topology, in which case previous studies have shown that the latter is most often the worse. Apart from that LBA to outgroups is the major and most common problem; scanning the literature also detected the ill advised comfort of high support values from thousands of characters, but very few taxa, in the age of genomics. Taxon sampling is crucial for an accurate phylogenetic estimate and trust cannot be put on whole mitochondrial or chloroplast genome studies with only a few taxa, despite their high support values. The placental mammal example demonstrates that parsimony analysis will be prone to LBA by the attraction of the tenrec to the distant marsupial outgroups. In addition, the murid rodents, creating the classic “the guinea‐pig is not a rodent” hypothesis in 1996, are also shown to be attracted to the outgroup by nuclear genes, although including the morphological evidence for rodents and Glires overcomes the artifact. The gall wasp example illustrates that Bayesian analyses with a partition‐specific GTR + Γ + I model give a conflicting resolution of clades, with a posterior probability of 1.0 when comparing ingroup alone versus outgroup rooted topologies, and this is due to long‐branch attraction to the outgroup. © The Willi Hennig Society 2005.     

Different gymnosperm outgroups have (mostly) congruent signal regarding the root of flowering plant phylogeny

We examined multiple plastid genes from a diversity of gymnosperm lineages to explore the consistency of signal among different outgroups for rooting flowering plant phylogeny. For maximum parsimony (MP), most outgroups attach on a branch of the underlying ingroup tree that leads to Amborella. Maximum likelihood (ML) analyses either root angiosperms on a nearby branch or find split support for these neighboring root placements, depending on the outgroup. The inclusion of two species of Hydatellaceae, recently recognized as an ancient line of angiosperms, does not aid in inference of the root. Cost profiles for placing the root in suboptimal locations are highly correlated across most outgroup comparisons, even comparing MP and ML profiles. Those for Gnetales are the most deviant of all those considered. This divergent outgroup either attaches on a long eudicot branch with moderate bootstrap support in MP analyses or supports no particular root location in ML analysis. Removing the most rapidly evolving sites in rate classifications based on two divergent angiosperm root placements with Gnetales yields strongly conflicting root placements in MP analysis, despite substantial overlap in the estimated sets of conservative sites. However, the generally high consistency in rooting signal among distantly related gymnosperm clades suggests that the long branch connecting angiosperms to their extant relatives may not interfere substantially with inference of the angiosperm root.     

Phylogenetics of Chondrichthyes and the problem of rooting phylogenies with distant outgroups

Erroneous estimates of ingroup relationships can be caused by attributes in the outgroup chosen to root the tree. Phylogenetic analyses of DNA sequences frequently yield incorrect estimates of ingroup relationships when the outgroup used to "root" the tree is highly divergent from the ingroup. This is especially the case when the outgroup has a different base composition than the ingroup. Unfortunately, in many instances, alternative less divergent outgroups are not available. In such cases, investigators must either target genes with attributes that minimize the problem (slowly evolving genes with stationary base compositions--which are often not ideal for estimating relationships among the more closely related ingroup taxa) or use inference models that are explicitly tailored to deal with an attenuated historical signal with a superimposed non-stationary base composition. In this paper we explore the problem both empirically and through simulation. For the empirical component we looked at the phylogenetic relationships among elasmobranch fishes (sharks and rays), a group whose closest living outgroup, the holocephalan Ghost fishes, are separated from the elasmobranchs by more than 100 million years of evolution. We compiled a data set for analysis comprising 10 single-copy nuclear protein-coding genes (12,096 bp) for representatives of the major lineages within elasmobranchs and holocephalans. For the simulation, we used an evolutionary model on a fixed tree topology to generate DNA sequence data sets which varied both in their distance to the outgroup, and in their base compositional difference between ingroup and outgroup. Results from both the empirical data set and the simulation, support the idea that deviation from base compositional stationarity, in conjunction with distance from the root can act in concert to compromise accuracy of estimated relationships within the ingroup. We tested several approaches to mitigate such problems. We found, that excluding genes with overall faster rates and heterogeneous base compositions, while the least sophisticated of the methods evaluated, seemed to be the most effective.     

