Nicolas Rodrigue

PhyloBayes MPI: Phylogenetic Reconstruction with Infinite Mixtures of Profiles in a Parallel Environment

Modeling across site variation of the substitution process is increasingly recognized as important for obtaining more accurate phylogenetic reconstructions. Both finite and infinite mixture models have been proposed and have been shown to significantly improve on classical single-matrix models. Compared with their finite counterparts, infinite mixtures have a greater expressivity. However, they are computationally more challenging. This has resulted in practical compromises in the design of infinite mixture models. In particular, a fast but simplified version of a Dirichlet process model over equilibrium frequency profiles implemented in PhyloBayes has often been used in recent phylogenomics studies, while more refined model structures, more realistic and empirically more fit, have been practically out of reach. We introduce a message passing interface version of PhyloBayes, implementing the Dirichlet process mixture models as well as more classical empirical matrices and finite mixtures. The parallelization is made efficient thanks to the combination of two algorithmic strategies: a partial Gibbs sampling update of the tree topology and the use of a truncated stick-breaking representation for the Dirichlet process prior. The implementation shows close to linear gains in computational speed for up to 64 cores, thus allowing faster phylogenetic reconstruction under complex mixture models. PhyloBayes MPI is freely available from our website www.phylobayes.org.

Heterotachy and long-branch attraction in phylogenetics

BackgroundProbabilistic methods have progressively supplanted the Maximum Parsimony (MP) method for inferring phylogenetic trees. One of the major reasons for this shift was that MP is much more sensitive to the Long Branch Attraction (LBA) artefact than is Maximum Likelihood (ML). However, recent work by Kolaczkowski and Thornton suggested, on the basis of simulations, that MP is less sensitive than ML to tree reconstruction artefacts generated by heterotachy, a phenomenon that corresponds to shifts in site-specific evolutionary rates over time. These results led these authors to recommend that the results of ML and MP analyses should be both reported and interpreted with the same caution. This specific conclusion revived the debate on the choice of the most accurate phylogenetic method for analysing real data in which various types of heterogeneities occur. However, variation of evolutionary rates across species was not explicitly incorporated in the original study of Kolaczkowski and Thornton, and in most of the subsequent heterotachous simulations published to date, where all terminal branch lengths were kept equal, an assumption that is biologically unrealistic.ResultsIn this report, we performed more realistic simulations to evaluate the relative performance of MP and ML methods when two kinds of heterogeneities are considered: (i) within-site rate variation (heterotachy), and (ii) rate variation across lineages. Using a similar protocol as Kolaczkowski and Thornton to generate heterotachous datasets, we found that heterotachy, which constitutes a serious violation of existing models, decreases the accuracy of ML whatever the level of rate variation across lineages. In contrast, the accuracy of MP can either increase or decrease when the level of heterotachy increases, depending on the relative branch lengths. This result demonstrates that MP is not insensitive to heterotachy, contrary to the report of Kolaczkowski and Thornton. Finally, in the case of LBA (i.e. when two non-sister lineages evolved faster than the others), ML outperforms MP over a wide range of conditions, except for unrealistic levels of heterotachy.ConclusionFor realistic combinations of both heterotachy and variation of evolutionary rates across lineages, ML is always more accurate than MP. Therefore, ML should be preferred over MP for analysing real data, all the more so since parametric methods also allow one to handle other types of biological heterogeneities much better, such as among sites rate variation. The confounding effects of heterotachy on tree reconstruction methods do exist, but can be eschewed by the development of mixture models in a probabilistic framework, as proposed by Kolaczkowski and Thornton themselves.

Evolutionary insight from whole-genome sequencing of experimentally evolved microbes.

Experimental evolution (EE) combined with whole‐genome sequencing (WGS) has become a compelling approach to study the fundamental mechanisms and processes that drive evolution. Most EE‐WGS studies published to date have used microbes, owing to their ease of propagation and manipulation in the laboratory and relatively small genome sizes. These experiments are particularly suited to answer long‐standing questions such as: How many mutations underlie adaptive evolution, and how are they distributed across the genome and through time? Are there general rules or principles governing which genes contribute to adaptation, and are certain kinds of genes more likely to be targets than others? How common is epistasis among adaptive mutations, and what does this reveal about the variety of genetic routes to adaptation? How common is parallel evolution, where the same mutations evolve repeatedly and independently in response to similar selective pressures? Here, we summarize the significant findings of this body of work, identify important emerging trends and propose promising directions for future research. We also outline an example of a computational pipeline for use in EE‐WGS studies, based on freely available bioinformatics tools.

Mutation-selection models of coding sequence evolution with site-heterogeneous amino acid fitness profiles.

