Inferring ancestral sequences in taxon-rich phylogenies

Statistical consistency in phylogenetics has traditionally referred to the accuracy of estimating phylogenetic parameters for a fixed number of species as we increase the number of characters. However, it is also useful to consider a dual type of statistical consistency where we increase the number of species, rather than characters. This raises some basic questions: what can we learn about the evolutionary process as we increase the number of species? In particular, does having more species allow us to infer the ancestral state of characters accurately? This question is particularly important when sequence evolution varies in a complex way from character to character, as methods applicable for i.i.d. models may no longer be valid. In this paper, we assemble a collection of results to analyse various approaches for inferring ancestral information with increasing accuracy as the number of taxa increases.     

Bayesian estimation of ancestral character states on phylogenies

Biologists frequently attempt to infer the character states at ancestral nodes of a phylogeny from the distribution of traits observed in contemporary organisms. Because phylogenies are normally inferences from data, it is desirable to account for the uncertainty in estimates of the tree and its branch lengths when making inferences about ancestral states or other comparative parameters. Here we present a general Bayesian approach for testing comparative hypotheses across statistically justified samples of phylogenies, focusing on the specific issue of reconstructing ancestral states. The method uses Markov chain Monte Carlo techniques for sampling phylogenetic trees and for investigating the parameters of a statistical model of trait evolution. We describe how to combine information about the uncertainty of the phylogeny with uncertainty in the estimate of the ancestral state. Our approach does not constrain the sample of trees only to those that contain the ancestral node or nodes of interest, and we show how to reconstruct ancestral states of uncertain nodes using a most-recent-common-ancestor approach. We illustrate the methods with data on ribonuclease evolution in the Artiodactyla. Software implementing the methods (BayesMultiState) is available from the authors.     

On the impossibility of reconstructing ancestral data and phylogenies.

We prove that it is impossible to reconstruct ancestral data at the root of "deep" phylogenetic trees with high mutation rates. Moreover, we prove that it is impossible to reconstruct the topology of "deep" trees with high mutation rates from a number of characters smaller than a low-degree polynomial in the number of leaves. Our impossibility results hold for all reconstruction methods. The proofs apply tools from information theory and percolation theory.     

Computational complexity of inferring phylogenies from dissimilarity matrices

Molecular biologists strive to infer evolutionary relationships from quantitative macromolecular comparisons obtained by immunological, DNA hybridization, electrophoretic or amino acid sequencing techniques. The problem is to find unrooted phylogenies that best approximate a given dissimilarity matrix according to a goodness-of-fit measure, for example the least-squares-fit criterion or Farris'sf statistic. Computational costs of known algorithms guaranteeing optimal solutions to these problems increase exponentially with problem size; practical computational considerations limit the algorithms to analyzing small problems. It is established here that problems of phylogenetic inference based on the least-squares-fit criterion and thef statistic are NP-complete and thus are so difficult computationally that efficient optimal algorithms are unlikely to exist for them. The Natural Sciences and Engineering Research Council of Canada partially supported this research through an individual operating grant (A4142) to W.H.E. Day.     

Surprising complexity of the ancestral apoptosis network

Background

Apoptosis, one of the main types of programmed cell death, is regulated and performed by a complex protein network. Studies in model organisms, mostly in the nematode Caenorhabditis elegans, identified a relatively simple apoptotic network consisting of only a few proteins. However, analysis of several recently sequenced invertebrate genomes, ranging from the cnidarian sea anemone Nematostella vectensis, representing one of the morphologically simplest metazoans, to the deuterostomes sea urchin and amphioxus, contradicts the current paradigm of a simple ancestral network that expanded in vertebrates.

Results

Here we show that the apoptosome-forming CED-4/Apaf-1 protein, present in single copy in vertebrate, nematode, and insect genomes, had multiple paralogs in the cnidarian-bilaterian ancestor. Different members of this ancestral Apaf-1 family led to the extant proteins in nematodes/insects and in deuterostomes, explaining significant functional differences between proteins that until now were believed to be orthologous. Similarly, the evolution of the Bcl-2 and caspase protein families appears surprisingly complex and apparently included significant gene loss in nematodes and insects and expansions in deuterostomes.