Random roots and lineage sorting

Lineage sorting has been suggested as a major force in generating incongruent phylogenetic signal when multiple gene partitions are examined. The degree of lineage sorting can be estimated using the coalescent process and simulation studies have also pointed to a major role for incomplete lineage sorting as a factor in phylogenetic inference. Some recent empirical studies point to an extreme role for this phenomenon with up to 50-60% of all informative genes showing incongruence as a result of lineage sorting. Here, we examine seven large multi-partition genome level data sets over a large range of taxonomic representation. We took the approach of examining outgroup choice and its impact on tree topology, by swapping outgroups into analyses with successively larger genetics distances to the ingroup. Our results indicate a linear relationship of outgroup distance with incongruence in the data sets we examined suggesting a strong random rooting effect. In addition, we attempted to estimate the degree of lineage sorting in several large genome level data sets by examining triads of very closely related taxa. This exercise resulted in much lower estimates of incongruent genes that could be the result of lineage sorting, with an overall estimate of around 10% of the total number of genes in a genome showing incongruence as a result of true lineage sorting. Finally we examined the behavior of likelihood and parsimony approaches on the random rooting phenomenon. Likelihood tends to stabilize incongruence as outgroups get further and further away from the ingroup. In one extreme case, likelihood overcompensates for sequence divergence but increases random rooting causing long branch repulsion.     

Phylogeny of marattioid ferns (Marattiaceae): inferring a root in the absence of a closely related outgroup

Closely related outgroups are optimal for rooting phylogenetic trees; however, such ideal outgroups are not always available. A phylogeny of the marattioid ferns (Marattiaceae), an ancient lineage with no close relatives, was reconstructed using nucleotide sequences of multiple chloroplast regions (rps4 + rps4-trnS spacer, trnS-trnG spacer + trnG intron, rbcL, atpB), from 88 collections, selected to cover the broadest possible range of morphologies and geographic distributions within the extant taxa. Because marattioid ferns are phylogenetically isolated from other lineages, and internal branches are relatively short, rooting was problematic. Root placement was strongly affected by long-branch attraction under maximum parsimony and by model choice under maximum likelihood. A multifaceted approach to rooting was employed to isolate the sources of bias and produce a consensus root position. In a statistical comparison of all possible root positions with three different outgroups, most root positions were not significantly less optimal than the maximum likelihood root position, including the consensus root position. This phylogeny has several important taxonomic implications for marattioid ferns: Marattia in the broad sense is paraphyletic; the Hawaiian endemic Marattia douglasii is most closely related to tropical American taxa; and Angiopteris is monophyletic only if Archangiopteris and Macroglossum are included.     

NUCLEIC ACID SEQUENCE PHYLOGENY AND RANDOM OUTGROUPS

Abstract— When divergent taxa are used to root networks, it is assumed that the character stales in the outgroup have historical similarity to those in the ingroup. Yet, if the data are nucleic acid sequences, the character stales shared by a divergent outgroup may be based not on history but on random similarity. A simple procedure is proposed to test this possibility. In the absence of an appropriate outgroup, root position can be estimated with the use of an asymmetrical character transformation matrix. If the matrix is sufficiently biased, it can supply the polarity information usually derived from an outgroup. This outgroup test and rooting procedure are demonstrated with ADH sequences from the genus Drosophila .     

