Local search for the generalized tree alignment problem

Multi-armed bandits, Thomson sampling and unsupervised machine learning in phylogenetic graph search

A phylogenetic graph search relies on a large number of highly parameterized search procedures (e.g. branch-swapping, perturbation, simulated annealing, genetic algorithm). These procedures vary in effectiveness over datasets and at alternative points in analytical pipelines. The multi-armed bandit problem is applied to phylogenetic graph searching to more effectively utilize these procedures. Thompson sampling is applied to a collection of search and optimization "bandits" to favour productive search strategies over those that are less successful. This adaptive random sampling strategy is shown to be more effective in producing heuristically optimal phylogenetic graphs and more time efficient than existing uniform probability randomized search strategies. The strategy acts as a form of unsupervised machine learning that can be applied to a diversity of phylogenetic datasets without prior knowledge of their properties.

PhylogeneticGraph (PhyG) a new phylogenetic graph search and optimization program

We present Phylogenetic Graph (PhyG), an open-source, phylogenetic search tool for diverse data types and graphs, including softwired and hardwired networks, in addition to trees. This allows for analysis of horizontal transfer and hybridization scenarios, as well as the necessary vertical inheritance of trees. PhyG is the successor to POY5 in performing combined data tree-alignment with enhancements in heuristic optimality (up to 7% in example data) and execution time (up to a factor of 200). Input data may exhibit a practically unlimited number of character states in qualitative or sequence (aligned and unaligned) types. Novel graph construction and refinement algorithms have been implemented and integrated into a variety of search procedures. Currently, PhyG implements parsimony and No-Common-Mechanism Likelihood optimization.

The phylogeny of Dendropsophini (Anura: Hylidae: Hylinae)

The relationships of the hyline tribe Dendropsophini remain poorly studied, with most published analyses dealing with few of the species groups of Dendropsophus. In order to test the monophyly of Dendropsophini, its genera, and the species groups currently recognized in Dendropsophus, we performed a total evidence phylogenetic analysis. The molecular dataset included sequences of three mitochondrial and five nuclear genes from 210 terminals, including 12 outgroup species, the two species of Xenohyla, and 93 of the 108 recognized species of Dendropsophus. The phenomic dataset includes 46 terminals, one per species (34 Dendropsophus, one Xenohyla, and 11 outgroup species). Our results corroborate the monophyly of Dendropsophini and the reciprocal monophyly of Dendropsophus and Xenohyla. Some species groups of Dendropsophus are paraphyletic (the D. microcephalus, D. minimus, and D. parviceps groups, and the D. rubicundulus clade). On the basis of our results, we recognize nine species groups; for three of them (D. leucophyllatus, D. microcephalus, and D. parviceps groups) we recognize some nominal clades to highlight specific morphology or relationships and facilitate species taxonomy. We further discuss the evolution of oviposition site selection, where our results show multiple instances of independent evolution of terrestrial egg clutches during the evolutionary history of Dendropsophus.

Evidence of absence treated as absence of evidence: The effects of variation in the number and distribution of gaps treated as missing data on the results of standard maximum likelihood analysis

Although numerous studies have demonstrated the theoretical and empirical importance of treating gaps as insertion/deletion (indel) events in phylogenetic analyses, the standard approach to maximum likelihood (ML) analysis employed in the vast majority of empirical studies codes gaps as nucleotides of unknown identity ("missing data"). Therefore, it is imperative to understand the empirical consequences of different numbers and distributions of gaps treated as missing data. We evaluated the effects of variation in the number and distribution of gaps (i.e., no base, coded as IUPAC "." or "-") treated as missing data (i.e., any base, coded as "?" or IUPAC "N") in standard ML analysis. We obtained alignments with variable numbers and arrangements of gaps by aligning seven diverse empirical datasets under different gap opening costs using MAFFT. We selected the optimal substitution model for each alignment using the corrected Akaike Information Criterion in jModelTest2 and searched for optimal trees using GARLI. We also employed a Monte Carlo approach to randomly replace nucleotides with gaps (treated as missing data) in an empirical dataset to understand more precisely the effects of varying their number and distribution. To compare alignments, we developed four new indices and used several existing measures to quantify the number and distribution of gaps in all alignments. Our most important finding is that ML scores correlate negatively with gap opening costs and the amount of missing data. However, this negative relationship is not due to the increase in missing data per se-which increases ML scores-but instead to the effect of gaps on nucleotide homology. These variables also cause significant but largely unpredictable effects on tree topology.

A total evidence analysis of the phylogeny of hatchet-faced treefrogs (Anura: Hylidae: Sphaenorhynchus)

The Neotropical hylid genus Sphaenorhynchus includes 15 species of small, greenish treefrogs widespread in the Amazon and Orinoco basins, and in the Atlantic Forest of Brazil. Although some studies have addressed the phylogenetic relationships of the genus with other hylids using a few exemplar species, its internal relationships remain poorly understood. In order to test its monophyly and the relationships among its species, we performed a total evidence phylogenetic analysis of sequences of three mitochondrial and three nuclear genes, and 193 phenotypic characters from all species of Sphaenorhynchus. Our results support the monophyly of Sphaenorhynchus with molecular and phenotypic evidence, with S. pauloalvini as the earliest diverging taxon, followed by S. carneus, as the sister taxon of all remaining species of the genus. We recognize three species groups in Sphaenorhynchus (the S. lacteus, S. planicola and S. platycephalus groups), to facilitate its taxonomic study; only three species (S. carneus, S. pauloalvini and S. prasinus) remain unassigned to any group. Sequence data were not available for only two species (S. bromelicola and S. palustris) for which we scored phenotypic data; wildcard behaviour was detected only in S. bromelicola nested inside the S. platycephalus group. On the basis of the resulting phylogenetic hypothesis, we discuss the evolution of oviposition site and a number of phenotypic characters that could be associated with heterochronic events in the evolutionary history of this group.