Conclusion

The emerging picture of the evolution of the apoptosis network is one of a succession of lineage-specific expansions and losses, which combined with the limited number of 'apoptotic' protein families, resulted in apparent similarities between networks in different organisms that mask an underlying complex evolutionary history. Similar results are beginning to surface for other regulatory networks, contradicting the intuitive notion that regulatory networks evolved in a linear way, from simple to complex.     

Cnidarians and ancestral genetic complexity in the animal kingdom

Eleven of the twelve recognized wingless (Wnt) subfamilies are represented in the sea anemone Nematostella vectensis, indicating that this developmentally important gene family was already fully diversified in the common ancestor of 'higher' animals. In deuterostomes, although duplications have occurred, no novel subfamilies of Wnts have evolved. By contrast, the protostomes Drosophila and Caenorhabditis have lost half of the ancestral Wnts. This pattern -- loss of genes from an ancestrally complex state -- might be more important in animal evolution than previously recognized.     

Computational complexity of inferring phylogenies from chromosome inversion data

In systematics, parsimony methods construct phylogenies, or evolutionary trees, in which characters evolve with the least evolutionary change. The chromosome inversion, or polymorphism, parsimony criterion is used when each character of a population may exhibit homozygous or heterozygous states, but when the heterozygous state must evolve uniquely. Variations of the criterion concern whether or not the ancestral states of characters are specified. We establish that problems of inferring phylogenies by these criteria are NP-complete and thus are so difficult computationally that efficient optimal algorithms for them are unlikely to exist.     

Molecular phylogeny of extant equids and effects of ancestral polymorphism in resolving species-level phylogenies

Short divergence times and processes such as incomplete lineage sorting and species hybridization are known to hinder the inference of species-level phylogenies due to the lack of sufficient informative genetic variation or the presence of shared but incongruent polymorphism among taxa. Extant equids (horses, zebras, and asses) are an example of a recently evolved group of mammals with an unresolved phylogeny, despite a large number of molecular studies. Previous surveys have proposed trees with rather poorly supported nodes, and the bias caused by genetic introgression or ancestral polymorphism has not been assessed. Here we studied the phylogenetic relationships of all extant species of Equidae by analyzing 22 partial mitochondrial and nuclear genes using maximum likelihood and Bayesian inferences that account for heterogeneous gene histories. We also examined genetic signatures of lineage sorting and/or genetic introgression in zebras by evaluating patterns of intraspecific genetic variation. Our study improved the resolution and support of the Equus phylogeny and in particular the controversial positions of the African wild ass (E. asinus) and mountain zebra (E. zebra): the African wild ass is placed as a sister species of the Asiatic asses and the mountain zebra as the sister taxon of Grevy's and Burchell's zebras. A shared polymorphism (indel) detected among zebra species in the Estrogen receptor 1 gene was likely due to incomplete lineage sorting and not genetic introgression as also indicated by other mitochondrial (Cytochrome b) and nuclear (Y chromosome and microsatellites) markers. Ancestral polymorphism in equids might have contributed to the long-standing lack of clarity in the phylogeny of this highly threatened group of mammals.     

Calculating complexity of large randomized libraries

Randomized libraries are increasingly popular in protein engineering and other biomedical research fields. Statistics of the libraries are useful to guide and evaluate randomized library construction. Previous works only give the mean of the number of unique sequences in the library, and they can only handle equal molar ratio of the four nucleotides at a small number of mutation sites. We derive formulas to calculate the mean and variance of the number of unique sequences in libraries generated by cassette mutagenesis with mixtures of arbitrary nucleotide ratios. Computer program was developed which utilizes arbitrary numerical precision software package to calculate the statistics of large libraries. The statistics of library with mutations in more than 20 amino acids can be calculated easily. Results show that the nucleotide ratios have significant effects on these statistics. The more skewed the ratio, the larger the library size is needed to obtain the same expected number of unique sequences. The program is freely available at http://graphics.med.yale.edu/cgi-bin/lib_comp.pl.     

