Search

International audience; Widely applicable, accurate and fast inference methods in phylodynamics are needed to fully profit from the richness of genetic data in uncovering the dynamics of epidemics. Standard methods, including maximum-likelihood and Bayesian approaches, generally rely on complex mathematical formulae and approximations, and do not scale with dataset size. We develop a likelihood-free, simulation-based approach, which combines deep learning with (1) a large set of summary statistics measured on phylogenies or (2) a complete and compact representation of trees, which avoids potential limitations of summary statistics and applies to any phylodynamics model. Our method enables both model selection and estimation of epidemiological parameters from very large phylogenies. We demonstrate its speed and accuracy on simulated data, where it performs better than the state-of-the-art methods. To illustrate its applicability, we assess the dynamics induced by superspreading individuals in an HIV dataset of men-having-sex-with-men in Zurich. Our tool PhyloDeep is available on github.com/evolbioinfo/phylodeep.

Widely applicable, accurate and fast inference methods in phylodynamics are needed to fully profit from the richness of genetic data in uncovering the dynamics of epidemics. Standard methods, including maximum-likelihood and Bayesian approaches, generally rely on complex mathematical formulae and approximations, and do not scale with dataset size. We develop a likelihood-free, simulation-based approach, which combines deep learning with (1) a large set of summary statistics measured on phylogenies or (2) a complete and compact representation of trees, which avoids potential limitations of summary statistics and applies to any phylodynamics model. Our method enables both model selection and estimation of epidemiological parameters from very large phylogenies. We demonstrate its speed and accuracy on simulated data, where it performs better than the state-of-the-art methods. To illustrate its applicability, we assess the dynamics induced by superspreading individuals in an HIV dataset of men-having-sex-with-men in Zurich. Our tool PhyloDeep is available on github.com/evolbioinfo/phylodeep.

International audience; Une fois une phylogénie estimée, se pose la question du support de ses branches et des inférences associées : Décrivent-elles la réalité du passé biologique ? Sontelles fiables, reproductibles, robustes ? Ou à l'inverse peut-on conclure que le signal contenu dans les données ne permet pas de trancher entre plusieurs hypothèses alternatives ? De nombreuses approches ont été proposées pour répondre à ces questions. Nous décrivons dans ce chapitre les principales méthodes utilisées aujourd'hui en phylogénétique. Certaines sont locales et demandent peu de temps calcul. D'autres comme le bootstrap requièrent à l'inverse des calculs lourds mais donnent une vision plus globale. On distingue les méthodes paramétriques qui font explicitement appel à un modèle de l'évolution des séquences et se basent sur la fonction de vraisemblance ou la distribution a posteriori dans un cadre Bayésien, et les approches non-paramétriques qui reposent sur moins d'hypothèses explicites. On trouvera quatre parties dans ce chapitre : (1) supports locaux ; (2) bootstrap phylogénétique de Felsenstein et bootstrap de transfert ; (3) supports Bayésiens basés sur les probabilités a posteriori; (4) comparaison de ces approches et discussion.

