Computational Molecular Evolution Research Articles

Spectral cluster supertree: fast and statistically robust merging of rooted phylogenetic trees

The algorithms for phylogenetic reconstruction are central to computational molecular evolution. The relentless pace of data acquisition has exposed their poor scalability and the conclusion that the conventional application of these methods is impractical and not justifiable from an energy usage perspective. Furthermore, the drive to improve the statistical performance of phylogenetic methods produces increasingly parameter-rich models of sequence evolution, which worsens the computational performance. Established theoretical and algorithmic results identify supertree methods as critical to divide-and-conquer strategies for improving scalability of phylogenetic reconstruction. Of particular importance is the ability to explicitly accommodate rooted topologies. These can arise from the more biologically plausible non-stationary models of sequence evolution. We make a contribution to addressing this challenge with Spectral Cluster Supertree, a novel supertree method for merging a set of overlapping rooted phylogenetic trees. It offers significant improvements over Min-Cut supertree and previous state-of-the-art methods in terms of both time complexity and overall topological accuracy, particularly for problems of large size. We perform comparisons against Min-Cut supertree and Bad Clade Deletion. Leveraging two tree topology distance metrics, we demonstrate that while Bad Clade Deletion generates more correct clades in its resulting supertree, Spectral Cluster Supertree’s generated tree is generally more topologically close to the true model tree. Over large datasets containing 10,000 taxa and ∼500 source trees, where Bad Clade Deletion usually takes ∼2 h to run, our method generates a supertree in on average 20 s. Spectral Cluster Supertree is released under an open source license and is available on the python package index as sc-supertree.

Understanding the evolution of cold sensing and menthol activation in TRPM8.

Transient receptor potential melastatin 8 (TRPM8) is a polymodally gated ion channel activated by a variety of stimuli including menthol, cold, and pH. These chemical and physical stimuli ultimately regulate the ion conductance pore. While the menthol binding site and ion conduction pathway gate have been identified, the physics that underlies the mechanism of activation among different stimuli remains elusive. Particularly, the temperature modulation, how TRPM8 converts small changes in temperature into a conformational change, ultimately resulting in signal transduction is not well understood. Here to shed light into polymodal regulation of TRPM8, we perform a comparative analysis of conformational dynamics and functional characterizations of ancestral and extant human TRPM8. We use ancestral sequence reconstruction, a computational molecular evolution tool, to explore that ability to disentangle TRPM8 activation modes, which presumably emerged distinctly over evolutionary time. Based on the extant sequences of TRPM8, ancestral reconstruction was used to determine the DNA sequences of the last common primate, mammalian, and vertebrate ancestors. Resurrection and experimental analysis of these ancestral TRPM8 proteins show them to be functional in heterologous mammalian cells. Whole-cell patch-clamp electrophysiology and calcium mobilization assays confirm our hypothesis that the cold and menthol responses differ in terms of emergence and sensitivity between the ancestral constructs suggesting a new method to dissect polymodal protein behavior. Furthermore, comparative conformational analysis of ancestral and extant TRMP8 through the Dynamic Flexibility and Dynamic Coupling analysis also further support this hypothesis and provide dynamical finger prints for menthol and cold sensation.

Structural evolution of an amphibian-specific globin: A computational evolutionary biochemistry approach

Studies on the globin family are continuously revealing insights into the mechanisms of gene and protein evolution. The rise of a new globin gene type in Pelobatoidea and Neobatrachia (Amphibia:Anura) from an α-globin precursor provides the opportunity to investigate the genetic and physical mechanisms underlying the origin of new protein structural and functional properties. This amphibian-specific globin (globin A/GbA) discovered in the heart of Rana catesbeiana is a monomer. As the ancestral oligomeric state of α-globins is a homodimer, we inferred that the ancestral state was lost somewhere in the GbA lineage. Here, we combined computational molecular evolution with structural bioinformatics to determine the extent to which the loss of the homodimeric state is pervasive in the GbA clade. We also characterized the loci of GbA genes in Bufo bufo. We found two GbA clades in Neobatrachia. One was deleted in Ranidae, but retained and expanded to yield a new globin cluster in Bufonidae species. Loss of the ancestral oligomeric state seems to be pervasive in the GbA clade. However, a taxonomic sampling that includes more Pelobatoidea, as well as early Neobatrachia, lineages would be necessary to determine the oligomeric state of the last common ancestor of all GbA. The evidence presented here points out a possible loss of oligomerization in Pelobatoidea GbA as a result of amino acid substitutions that weaken the homodimeric state. In contrast, the loss of oligomerization in both Neobatrachia GbA clades was linked to independent deletions that disrupted many packing contacts at the homodimer interface.

A Not-So-Long Introduction to Computational Molecular Evolution.

In this chapter, we give a not-so-long and self-contained introduction to computational molecular evolution. In particular, we present the emergence of the use of likelihood-based methods, review the standard DNA substitution models, and introduce how model choice operates. We also present recent developments in inferring absolute divergence times and rates on a phylogeny, before showing how state-of-the-art models take inspiration from diffusion theory to link population genetics, which traditionally focuses at a taxonomic level below that of the species, and molecular evolution. Although this is not a cookbook chapter, we try and point to popular programs and implementations along the way.

Crystal structure of an inferred ancestral bacterial pyruvate decarboxylase.