Maintenance of ancestral complexity and non-metazoan genes in two basal cnidarians

Cnidarians are among the simplest extant animals; however EST analyses reveal that they have a remarkably high level of genetic complexity. In this article, we show that the full diversity of metazoan signaling pathways is represented in this phylum, as are antagonists previously known only in chordates. Many of the cnidarian ESTs match genes previously known only in non-animal kingdoms. At least some of these represent ancient genes lost by all bilaterians examined so far, rather than genes gained by recent lateral gene transfer.     

HyPhy: hypothesis testing using phylogenies

SUMMARY: The HyPhypackage is designed to provide a flexible and unified platform for carrying out likelihood-based analyses on multiple alignments of molecular sequence data, with the emphasis on studies of rates and patterns of sequence evolution. AVAILABILITY: http://www.hyphy.org CONTACT: muse@stat.ncsu.edu SUPPLEMENTARY INFORMATION: HyPhydocumentation and tutorials are available at http://www.hyphy.org.     

The innate immune repertoire in Cnidaria - ancestral complexity and stochastic gene loss

Background  

Characterization of the innate immune repertoire of extant cnidarians is of both fundamental and applied interest - it not only provides insights into the basic immunological 'tool kit' of the common ancestor of all animals, but is also likely to be important in understanding the global decline of coral reefs that is presently occurring. Recently, whole genome sequences became available for two cnidarians, Hydra magnipapillata and Nematostella vectensis, and large expressed sequence tag (EST) datasets are available for these and for the coral Acropora millepora.     

Stable tRNA-based phylogenies using only 76 nucleotides

tRNAs are among the most ancient, highly conserved sequences on earth, but are often thought to be poor phylogenetic markers because they are short, often subject to horizontal gene transfer, and easily change specificity. Here we use an algorithm now commonly used in microbial ecology, UniFrac, to cluster 175 genomes spanning all three domains of life based on the phylogenetic relationships among their complete tRNA pools. We find that the overall pattern of similarities and differences in the tRNA pools recaptures universal phylogeny to a remarkable extent, and that the resulting tree is similar to the distribution of bootstrapped rRNA trees from the same genomes. In contrast, the trees derived from tRNAs of identical specificity or of individual isoacceptors generally produced trees of lower quality. However, some tRNA isoacceptors were very good predictors of the overall pattern of organismal evolution. These results show that UniFrac can extract meaningful biological patterns from even phylogenies with high level of statistical inaccuracy and horizontal gene transfer, and that, overall, the pattern of tRNA evolution tracks universal phylogeny and provides a background against which we can test hypotheses about the evolution of individual isoacceptors.     

Reconstruction of ancestral RNA sequences under multiple structural constraints

Background

Secondary structures form the scaffold of multiple sequence alignment of non-coding RNA (ncRNA) families. An accurate reconstruction of ancestral ncRNAs must use this structural signal. However, the inference of ancestors of a single ncRNA family with a single consensus structure may bias the results towards sequences with high affinity to this structure, which are far from the true ancestors.

Methods

In this paper, we introduce achARNement, a maximum parsimony approach that, given two alignments of homologous ncRNA families with consensus secondary structures and a phylogenetic tree, simultaneously calculates ancestral RNA sequences for these two families.

Results

We test our methodology on simulated data sets, and show that achARNement outperforms classical maximum parsimony approaches in terms of accuracy, but also reduces by several orders of magnitude the number of candidate sequences. To conclude this study, we apply our algorithms on the Glm clan and the FinP-traJ clan from the Rfam database.

Conclusions

Our results show that our methods reconstruct small sets of high-quality candidate ancestors with better agreement to the two target structures than with classical approaches. Our program is freely available at: http://csb.cs.mcgill.ca/acharnement.

     

Testing macro-evolutionary models using incomplete molecular phylogenies.

Phylogenies reconstructed from gene sequences can be used to investigate the tempo and mode of species diversification. Here we develop and use new statistical methods to infer past patterns of speciation and extinction from molecular phylogenies. Specifically, we test the null hypothesis that per-lineage speciation and extinction rates have remained constant through time. Rejection of this hypothesis may provide evidence for evolutionary events such as adaptive radiations or key adaptations. In contrast to previous approaches, our methods are robust to incomplete taxon sampling and are conservative with respect to extinction. Using simulation we investigate, first, the adverse effects of failing to take incomplete sampling into account and, second, the power and reliability of our tests. When applied to published phylogenies our tests suggest that, in some cases, speciation rates have decreased through time.     