Rooting and dating maples (Acer) with an uncorrelated-rates molecular clock: implications for north American/Asian disjunctions

Simulations suggest that molecular clock analyses can correctly identify the root of a tree even when the clock assumption is severely violated. Clock-based rooting of phylogenies may be particularly useful when outgroup rooting is problematic. Here, we explore relaxed-clock rooting in the Acer/Dipteronia clade of Sapindaceae, which comprises genera of highly uneven species richness and problematic mutual monophyly. Using an approach that does not presuppose rate autocorrelation between ancestral and descendant branches and hence does not require a rooted a priori topology, we analyzed data from up to seven chloroplast loci for some 50 ingroup species. For comparison, we used midpoint and outgroup rooting and dating methods that rely on rooted input trees, namely penalized likelihood, a Bayesian autocorrelated-rates model, and a strict clock. The chloroplast sequences used here reject a single global substitution rate, and the assumption of autocorrelated rates was also rejected. The root was placed between Acer and Dipteronia by all three rooting methods, albeit with low statistical support. Analyses of Acer diversification with a lineage-through-time plot and different survival models, although sensitive to missing data, suggest a gradual decrease in the average diversification rate. The nine North American species of Acer diverged from their nearest relatives at widely different times: eastern American Acer diverged in the Oligocene and Late Miocene; western American species in the Late Eocene and Mid Miocene; and the Acer core clade, including A. saccharum, dates to the Miocene. Recent diversification in North America is strikingly rare compared to diversification in eastern Asia.     

Optimal outgroup analysis

We present and critically examine a statistical criterion for the selection of outgroup taxa for rooting evolutionary trees. The criterion is the amount of phylogenetic signal for the ingroup when the states of the candidate outgroup taxa are assumed to be plesiomorphic relative to the ingroup for the purpose of measuring plesiomorphy content of the outgroup taxon. A statistical measure of rooted, ingroup signal was subjected to a suite of critical tests which indicate that it provides a proxy measure of plesiomorphy content. As the evolutionary distance between the ingroup ancestral node and outgroup taxa increases, the tree-independent measure of signal decreases, tracking the decay in plesiomorphy content and the increase in convergence to the ingroup states. We show that a priori generalizations about optimal outgroup taxon sampling strategies are likely to be misleading, and that testing for the suitability of available outgroup taxon sampling in specific instances is warranted. Software for optimal outgroup analysis is available.     

A multi criterion approach for the selection of optimal outgroups in phylogeny: recovering some support for Mandibulata over Myriochelata using mitogenomics

The choice of an appropriate outgroup is a fundamental prerequisite when the difference between two conflicting phylogenetic hypotheses depends on the position of the root. This is the case for the myriapods that may group either with Pancrustacea forming a clade called Mandibulata, in accordance with morphological characters, or with chelicerates to form Myriochelata (also called Paradoxopoda) as has recently been proposed by mitochondrial and ribosomal RNA gene phylogenies. In order to understand the impact that outgroup choice may have on phylogenetic reconstruction, we have investigated compositional heterogeneity and genetic distance in mtDNA sequences of several different outgroups to the arthropods, selected from deuterostomes, lophotrochozoans and ecdysozoans, and have used them to root a phylogenetically balanced and compositionarily homogeneous arthropod dataset. Results indicate that some outgroups, in particular from lophotrochozoans, nematodes and an onychophoran have G+C content and strand specific biases which are very different from those of arthropods, suggesting that the use of such outgroups may interfere with the stationarity of the model to create a random outgroup effect. We suggest a multi criterion approach for the selection of optimal outgroup species on the basis of (1) low substitution rate, (2) ingroup-like G+C composition, (3) a new strand bias estimator called the skew index, (4) the ability of the outgroup to avoid a "random branching effect" and (5) phylogenetic proximity to arthropods. Inference of phylogeny using various outgroups shows that use of phylogenetically distant and compositionally distinct lophotrochozoans as outgroups strongly supports Myriochelata and use of more closely related, but fast evolving nematodes supports Mandibulata. A dataset comprising multiple ecdysozoan outgroups also supports Mandibulata, unless the compositionally distant Onychophora are included. A group of the best outgroups selected according to our multi criteria selection, and including the most closely related, least genetically distant and most compositionally similar outgroup, a priapulid worm, supports Mandibulata. We conclude that support for the Myriochelata hypothesis from mitochondrial sequences may depend on the nature of the outgroup sequences rather than a true phylogenetic signal. Finally, we advocate a careful analysis and an objective choice of outgroup when dealing with derived sequences, such as mitochondrial genomes.     