International audience; The Zika virus (ZIKV) disease caused a public health emergency of international concern that started in February 2016. The overall number of ZIKV-related cases increased until November 2016, after which it declined sharply. While the evaluation of the potential risk and impact of future arbovirus epidemics remains challenging, intensified surveillance efforts along with a scale-up of ZIKV whole-genome sequencing provide an opportunity to understand the patterns of genetic diversity, evolution, and spread of ZIKV. However, a classification system that reflects the true extent of ZIKV genetic variation is lacking. Our objective was to characterize ZIKV genetic diversity and phylodynamics, identify genomic footprints of differentiation patterns, and propose a dynamic classification system that reflects its divergence levels. We analysed a curated dataset of 762 publicly available sequences spanning the full-length coding region of ZIKV from across its geographical span and collected between 1947 and 2021. The definition of genetic groups was based on comprehensive evolutionary dynamics analyses, which included recombination and phylogenetic analyses, within- and between-group pairwise genetic distances comparison, detection of selective pressure, and clustering analyses. Evidence for potential recombination events was detected in a few sequences. However, we argue that these events are likely due to sequencing errors as proposed in previous studies. There was evidence of strong purifying selection, widespread across the genome, as also detected for other arboviruses. A total of 50 sites showed evidence of positive selection, and for a few of these sites, there was amino acid (AA) differentiation between genetic clusters. Two main genetic clusters were defined, ZA and ZB, which correspond to the already characterized ‘African’ and ‘Asian’ genotypes, respectively. Within ZB, two subgroups, ZB.1 and ZB.2, represent the Asiatic and the American (and Oceania) lineages, respectively. ZB.1 is further subdivided into ZB.1.0 (a basal Malaysia sequence sampled in the 1960s and a recent Indian sequence), ZB.1.1 (South-Eastern Asia, Southern Asia, and Micronesia sequences), and ZB.1.2 (very similar sequences from the outbreak in Singapore). ZB.2 is subdivided into ZB.2.0 (basal American sequences and the sequences from French Polynesia, the putative origin of South America introduction), ZB.2.1 (Central America), and ZB.2.2 (Caribbean and North America). This classification system does not use geographical references and is flexible to accommodate potential future lineages. It will be a helpful tool for studies that involve analyses of ZIKV genomic variation and its association with pathogenicity and serve as a starting point for the public health surveillance and response to on-going and future epidemics and to outbreaks that lead to the emergence of new variants.The Zika virus (ZIKV) disease caused a public health emergency of international concern that started in February 2016. The overall number of ZIKV-related cases increased until November 2016, after which it declined sharply. While the evaluation of the potential risk and impact of future arbovirus epidemics remains challenging, intensified surveillance efforts along with a scale-up of ZIKV whole-genome sequencing provide an opportunity to understand the patterns of genetic diversity, evolution, and spread of ZIKV. However, a classification system that reflects the true extent of ZIKV genetic variation is lacking. Our objective was to characterize ZIKV genetic diversity and phylodynamics, identify genomic footprints of differentiation patterns, and propose a dynamic classification system that reflects its divergence levels. We analysed a curated dataset of 762 publicly available sequences spanning the full-length coding region of ZIKV from across its geographical span and collected between 1947 and 2021. The definition of genetic groups was based on comprehensive evolutionary dynamics analyses, which included recombination and phylogenetic analyses, within- and between-group pairwise genetic distances comparison, detection of selective pressure, and clustering analyses. Evidence for potential recombination events was detected in a few sequences. However, we argue that these events are likely due to sequencing errors as proposed in previous studies. There was evidence of strong purifying selection, widespread across the genome, as also detected for other arboviruses. A total of 50 sites showed evidence of positive selection, and for a few of these sites, there was amino acid (AA) differentiation between genetic clusters. Two main genetic clusters were defined, ZA and ZB, which correspond to the already characterized ‘African’ and ‘Asian’ genotypes, respectively. Within ZB, two subgroups, ZB.1 and ZB.2, represent the Asiatic and the American (and Oceania) lineages, respectively. ZB.1 is further subdivided into ZB.1.0 (a basal Malaysia sequence sampled in the 1960s and a recent Indian sequence), ZB.1.1 (South-Eastern Asia, Southern Asia, and Micronesia sequences), and ZB.1.2 (very similar sequences from the outbreak in Singapore). ZB.2 is subdivided into ZB.2.0 (basal American sequences and the sequences from French Polynesia, the putative origin of South America introduction), ZB.2.1 (Central America), and ZB.2.2 (Caribbean and North America). This classification system does not use geographical references and is flexible to accommodate potential future lineages. It will be a helpful tool for studies that involve analyses of ZIKV genomic variation and its association with pathogenicity and serve as a starting point for the public health surveillance and response to on-going and future epidemics and to outbreaks that lead to the emergence of new variants.

Evolutionary convergences are observed at all levels, from phenotype to DNA and protein sequences, and the changes observed at these different levels tend to be strongly correlated. Here we propose a simulation-based method to detect positions under convergent evolution in large protein alignments, without prior knowledge on the phenotype and environmental constraints. A phylogeny is inferred from the data and used in simulations to estimate the expected number of amino-acid changes in stable evolutionary constraints (null model) for each position. Similarly, we count the number of mutations towards the same amino acid in the data and test if they are occurring more often than expected. We applied our method to two real datasets: HIV reverse transcriptase and fish rhodopsin, and to HIVlike simulated data. On the latter, with known convergent events and substitution model, we detected on average two third of these events, with a low fraction of false positives. With HIV data, one knows that drug resistance mutations (DRMs) are convergent. Even without any knowledge of patient treatment status, we retrieved more than 70% of positions corresponding to known DRMs. On the rhodopsin dataset, four substitutions are supposed to be convergent, as they change the maximum wavelength absorption of the photoreceptor and occurred several times independently during evolution. We detected three of them. These results demonstrate the potential of the method to target specific mutations to be further studied experimentally or, for example, using a nonsynonymous/synonymous rate ratio approach. Our software named ConDor is available at http://condor.pasteur.cloud.