Modeling the interplay between mutation and selection at the molecular level is key to evolutionary studies. To this end, codon-based evolutionary models have been proposed as pertinent means of studying long-range evolutionary patterns and are widely used. However, these approaches have not yet consolidated results from amino acid level phylogenetic studies showing that selection acting on proteins displays strong site-specific effects, which translate into heterogeneous amino acid propensities across the columns of alignments; related codon-level studies have instead focused on either modeling a single selective context for all codon columns, or a separate selective context for each codon column, with the former strategy deemed too simplistic and the latter deemed overparameterized. Here, we integrate recent developments in nonparametric statistical approaches to propose a probabilistic model that accounts for the heterogeneity of amino acid fitness profiles across the coding positions of a gene. We apply the model to a dozen real protein-coding gene alignments and find it to produce biologically plausible inferences, for instance, as pertaining to site-specific amino acid constraints, as well as distributions of scaled selection coefficients. In their account of mutational features as well as the heterogeneous regimes of selection at the amino acid level, the modeling approaches studied here can form a backdrop for several extensions, accounting for other selective features, for variable population size, or for subtleties of mutational features, all with parameterizations couched within population-genetic theory.

Genomics of Adaptation during Experimental Evolution of the Opportunistic Pathogen Pseudomonas aeruginosa

Adaptation is likely to be an important determinant of the success of many pathogens, for example when colonizing a new host species, when challenged by antibiotic treatment, or in governing the establishment and progress of long-term chronic infection. Yet, the genomic basis of adaptation is poorly understood in general, and for pathogens in particular. We investigated the genetics of adaptation to cystic fibrosis-like culture conditions in the presence and absence of fluoroquinolone antibiotics using the opportunistic pathogen Pseudomonas aeruginosa. Whole-genome sequencing of experimentally evolved isolates revealed parallel evolution at a handful of known antibiotic resistance genes. While the level of antibiotic resistance was largely determined by these known resistance genes, the costs of resistance were instead attributable to a number of mutations that were specific to individual experimental isolates. Notably, stereotypical quinolone resistance mutations in DNA gyrase often co-occurred with other mutations that, together, conferred high levels of resistance but no consistent cost of resistance. This result may explain why these mutations are so prevalent in clinical quinolone-resistant isolates. In addition, genes involved in cyclic-di-GMP signalling were repeatedly mutated in populations evolved in viscous culture media, suggesting a shared mechanism of adaptation to this CF–like growth environment. Experimental evolutionary approaches to understanding pathogen adaptation should provide an important complement to studies of the evolution of clinical isolates.

Evolutionary genomics of epidemic and nonepidemic strains of Pseudomonas aeruginosa

Significance The bacterium Pseudomonas aeruginosa is an opportunistic pathogen of humans and is the leading cause of death in patients with cystic fibrosis (CF). We sequenced the genomes of P. aeruginosa isolated from respiratory tracts of patients with CF to investigate general patterns of adaptation associated with chronic infection. Selection imposed by the CF lung environment has had a major influence on genomic evolution and the genetic characteristics of isolates causing contemporary infection. Many of the genes and pathways implicated in adaptive evolution within the host had obvious roles in the pathogenic lifestyle of this bacteria. Genome sequence data indicated that an epidemic strain, with increased virulence and multidrug resistance, has spread between clinics in the United Kingdom and North America. Pseudomonas aeruginosa is an opportunistic pathogen of humans and is a major cause of morbidity and mortality in patients with cystic fibrosis (CF). Prolonged infection of the respiratory tract can lead to adaptation of the pathogen to the CF lung environment. To examine the general patterns of adaptation associated with chronic infection, we obtained genome sequences from a collection of P. aeruginosa isolated from airways of patients with CF. Our analyses support a nonclonal epidemic population structure, with a background of unique, recombining genotypes, and the rare occurrence of successful epidemic clones. We present unique genome sequence evidence for the intercontinental spread of an epidemic strain shared between CF clinics in the United Kingdom and North America. Analyses of core and accessory genomes identified candidate genes and important functional pathways associated with adaptive evolution. Many genes of interest were involved in biological functions with obvious roles in this pathosystem, such as biofilm formation, antibiotic metabolism, pathogenesis, transport, reduction/oxidation, and secretion. Key factors driving the adaptive evolution of this pathogen within the host appear to be the presence of oxidative stressors and antibiotics. Regions of the accessory genome unique to the epidemic strain were enriched for genes in transporter families that efflux heavy metals and antibiotics. The epidemic strain was significantly more resistant than nonepidemic strains to three different antibiotics. Multiple lines of evidence suggest that selection imposed by the CF lung environment has a major influence on genomic evolution and the genetic characteristics of P. aeruginosa isolates causing contemporary infection.