Revising the Bantu tree

Phylogenetic methods offer a promising advance for the historical study of language and cultural relationships. Applications to date, however, have been hampered by traditional approaches dependent on unfalsifiable authority statements: in this regard, historical linguistics remains in a similar position to evolutionary biology prior to the cladistic revolution. Influential phylogenetic studies of Bantu languages over the last two decades, which provide the foundation for multiple analyses of Bantu sociocultural histories, are a major case in point. Comparative analyses of basic lexica, instead of directly treating written words, use only numerical symbols that express non-replicable authority opinion about underlying relationships. Building on a previous study of Uto-Aztecan, here we analyse Bantu language relationships with methods deriving from DNA sequence optimization algorithms, treating basic vocabulary as sequences of sounds. This yields finer-grained results that indicate major revisions to the Bantu tree, and enables more robust inferences about the history of Bantu language expansion and/or migration throughout sub-Saharan Africa. "Early-split" versus "late-split" hypotheses for East and West Bantu are tested, and overall results are compared to trees based on numerical reductions of vocabulary data. Reconstruction of language histories is more empirically based and robust than with previous methods.

Phylogeny of Riama (Squamata: Gymnophthalmidae), impact of phenotypic evidence on molecular datasets, and the origin of the Sierra Nevada de Santa Marta endemic fauna

Riama is the most speciose genus of the Neotropical lizard family Gymnophthalmidae. Its more than 30 montane species occur throughout the northern Andes, the Cordillera de la Costa (CC) in Venezuela, and Trinidad. We present the most comprehensive phylogenetic analysis of Riama to date based on a total evidence (TE) approach and direct optimization of molecular and morphological evidence. Analyses use DNA sequences from four loci and 35 phenotypic characters. The dataset consists of 55 ingroup terminals representing 25 of the 30 currently recognized species of Riama plus five undescribed taxa, including an endemic species from the Sierra Nevada de Santa Marta (SNSM) in Colombia, and 66 outgroup terminals of 47 species. Analysis results in a well-supported hypothesis in which Riama is polyphyletic, with its species falling into three clades. The Tepuian Anadia mcdiarmidi nests within one clade of Riama, and the recently resurrected Pantodactylus nests within Cercosaura. Accordingly, we propose a monophyletic taxonomy that reflects historical relationships. Analysis of character evolution indicates that the presence/absence of prefrontals-a cornerstone of the early genus-level taxonomy of cercosaurines-is optimally explained as having been plesiomorphically present in the most recent common ancestor of Cercosaurinae and lost in that of the immediately less inclusive clade. Multiple independent reversals to present and subsequent returns to absent occur within this clade. To evaluate the impact of phenotypic evidence on our results, we compare our TE results with results obtained from analyses using only molecular data. Although phenotypic evidence comprises only 1.2% of the TE matrix, its inclusion alters both the topology and support values of the clades that do not differ. Finally, current phylogenetic evidence reveals a SNSM-CC-Trinidad-tepuis biogeographical link. We hypothesize that an ancient connection facilitated the exchange of species between the SNSM and the CC.

Characterization of VP1 sequence of Coxsackievirus A16 isolates by Bayesian evolutionary method

BACKGROUND

Coxsackievirus A16 (CV-A16), a major etiopathologic cause of pediatric hand, foot, and mouth disease (HFMD) worldwide, has been reported to have caused several fatalities. Revealing the evolutionary and epidemiologic dynamics of CV-A16 across time and space is central to understanding its outbreak potential.

METHODS

In this study, we isolated six CV-A16 strains in China's Jilin province and construct a maximum clade credibility (MCC) tree for CV-A16 VP1 gene by the Bayesian Markov Chain Monte Carlo method using 708 strains from GenBank with epidemiological information. The evolution characteristics of CV-A16 VP1 gene was also analysed dynamicly through Bayesian skyline plot.

RESULTS

All CV-A16 strains identified could be classified into five major genogroups, denoted by GI-GV. GIV and GV have co-circulated in China since 2007, and the CV-A16 epidemic strain isolated in the Jilin province, China, can be classified as GIV-3. The CV-A16 genogroups circulating recently in China have the same ancestor since 2007. The genetic diversity of the CV-A16 VP1 gene shows a continuous increase since the mid-1990s, with sharp increases in genetic diversity in 1997 and 2007 and reached peak in 2007. Very low genetic diversity existed after 2010. The CV-A16 VP1 gene evolutionary rate was 6.656E-3 substitutions per site per year.

CONCLUSIONS

We predicted the dynamic phylogenetic trends, which indicate outbreak trends of CV-A16, and provide theoretical foundations for clinical prevention and treatment of HFMD which caused by a CV-A16.

Historical linguistics as a sequence optimization problem: the evolution and biogeography of Uto-Aztecan languages

Language origins and diversification are vital for mapping human history. Traditionally, the reconstruction of language trees has been based on cognate forms among related languages, with ancestral protolanguages inferred by individual investigators. Disagreement among competing authorities is typically extensive, without empirical grounds for resolving alternative hypotheses. Here, we apply analytical methods derived from DNA sequence optimization algorithms to Uto-Aztecan languages, treating words as sequences of sounds. Our analysis yields novel relationships and suggests a resolution to current conflicts about the Proto-Uto-Aztecan homeland. The techniques used for Uto-Aztecan are applicable to written and unwritten languages, and should enable more empirically robust hypotheses of language relationships, language histories, and linguistic evolution.