International audience; SARS-CoV-2 is the virus responsible for the global COVID19 pandemic. We review what is known about the origin of this virus, detected in China at the end of December 2019. The genome of this virus mainly evolves under the effect of point mutations. These are generally neutral and have no impact on virulence and severity, but some appear to influence infectivity, notably the D614G mutation of the Spike protein. To date (30/09/2020) no recombination of the virus has been documented in the human host, and very few insertions and deletions. The worldwide spread of the virus was the subject of controversies that we summarize, before proposing a new approach free from the limitations of previous methods. The results show a complex scenario with, for example, numerous introductions to the USA and returns of the virus from the USA to certain countries including France.; Le SARS-CoV-2 est le virus responsable de la pandémie mondiale de COVID19. On dresse ici un bilan de ce qui est connu sur l’origine de ce virus, détecté en Chine fin décembre 2019. Le génome de ce virus évolue sous l’effet de mutations ponctuelles. Celles-ci sont généralement neutres et sans impact sur la virulence et la sévérité, mais certaines semblent influer sur l’infectiosité, notamment la mutation D614G de la protéine Spike. A l’inverse, on n’a à ce jour (30/09/2020) documenté aucune recombinaison du virus chez l’hôte humain, et très peu d’insertions et de délétions. La propagation mondiale du virus a fait l’objet de polémiques sur lesquelles nous revenons, avant de proposer une nouvelle approche débarrassée des limites des méthodes précédentes. Les résultats montrent une propagation complexe avec, par exemple, de très nombreuses introductions aux USA et des retours du virus depuis les USA vers certains pays dont la France.

Motivation The first cases of the COVID-19 pandemic emerged in December 2019. Until the end of February 2020, the number of available genomes was below 1000 and their multiple alignment was easily achieved using standard approaches. Subsequently, the availability of genomes has grown dramatically. Moreover, some genomes are of low quality with sequencing/assembly errors, making accurate re-alignment of all genomes nearly impossible on a daily basis. A more efficient, yet accurate approach was clearly required to pursue all subsequent bioinformatics analyses of this crucial data. Results hCoV-19 genomes are highly conserved, with very few indels and no recombination. This makes the profile HMM approach particularly well suited to align new genomes, add them to an existing alignment and filter problematic ones. Using a core of ∼2500 high quality genomes, we estimated a profile using HMMER, and implemented this profile in COVID-Align, a user-friendly interface to be used online or as standalone via Docker. The alignment of 1000 genomes requires ∼50 minutes on our cluster. Moreover, COVID-Align provides summary statistics, which can be used to determine the sequencing quality and evolutionary novelty of input genomes (e.g. number of new mutations and indels). Availability and implementation https://covalign.pasteur.cloud, hub.docker.com/r/evolbioinfo/covid-align. Supplementary information Supplementary data are available at Bioinformatics online.

Plasmid-mediated horizontal gene transfer of antibiotic resistance and virulence in pathogenic bacteria underlies a major public health issue. Understanding how, in the absence of antibiotic-mediated selection, plasmid-bearing cells avoid being outnumbered by plasmid-free cells is key to developing counterstrategies. Here, we quantified the induction of the plasmidial sex pheromone pathway of Enterococcus faecalis to show that the integration of the stimulatory (mate-sensing) and inhibitory (self-sensing) signaling modules from the pCF10 conjugative plasmid provides a precise measure of the recipient-to-donor ratio, agnostic to variations in population size. Such ratiometric control of conjugation favors vertical plasmid transfer under low mating likelihood and allows activation of conjugation functions only under high mating likelihood. We further show that this strategy constitutes a cost-effective investment into mating effort because overstimulation produces unproductive self-aggregation and growth rate reduction. A mathematical model suggests that ratiometric control of conjugation increases plasmid fitness and predicts a robust long-term, stable coexistence of donors and recipients. Our results demonstrate how population-level parameters can control transfer of antibiotic resistance in bacteria, opening the door for biotic control strategies., The pathogenic bacterium Enterococcus faecalis controls the horizontal gene transfer of antibiotic resistance according to the ratio of non-resistant to resistant bacteria, while remaining agnostic to variation in the total culture density. This ratiometric twist on the well-known phenomenon of quorum-sensing allows investment in expensive conjugative transfer of plasmids only when chances are high, akin to sexual systems in more complex organisms.