Uniformization for sampling realizations of Markov processes

MOTIVATION Mapping character state changes over phylogenetic trees is central to the study of evolution. However, current probabilistic methods for generating such mappings are ill-suited to certain types of evolutionary models, in particular, the widely used models of codon substitution. RESULTS We describe a general method, based on a uniformization technique, which can be utilized to generate realizations of a Markovian substitution process conditional on an alignment of character states and a given tree topology. The method is applicable under a wide range of evolutionary models, and to illustrate its usefulness in practice, we embed it within a data augmentation-based Markov chain Monte Carlo sampler, for approximating posterior distributions under previously proposed codon substitution models. The sampler is found to be more efficient than the conventional pruning-based sampler with the decorrelation times between draws from the posterior reduced by a factor of 20 or more.

Statistical Potentials for Improved Structurally Constrained Evolutionary Models

Assessing the influence of three-dimensional protein structure on sequence evolution is a difficult task, mainly because of the assumption of independence between sites required by probabilistic phylogenetic methods. Recently, models that include an explicit treatment of protein structure and site interdependencies have been developed: a statistical potential (an energy-like scoring system for sequence-structure compatibility) is used to evaluate the probability of fixation of a given mutation, assuming a coarse-grained protein structure that is constant through evolution. Yet, due to the novelty of these models and the small degree of overlap between the fields of structural and evolutionary biology, only simple representations of protein structure have been used so far. In this work, we present new forms of statistical potentials using a probabilistic framework recently developed for evolutionary studies. Terms related to pairwise distance interactions, torsion angles, solvent accessibility, and flexibility of the residues are included in the potentials, so as to study the effects of the main factors known to influence protein structure. The new potentials, with a more detailed representation of the protein structure, yield a better fit than the previously used scoring functions, with pairwise interactions contributing to more than half of this improvement. In a phylogenetic context, however, the structurally constrained models are still outperformed by some of the available site-independent models in terms of fit, possibly indicating that alternatives to coarse-grained statistical potentials should be explored in order to better model structural constraints.

Computational Methods for Evaluating Phylogenetic Models of Coding Sequence Evolution with Dependence between Codons

In recent years, molecular evolutionary models formulated as site-interdependent Markovian codon substitution processes have been proposed as means of mechanistically accounting for selective features over long-range evolutionary scales. Under such models, site interdependencies are reflected in the use of a simplified protein tertiary structure representation and predefined statistical potential, which, along with mutational parameters, mediate nonsynonymous rates of substitution; rates of synonymous events are solely mediated by mutational parameters. Although theoretically attractive, the models are computationally challenging, and the methods used to manipulate them still do not allow for quantitative model evaluations in a multiple-sequence context. Here, we describe Markov chain Monte Carlo computational methodologies for sampling parameters from their posterior distribution under site-interdependent codon substitution models within a phylogenetic context and allowing for Bayesian model assessment and ranking. Specifically, the techniques we expound here can form the basis of posterior predictive checking under these models and can be embedded within thermodynamic integration algorithms for computing Bayes factors. We illustrate the methods using two data sets and find that although current forms of site-interdependent models of codon substitution provide an improved fit, they are outperformed by the extended site-independent versions. Altogether, the methodologies described here should enable a quantified contrasting of alternative ways of modeling structural constraints, or other site-interdependent criteria, and establish if such formulations can match (or supplant) site-independent model extensions.

The effect of selection environment on the probability of parallel evolution

Across the great diversity of life, there are many compelling examples of parallel and convergent evolution-similar evolutionary changes arising in independently evolving populations. Parallel evolution is often taken to be strong evidence of adaptation occurring in populations that are highly constrained in their genetic variation. Theoretical models suggest a few potential factors driving the probability of parallel evolution, but experimental tests are needed. In this study, we quantify the degree of parallel evolution in 15 replicate populations of Pseudomonas fluorescens evolved in five different environments that varied in resource type and arrangement. We identified repeat changes across multiple levels of biological organization from phenotype, to gene, to nucleotide, and tested the impact of 1) selection environment, 2) the degree of adaptation, and 3) the degree of heterogeneity in the environment on the degree of parallel evolution at the gene-level. We saw, as expected, that parallel evolution occurred more often between populations evolved in the same environment; however, the extent of parallel evolution varied widely. The degree of adaptation did not significantly explain variation in the extent of parallelism in our system but number of available beneficial mutations correlated negatively with parallel evolution. In addition, degree of parallel evolution was significantly higher in populations evolved in a spatially structured, multiresource environment, suggesting that environmental heterogeneity may be an important factor constraining adaptation. Overall, our results stress the importance of environment in driving parallel evolutionary changes and point to a number of avenues for future work for understanding when evolution is predictable.