Pyruvate decarboxylase (PDC; EC 4.1.1.1) is a key enzyme in homofermentative metabolism where ethanol is the major product. PDCs are thiamine pyrophosphate- and Mg2+ ion-dependent enzymes that catalyse the non-oxidative decarboxylation of pyruvate to acetaldehyde and carbon dioxide. As this enzyme class is rare in bacteria, current knowledge of bacterial PDCs is extremely limited. One approach to further the understanding of bacterial PDCs is to exploit the diversity provided by evolution. Ancestral sequence reconstruction (ASR) is a method of computational molecular evolution to infer extinct ancestral protein sequences, which can then be synthesized and experimentally characterized. Through ASR a novel PDC was generated, designated ANC27, that shares only 78% amino-acid sequence identity with its closest extant homologue (Komagataeibacter medellinensis PDC, GenBank accession No. WP_014105323.1), yet is fully functional. Crystals of this PDC diffracted to 3.5 Å resolution. The data were merged in space group P3221, with unit-cell parameters a = b = 108.33, c = 322.65 Å, and contained two dimers (two tetramer halves) in the asymmetric unit. The structure was solved by molecular replacement using PDB entry 2wvg as a model, and the final R values were Rwork = 0.246 (0.3671 in the highest resolution bin) and Rfree = 0.319 (0.4482 in the highest resolution bin). Comparison with extant bacterial PDCs supports the previously observed correlation between decreased tetramer interface area (and number of interactions) and decreased thermostability.

Darwin and Fisher meet at biotech: on the potential of computational molecular evolution in industry.

BackgroundToday computational molecular evolution is a vibrant research field that benefits from the availability of large and complex new generation sequencing data – ranging from full genomes and proteomes to microbiomes, metabolomes and epigenomes. The grounds for this progress were established long before the discovery of the DNA structure. Specifically, Darwin’s theory of evolution by means of natural selection not only remains relevant today, but also provides a solid basis for computational research with a variety of applications. But a long-term progress in biology was ensured by the mathematical sciences, as exemplified by Sir R. Fisher in early 20th century. Now this is true more than ever: The data size and its complexity require biologists to work in close collaboration with experts in computational sciences, modeling and statistics.ResultsNatural selection drives function conservation and adaptation to emerging pathogens or new environments; selection plays key role in immune and resistance systems. Here I focus on computational methods for evaluating selection in molecular sequences, and argue that they have a high potential for applications. Pharma and biotech industries can successfully use this potential, and should take the initiative to enhance their research and development with state of the art bioinformatics approaches.ConclusionsThis review provides a quick guide to the current computational approaches that apply the evolutionary principles of natural selection to real life problems – from drug target validation, vaccine design and protein engineering to applications in agriculture, ecology and conservation.

Intra-specific variability and biological relevance of P3N-PIPO protein length in potyviruses.

BackgroundPipo was recently described as a new ORF encoded within the genome of the Potyviridae family members (PNAS 105:5897–5902, 2008). It is embedded within the P3 cistron and is translated in the +2 reading frame relative to the potyviral long ORF as the P3N-PIPO fusion protein. In this work, we first collected pipo nucleotide sequences available for different isolates of 48 Potyvirus species. Second, to determine the biological implications of variation in pipo length, we measured infectivity, viral accumulation, cell-to-cell and systemic movements for two Turnip mosaic virus (TuMV) variants with pipo alleles of different length in three different susceptible host species, and tested for differences between the two variants.ResultsIn addition to inter-specific variation, there was high variation in the length of the PIPO protein among isolates within species (ranging from 1 to 89 amino acids). Furthermore, selection analyses on the P3 cistron did not account for the existence of stop codons in the pipo ORF, but showed that positive selection was significant in the overlapping region for Potato virus Y (PVY) and TuMV. In some cases, variability in length was associated with host species, geographic provenance and/or other strain features. We found significant empirical differences among the phenotypes associated with TuMV pipo alleles, though the magnitude and sign of the effects were host-dependent.ConclusionsThe combination of computational molecular evolution analyses and experiments stemming from these analyses provide clues about the selective pressures acting upon the different-length pipo alleles and show that variation in length may be maintained by host-driven selection.

Site-specific evolutionary rates in proteins are better modeled as non-independent and strictly relative

In a nucleotide or amino acid sequence, not all sites evolve at the same rate, due to differing selective constraints at each site. Currently in computational molecular evolution, models incorporating rate heterogeneity always share two assumptions. First, the rate of evolution at each site is assumed to be independent of every other site. Second, the values of these rates are assumed to be drawn from a known prior distribution. Although often assumed to be small, the actual effect of these assumptions has not been previously quantified in the literature. Herein we describe an algorithm to simultaneously infer the set of n-1 relative rates that parameterize the likelihood of an n-site alignment. Unlike previous work (a) these relative rates are completely identifiable and distinct from the branch-length parameters, and (b) a far more general class of rate priors can be used, and their effects quantified. Although described in a Bayesian framework, we discuss a future maximum likelihood extension. Using both synthetic data and alignments from the Myc, Max and p53 protein families, we find that inferring relative rather than absolute rates has several advantages. First, both empirical likelihoods and Bayes factors show strong preference for the relative-rate model, with a mean Delta ln P=-0.458 per alignment site. Second, the computed likelihoods and Bayes factors were essentially independent of the relative-rate prior, indicating that good estimates of the posterior rate distribution are not required a priori. Third, a novel finding is that rates can be accurately inferred even when up to approximately 4 substitutions per site have occurred. Thus biologically relevant putative hypervariable sites can be identified as easily as conserved sites. Lastly, our model treats rates and tree branch-lengths as completely identifiable, allowing for the first time coherent simultaneous inference of branch-lengths and site-specific evolutionary rates. Source code for the utility described is available under a BSD-style license at http://www.fernandes.org/txp/article/9/site-specific-relative-evolutionary-rates.