Thioredoxin - structural and functional complexity

Thioredoxins are small globular proteins that proved to be excellent model for investigating the relationship between the structure of protein and their physico-chemical and functional properties. The results from the experiments on thioredoxins offer the basic for the development of the new paradigms in the field of chemistry, biophysics and biology of proteins, with special attention to redox reaction in living cells, protein stability and design. It is a good example of broad class of sulphur-containing redox proteins.     

Topology testing of phylogenies using least squares methods

Background  

The least squares (LS) method for constructing confidence sets of trees is closely related to LS tree building methods, in which the goodness of fit of the distances measured on the tree (patristic distances) to the observed distances between taxa is the criterion used for selecting the best topology. The generalized LS (GLS) method for topology testing is often frustrated by the computational difficulties in calculating the covariance matrix and its inverse, which in practice requires approximations. The weighted LS (WLS) allows for a more efficient albeit approximate calculation of the test statistic by ignoring the covariances between the distances.     

Re-examining alveolate evolution using multiple protein molecular phylogenies

Alveolates are a diverse group of protists that includes three major lineages: ciliates, apicomplexa, and dinoflagellates. Among these three, it is thought that the apicomplexa and dinoflagellates are more closely related to one another than to ciliates. However, this conclusion is based almost entirely on results from ribosomal RNA phylogeny because very few morphological characters address this issue and scant molecular data are available from dinoflagellates. To better examine the relationships between the three major alveolate groups, we have sequenced six genes from the non-photosynthetic dinoflagellate, Crypthecodinium cohnii: actin, beta-tubulin, hsp70, BiP, hsp90, and mitochondrial hsp10. Beta-tubulin, hsp70, BiP, and hsp90 were found to be useful for intra-alveolate phylogeny, and trees were inferred from these genes individually and in combination. Trees inferred from individual genes generally supported the apicomplexa-dinoflagellate grouping, as did a combined analysis of all four genes. However, it was also found that the outgroup had a significant effect on the topology within alveolates when using certain methods of phylogenetic reconstruction, and an alternative topology clustering dinoflagellates and ciliates could not be rejected by the combined data. Altogether, these results support the sisterhood of apicomplexa and dinoflagellates, but point out that the relationship is not as strong as is often assumed.     

The structural complexity of DNA templates—Implications on cellular complexity

A previously formulated procedure for the quantitative evaluation of the complexities of molecules and biostructures is applied to assess the complexities of selected genomic DNA sequences. These include: (1) Several E. coli genes, including lacI, as examples of DNA sequences which are nearly as complex as possible (relative complexity=∼1). This is verified by the Lempel-Ziv (LZ) complexity analysis. (2) The telomere of a yeast chromosome, which has a considerable number of regular features that reduce complexity; the telomere shows indeed a lower structural complexity value. (3) A segment of human DNA, gene p53, which has a certain number of regular features such as 29 interspersed alu elements; these features cause a certain reduction in the complexity of the p53 gene, but do not invalidate the (previous) overall conclusion that template complexity is very high. The close to maximal complexity of the transcribed regions of p53 is validated by the LZ compression analysis. The general conclusion is that DNA base sequence composition is the dominant factor determining cellular complexity. The high complexity of DNA arrived at is a direct consequence of the template character of DNA and reflects the role of genomic DNA as a principal regulating element of a cell. It will be a challenge to find systems of lower complexity with the ability to respond to challenges from the environment to the extent that DNA templated systems do. Cellular complexity and template directed activity are thus highly intertwined properties, at the heart of many developmental, behavioral and evolutionary processes.     

Social scale and structural complexity in human languages

The complexity of different components of the grammars of human languages can be quantified. For example, languages vary greatly in the size of their phonological inventories, and in the degree to which they make use of inflectional morphology. Recent studies have shown that there are relationships between these types of grammatical complexity and the number of speakers a language has. Languages spoken by large populations have been found to have larger phonological inventories, but simpler morphology, than languages spoken by small populations. The results require further investigation, and, most importantly, the mechanism whereby the social context of learning and use affects the grammatical evolution of a language needs elucidation.