Rooting with Multiple Outgroups: Consensus Versus Parsimony

Using outgroup(s) is the most frequent method to root trees. Rooting through unconstrained simultaneous analysis of several outgroups is a favoured option because it serves as a test of the supposed monophyly of the ingroup. When contradiction occurs among the characters of the outgroups, the branching pattern of basal nodes of the rooted tree is dependent on the order of the outgroups listed in the data matrix, that is, on the prime outgroup (even in the case of exhaustive search). Different equally parsimonious rooted trees (=cladograms) can be obtained by permutation of prime outgroups. An alternative to a common implicit practice (select one outgroup to orientate the tree) is that the accepted cladogram is the strict consensus of the different equally parsimonious rooted trees. The consensus tree is less parsimonious but is not hampered with extra assumption such as the choice of one outgroup (or more) among the initial number of outgroup terminals. It also does not show sister-group relations that are ambiguously resolved or not resolved at all.     

Use of the nuclear gene glyceraldehyde 3-phosphate dehydrogenase for phylogeny reconstruction of recently diverged lineages in Mitthyridium (Musci: Calymperaceae)

A portion of the nuclear gene glyceraldehyde 3-phosphate dehydrogenase (gpd) was sequenced in 26 representatives of the paleotropical moss, Mitthyridium, and a group of 20 outgroup taxa to assess its utility for phylogenetic reconstruction compared with the better understood chloroplast markers, rps4 and trnL. Primers based on plant and fungal sequences were designed to amplify gpd in plants universally with the exclusion of fungal contaminants. The piece amplified spanned 4 introns and 3 of 9 exons, based on comparisons with complete sequence from Arabidopsis. Size variation in gpd ranged from 891 to 1007 bp, in part attributable to 6 indels of variable length found within the introns. Intron 6 contributed most of the length variation and contained a variable purine-repeat motif of possible use as a microsatellite. Phylogenetic analyses of the full gpd amplicon yielded well-resolved trees that were in nearly full accord with the trees derived from the cpDNA partitions for analyses of both the ingroup and ingroup + outgroup taxon sets. Pairwise nucleotide substitution rates of gpd were as much as 2.2 times higher than those in rps4 and 2.8 times higher than in trnL. Excision of the introns left suitable numbers of parsimony informative characters and demonstrated that the full gpd amplicon could be compartmentalized to provide resolution for both shallow and deep phylogenetic branches. Exons of gpd were found to behave in a clock-like fashion for the 26 ingroup taxa and select outgroups. In general, gpd was found to hold great promise not only for improving resolution of chloroplast-derived phylogenies, but also for phylogenetic reconstruction of recent, diversifying lineages.     

The problem of rooting rapid radiations

There are many examples of groups (such as birds, bees, mammals, multicellular animals, and flowering plants) that have undergone a rapid radiation. In such cases, where there is a combination of short internal and long external branches, correctly estimating and rooting phylogenetic trees is known to be a difficult problem. In this simulation study, we tested the performances of different phylogenetic methods at estimating a tree that models a rapid radiation. We found that maximum likelihood, corrected and uncorrected neighbor-joining, and corrected and uncorrected parsimony, all suffer from biases toward specific tree topologies. In addition, we found that using a single-taxon outgroup to root a tree frequently disrupts an otherwise correct ingroup phylogeny. Moreover, for uncorrected parsimony, we found cases where several individual trees (in which the outgroup was placed incorrectly) were selected more frequently than the correct tree. Even for parameter settings where the correct tree was selected most frequently when using extremely long sequences, for sequences of up to 60,000 nucleotides the incorrectly rooted trees were each selected more frequently than the correct tree. For all the cases tested here, tree estimation using a two taxon outgroup was more accurate than when using a single-taxon outgroup. However, the ingroup was most accurately recovered when no outgroup was used.     