International audience

MotivationThe reconstruction of ancestral genetic sequences from the analysis of contemporaneous data is a powerful tool to improve our understanding of molecular evolution. Various statistical criteria defined in a phylogenetic framework can be used to infer nucleotide, amino-acid or codon states at internal nodes of the tree, for every position along the sequence. These criteria generally select the state that maximizes (or minimizes) a given criterion. Although it is perfectly sensible from a statistical perspective, that strategy fails to convey useful information about the level of uncertainty associated to the inference.ResultsThe present study introduces a new criterion for ancestral sequence reconstruction, the minimum posterior expected error (MPEE), that selects a single state whenever the signal conveyed by the data is strong, and a combination of multiple states otherwise. We also assess the performance of a criterion based on the Brier scoring scheme which, like MPEE, does not rely on any tuning parameters. The precision and accuracy of several other criteria that involve arbitrarily set tuning parameters are also evaluated. Large scale simulations demonstrate the benefits of using the MPEE and Brier-based criteria with a substantial increase in the accuracy of the inference of past sequences compared to the standard approach and realistic compromises on the precision of the solutions returned.Availability and implementationThe software package PhyML (https://github.com/stephaneguindon/phyml) provides an implementation of the Maximum A Posteriori (MAP) and MPEE criteria for reconstructing ancestral nucleotide and amino-acid sequences.

International audience

International audience

International audience; In this chapter, we present recent computational and algorithmic advances for improving the inference of phylogenetic trees from the analysis of homologous genetic sequences under the maximum likelihood criterion. In particular, we detail how the use of matrix algebra at the core of Felsenstein’s pruning algorithm, combined with the architecture of modern day computer processors, leads to efficient techniques for optimizing edge lengths. We also discuss some properties of the likelihood function when considering the optimization of the parameters of mixture models that are used to describe the variation of rates-across sites .

Motivation: Most evolutionary analyses are based on pre-estimated multiple sequence alignment. Wong et al. established the existence of an uncertainty induced by multiple sequence alignment when reconstructing phylogenies. They were able to show that in many cases different aligners produce different phylogenies, with no simple objective criterion sufficient to distinguish among these alternatives. Results: We demonstrate that incorporating MSA induced uncertainty into bootstrap sampling can significantly increase correlation between clade correctness and its corresponding bootstrap value. Our procedure involves concatenating several alternative multiple sequence alignments of the same sequences, produced using different commonly used aligners. We then draw bootstrap replicates while favoring columns of the more unique aligner among the concatenated aligners. We named this concatenation and bootstrapping method, Weighted Partial Super Bootstrap (wpSBOOT). We show on three simulated datasets of 16, 32 and 64 tips that our method improves the predictive power of bootstrap values. We also used as a benchmark an empirical collection of 853 1-to-1 orthologous genes from seven yeast species and found wpSBOOT to significantly improve discrimination capacity between topologically correct and incorrect trees. Bootstrap values of wpSBOOT are comparable to similar readouts estimated using a single method. However, for reduced trees by 50% and 95% bootstrap thresholds, wpSBOOT comes out the lowest Type I error (less FP). Availability: The automated generation of replicates has been implemented in the T-Coffee package, which is available as open source freeware available from www.tcoffee.org. Supplementary information: Supplementary data are available at Bioinformatics online. This work was supported by the Spanish Ministry of Science Plan Nacional [BFU2008-00419 to P.D.T. and C.N.]; the Wellcome Trust [WT095908 to P.F.]; the INCEPTION project [PIA/ANR-16-CONV-0005 to O.G.]; the Taiwan Ministry of Science and Technology [106-2221-E-004-011-MY2 to J.-M.C.]. We acknowledge support of the European Molecular Biology Laboratory, the Spanish Ministry of Economy and Competitiveness, “Centro de Excelencia Severo Ochoa 2013-2017” and “The Human Project from Mind, Brain and Learning” of NCCU from the Higher Education Sprout Project by the Ministry of Education in Taiwan

International audience

International audience

The reconstruction of ancestral scenarios is widely used to study the evolution of characters along phylogenetic trees. One commonly uses the marginal posterior probabilities of the character states, or the joint reconstruction of the most likely scenario. However, marginal reconstructions provide users with state probabilities, which are difficult to interpret and visualize, whereas joint reconstructions select a unique state for every tree node and thus do not reflect the uncertainty of inferences. We propose a simple and fast approach, which is in between these two extremes. We use decision-theory concepts (namely, the Brier score) to associate each node in the tree to a set of likely states. A unique state is predicted in tree regions with low uncertainty, whereas several states are predicted in uncertain regions, typically around the tree root. To visualize the results, we cluster the neighboring nodes associated with the same states and use graph visualization tools. The method is implemented in the PastML program and web server. The results on simulated data demonstrate the accuracy and robustness of the approach. PastML was applied to the phylogeography of Dengue serotype 2 (DENV2), and the evolution of drug resistances in a large HIV data set. These analyses took a few minutes and provided convincing results. PastML retrieved the main transmission routes of human DENV2 and showed the uncertainty of the human-sylvatic DENV2 geographic origin. With HIV, the results show that resistance mutations mostly emerge independently under treatment pressure, but resistance clusters are found, corresponding to transmissions among untreated patients.