Was the ANITA rooting of the angiosperm phylogeny affected by long-branch attraction? Amborella, Nymphaeales, Illiciales, Trimeniaceae, and Austrobaileya

Five groups of basal angiosperms, Amborella, Nymphaeales, Illiciales, Trimeniaceae, and Austrobaileya (ANITA), were identified in several recent studies as representing a series of the earliest-diverging lineages of the angiosperm phylogeny. All of these studies except one employed a multigene analysis approach and used gymnosperms as the outgroup to determine the ingroup topology. The high level of divergence between gymnosperms and angiosperms, however, has long been implicated in the difficulty of reconstructing relationships at the base of angiosperm phylogeny using DNA sequences, for fear of long-branch attraction (LBA). In this study, we replaced the gymnosperm sequences from the five-gene matrix (mitochondrial atp1 and matR, plastid atpB and rbcL, and nuclear 18S rDNA) used in our earlier study with four categories of divergent sequences--random sequences with equal base frequencies or equally AT- and GC-rich contents, homopolymers and heteropolymers, misaligned gymnosperm sequences, and aligned lycopod and bryophyte sequences--to evaluate whether the gymnosperms were an appropriate outgroup to angiosperms in our earlier study that identified the ANITA rooting. All 24 analyses performed rooted the angiosperm phylogeny at either Acorus or Alisma (or Alisma-Triglochin-Potamogeton in one case due to use of a slightly different alignment) and placed the monocots as a basal grade, producing genuine LBA results. These analyses demonstrate that the identification of ANITA as the basalmost extant angiosperms was based on historical signals preserved in the gymnosperm sequences and that the gymnosperms were an appropriate outgroup with which to root the angiosperm phylogeny in the multigene sequence analysis. This strategy of evaluating the appropriateness of an outgroup using artificial sequences and a series of outgroups with increments of divergence levels can be applied to investigations of phylogenetic patterns at the bases of other major clades, such as land plants, animals, and eukaryotes.     

Rapid evolutionary divergences in reef fishes of the family Acanthuridae (Perciformes: Teleostei)

A phylogenetic analysis of the sugeonfish family Acanthuridae was conducted to investigate: (a) the pattern of divergences among outgroup and basal ingroup taxa, (b) the pattern of species divergences within acanthurid genera, (c) monophyly in the genus Acanthurus, and (d) the evolution of thick-walled stomach morphology in the genera Acanthurus and Ctenochaetus. Fragments of the 12S, 16S, t-Pro, and control region mitochondrial genes were sequenced for 21 acanthurid taxa (representing all extant genera) and four outgroup taxa. Unweighted parsimony analysis produced two optimal trees. Both of these were highly incongruent with a previous morphological phylogeny, especially with regard to the placement of the monotypic outgroups Zanclus and Luvarus. The maximum likelihood tree and the morphological phylogeny were not significantly different and the conflicting branches were very short. Split decomposition analysis identified conflict in the placement of long basal branches separated by short internodes, providing further evidence that long branch attraction is an important cause of disagreement between molecular and morphological trees. Parametric bootstrapping rejected hypotheses of monophyly of: (a) the genus Acanthurus and (b) a group containing representatives of Acanthurus/Ctenochaetus with thick-walled stomachs. The branching pattern of the likelihood and split decomposition trees indicates that evolution in the acanthurid clade has involved at least three periods of intense speciation.     