Inferring epidemiological parameters such as the R0 from time-scaled phylogenies is a timely challenge. Most current approaches rely on likelihood functions, which raise specific issues that range from computing these functions to finding their maxima numerically. Here, we present a new regression-based Approximate Bayesian Computation (ABC) approach, which we base on a large variety of summary statistics intended to capture the information contained in the phylogeny and its corresponding lineage-through-time plot. The regression step involves the Least Absolute Shrinkage and Selection Operator (LASSO) method, which is a robust machine learning technique. It allows us to readily deal with the large number of summary statistics, while avoiding resorting to Markov Chain Monte Carlo (MCMC) techniques. To compare our approach to existing ones, we simulated target trees under a variety of epidemiological models and settings, and inferred parameters of interest using the same priors. We found that, for large phylogenies, the accuracy of our regression-ABC is comparable to that of likelihood-based approaches involving birth-death processes implemented in BEAST2. Our approach even outperformed these when inferring the host population size with a Susceptible-Infected-Removed epidemiological model. It also clearly outperformed a recent kernel-ABC approach when assuming a Susceptible-Infected epidemiological model with two host types. Lastly, by re-analyzing data from the early stages of the recent Ebola epidemic in Sierra Leone, we showed that regression-ABC provides more realistic estimates for the duration parameters (latency and infectiousness) than the likelihood-based method. Overall, ABC based on a large variety of summary statistics and a regression method able to perform variable selection and avoid overfitting is a promising approach to analyze large phylogenies., Author summary Given the rapid evolution of many pathogens, analysing their genomes by means of phylogenies can inform us about how they spread. This is the focus of the field known as “phylodynamics”. Most existing methods inferring epidemiological parameters from virus phylogenies are limited by the difficulty of handling complex likelihood functions, which commonly incorporate latent variables. Here, we use an alternative method known as regression-based Approximate Bayesian Computation (ABC), which circumvents this problem by using simulations and dataset comparisons. Since phylogenies are difficult to compare to one another, we introduce many summary statistics to describe them and take advantage of current machine learning techniques able to perform variable selection. We show that the accuracy we reach is comparable to that of existing methods. This accuracy increases with phylogeny size and can even be higher than that of existing methods for some parameters. Overall, regression-based ABC opens new perspectives to infer epidemiological parameters from large phylogenies.

International audience; Phylogenetic reconstructions are essential in genomics data analyses and depend on accurate multiple sequence alignment (MSA) models. We show that all currently available large-scale progressive multiple alignment methods are numerically unstable when dealing with amino-acid sequences. They produce significantly different output when changing sequence input order. We used the HOMFAM protein sequences dataset to show that on datasets larger than 100 sequences, this instability affects on average 21.5% of the aligned residues. The resulting Maximum Likelihood (ML) trees estimated from these MSAs are equally unstable with over 38% of the branches being sensitive to the sequence input order. We established that about two-thirds of this uncertainty stems from the unordered nature of children nodes within the guide trees used to estimate MSAs. To quantify this uncertainty we developed unistrap, a novel approach that estimates the combined effect of alignment uncertainty and site sampling on phylogenetic tree branch supports. Compared with the regular bootstrap procedure, unistrap provides branch support estimates that take into account a larger fraction of the parameters impacting tree instability when processing datasets containing a large number of sequences.

International audience; Viral evolutionary analyses are confronted with increasingly large sequence data sets, both in terms of sequence length and number of sequences. This can result in considerable computational burden, not only to infer phylogenies but also to obtain associated estimates such as their time scales and phylogeographic patterns. Here, we illustrate two frequently-used approaches to obtain phylogenomic estimates of time-measured trees and spatial dispersal patterns for fast-evolving viruses. First, we discuss computationally efficient procedures that employ a fixed tree topology obtained through maximum likelihood inference to estimate molecular clock rates and phylogeographic spread for Dengue virus genomes. Using the same viral example, we also illustrate Bayesian phylodynamic inference that jointly infers time-measured trees and phylogeo-graphy, including covariates of spatial dispersal, from sequence and trait data. We highlight state-of-the-art efforts to perform such computations more efficiently. Finally, we compare the estimates obtained by both approaches and discuss their strengths and potential pitfalls.