Turning the crown upside down: gene tree parsimony roots the eukaryotic tree of life

The first analyses of gene sequence data indicated that the eukaryotic tree of life consisted of a long stem of microbial groups "topped" by a crown-containing plants, animals, and fungi and their microbial relatives. Although more recent multigene concatenated analyses have refined the relationships among the many branches of eukaryotes, the root of the eukaryotic tree of life has remained elusive. Inferring the root of extant eukaryotes is challenging because of the age of the group (～1.7-2.1 billion years old), tremendous heterogeneity in rates of evolution among lineages, and lack of obvious outgroups for many genes. Here, we reconstruct a rooted phylogeny of extant eukaryotes based on minimizing the number of duplications and losses among a collection of gene trees. This approach does not require outgroup sequences or assumptions of orthology among sequences. We also explore the impact of taxon and gene sampling and assess support for alternative hypotheses for the root. Using 20 gene trees from 84 diverse eukaryotic lineages, this approach recovers robust eukaryotic clades and reveals evidence for a eukaryotic root that lies between the Opisthokonta (animals, fungi and their microbial relatives) and all remaining eukaryotes.     

An empirical test of the midpoint rooting method

The outgroup method is widely used to root phylogenetic trees. An accurate root indication, however, strongly depends on the availability of a proper outgroup. An alternate rooting method is the midpoint rooting (MPR). In this case, the root is set at the midpoint between the two most divergent operational taxonomic units. Although the midpoint rooting algorithm has been extensively used, the efficiency of this method in retrieving the correct root remains untested. In the present study, we empirically tested the success rate of the MPR in obtaining the outgroup root for a given phylogenetic tree. This was carried out by eliminating outgroups in 50 selected data sets from 33 papers and rooting the trees with the midpoint method. We were thus able to compare the root position retrieved by each method. Data sets were separated into three categories with different root consistencies: data sets with a single outgroup taxon (54% success rate for MPR), data sets with multiple outgroup taxa that showed inconsistency in root position (82% success rate), and data sets with multiple outgroup taxa in which root position was consistent (94% success rate). Interestingly, the more consistent the outgroup root is, the more successful MPR appears to be. This is a strong indication that the MPR method is valuable, particularly for cases where a proper outgroup is unavailable.  © 2007 The Linnean Society of London, Biological Journal of the Linnean Society , 2007, 92 , 669–674.     

Phylogenetics of Coenonymphina (Nymphalidae: Satyrinae) and the problem of rooting rapid radiations

We report a rapid radiation of a group of butterflies within the family Nymphalidae and examine some aspects of popular analytical methods in dealing with rapid radiations. We attempted to infer the phylogeny of butterflies belonging to the subtribe Coenonymphina sensu lato using five genes (4398 bp) with Maximum Parsimony, Maximum Likelihood and Bayesian analyses. Initial analyses suggested that the group has undergone rapid speciation within Australasia. We further analyzed the dataset with different outgroup combinations the choice of which had a profound effect on relationships within the ingroup. Modelling methods recovered Coenonymphina as a monophyletic group to the exclusion of Zipaetis and Orsotriaena, irrespective of outgroup combination. Maximum Parsimony occasionally returned a polyphyletic Coenonymphina, with Argyronympha grouping with outgroups, but this was strongly dependent on the outgroups used. We analyzed the ingroup without any outgroups and found that the relationships inferred among taxa were different from those inferred when either of the outgroup combinations was used, and this was true for all methods. We also tested whether a hard polytomy is a better hypothesis to explain our dataset, but could not find conclusive evidence. We therefore conclude that the major lineages within Coenonymphina form a near-hard polytomy with regard to each other. The study highlights the importance of testing different outgroups rather than using results from a single outgroup combination of a few taxa, particularly in difficult cases where basal nodes appear to receive low support. We provide a revised classification of Coenonymphina; Zipaetis and Orsotriaena are transferred to the tribe Eritina.     

ON OUTGROUPS

Abstract— The relations among polarity, outgroups and rooting are clarified. The "outgroup algorithm" and "outgroup substitution method" are irrelevant forms of relaxed parsimony. They should be discarded in favor of unconstrained, simultaneous analysis of all terminals. A revised outgroup method is described both in text and with a computer-generated flowchart. Lundberg rooting is consistent with cladistic parsimony only under specific circumstances involving hypothetical ancestors.
"Both sides seemed convinced that the 'real enemy'is a vicious conspiracy of some kind."
Hunter S. Thompson (1979: 145).