Institut de Mathématiques de Marseille, UMR 7373


Accueil > Séminaires > Séminaires et Groupes de travail hebdomadaires > Mathématiques, Évolution, Biologie (MEB)

Séminaire Mathématiques, Évolution, Biologie (MEB)

par Didier Gilles, Keller Julien, Kopp Michael, Lozingot Eric, Tichit Laurent - publié le , mis à jour le



  • Mercredi 31 mars 2010 11:00-12:00 - Céline Vens - Katholieke Universiteit Leuven and INRA Sophia-Ant

    Top-down phylogenetic tree reconstruction

    Résumé : We propose a novel distance based method for reconstruction of phylogenetic trees. Our method is based on a conceptual clustering method that extends the decision tree learning approach. Our method starts from a single cluster and repeatedly splits it into subclusters until all sequences form a different cluster. We assume that a split can be described by referring to particular polymorphic locations, which makes such a divisive method computationally feasible.
    To define the best split, our method uses a criterion that is close to Neighbor Joining’s (NJ) optimization criterion, namely, constructing a phylogenetic tree with minimal total branch length. Our approach has a number of important advantages.
    First, by listing the polymorphic locations at the internal nodes, it provides an explanation for the resulting tree topology.
    Second, the resulting trees can be used as classification trees to add new sequences to the phylogeny.
    Third, the top-down tree growing process can be stopped before a complete tree is generated, yielding an efficient gene or protein subfamily identification approach.
    Our algorithm, called Clus-I+, can be found at dtai/clus.

    JPEG - 8.7 ko
    Céline Vens

    Lieu : FRUMAM - Aix-Marseille Université - Site St Charles
    3, place Victor Hugo - case 39
    13331 MARSEILLE Cedex 03

    Exporter cet événement

  • Vendredi 31 janvier 2014 11:00-12:00 -

    Séminaire Mathématiques, Évolution, Génome (TBA)

  • Mardi 9 juin 2015 11:00-12:00 - Nicolas Lartillot - LBBE, Lyon

    The molecular comparative method : Bayesian integrative models of macro-evolutionary processes

    Résumé : Estimating divergence times, understanding molecular evolutionary mechanisms, or testing macroevolutionary hypotheses about patterns of diversification and morphological evolution, are usually considered as separate research questions, adressed by distinct, although overlapping, scientific communities. Yet, many connections would deserve to be made between these various topics in evolutionary sciences. Over the last years, several attempts at integrating some of these various aspects of macro-evolutionary sciences have been made, using hierarchical modeling approaches. After reviewing them, I will more specifically present a Bayesian framework for modeling the macroevolutionary process in an integrative manner. This framework can be seen as a fusion between the classical comparative method and methods for divergence time estimation. Taking as an input a multiple sequence alignment, data about life-history traits of extant species, and fossil calibrations, it then jointly estimates divergence times, life-history evolution and correlations between substitution patterns and quantitative traits. Application of the method to placental mammals reveals extensive correlations between life-history and molecular evolution, providing stimulating observations for testing macroevolutionary hypotheses.

    JPEG - 15.6 ko
    Nicolas LARTILLOT

    Lieu : Salle LSH 404 - Aix-Marseille Université - Site St Charles
    3, place Victor Hugo - case 39
    13331 MARSEILLE Cedex 03

    Exporter cet événement

  • Mercredi 24 juin 2015 11:00-12:00 - Sebastian MATUSZEWSKI - Université de Lausanne

    Catch me if you can : Adaptation from standing genetic variation to a moving phenotypic optimum

    Résumé : Adaptation lies at the heart of Darwinian evolution. Natural populations are constantly faced with environmental changes that force them to either adapt or go extinct - a problem that is aggravated by human-induced global change. Therefore, increasing our understanding of the adaptive process is important for both basic and applied research.
    Accordingly, numerous studies have tried to provide a formal framework for the description of the adaptive process. Out of these, two complementary modeling approaches have emerged : While so-called adaptive-walk models consider adaptation from the successive fixation of de-novo mutations only, quantitative genetic models, on the other hand, assume that adaptation proceeds exclusively from pre-existing standing genetic variation. The latter approach, however, has focused on short-term evolution of population means and variances rather than on the statistical properties of adaptive substitutions.
    Thus, I will here address what has been described as "the most obvious theoretical limitation when describing the adaptive process" and propose an analytical framework for the genetic basis of adaptation from standing genetic variation in terms of the effect-size distribution of individual alleles.
    This approach addresses one of the central questions in this context : From the multitude of standing genetic variants segregating in a population, which are the ones that ultimately become fixed and contribute to adaptation, and how does their distribution differ from that of de-novo mutations ?

    JPEG - 17.3 ko
    Sebastian MATUSZEWSKI

    Lieu : FRUMAM (3ème étage) - Aix-Marseille Université - Site St Charles
    3, place Victor Hugo - case 39
    13331 MARSEILLE Cedex 03

    Exporter cet événement

  • Mercredi 1er juillet 2015 11:00-12:00 - Olivier CHABROL - I2M, Marseille

    Détection de signatures moléculaires de convergences évolutives

    Résumé : La regretté Thylacine et le Loup ont une boite crânienne, des mâchoires, extrêmement similaires. Or l’un était un marsupial et l’autre est un placentaire.
    Ce bel exemple de convergence évolutive est-il observable à l’échelle génétique ?
    Nous proposons une nouvelle approche pour détecter la convergence au niveau génétique et évaluons ses performances par rapport à celle proposée par Foote et al (Nature 2015).

    Lieu : Salle des séminaires 304-306 (3ème étage) - Institut de Mathématiques de Marseille (UMR 7373)
    Site Sud
    Campus de Luminy, Case 907
    13288 MARSEILLE Cedex 9

    Exporter cet événement

  • Mercredi 23 septembre 2015 14:00-15:00 - Pierre PONTAROTTI - I2M, Marseille

    Roadmap to study convergent evolution, a proposition for evolutionary system biology approaches

    Résumé : Most of the authors studying convergent evolution think about the apparition of similar phenotype in two evolutionary independent lineages. (Conway Morris 2003 , McGee 2011, Losos 2011 , Gordon et al 2015). From this broad definition the authors focus in they case studies : echolocation (Parker et al 2013) or the repetitive adaptations of marine mammals (Foote et la 2015). One of The goal of my presentation is to show that these cases of “repetitive similarities” should be defined in a better way. Lot of tries have been done since the 20th century, but these add more confusions (Arendt and Reznick 2008). From ours point of view the confusion came from the fact that most of the authors gave sub definitions based on the genetic mechanisms involved in the evolution of the convergent evolving character. We propose two neologims that can be apply to all the biological level : isoconvergent and alloconvergent evolution Iso from the same ancestral state allo from a different ancestral state . This distinction is important as in the case iso convergent evolution , one can suspect that the underlying molecular mechanism could be similar(Stern and Orgozozo 2008) . Such cases could be used as meta-model (Kopp 2009) that in turn can be used to decipher biological mechanisms at genetic, epigenetic, transcriptional level and other biological levels. We propose 1) to reanalyse all the case of convergent evolutions at phenotypic level decribed in the litterature and sort out the case of isoconvergences,2) A strategy to identify undescribed cases of iso convergent evolution and 3) a road map to study thes cases at the different biological level and 4) the development of a data base. (joint work with Justine Dardaillon and Laurent Tichit, I2M)

    JPEG - 8.1 ko

    Lieu : FRUMAM - Aix-Marseille Université - Site St Charles
    3, place Victor Hugo - case 39
    13331 MARSEILLE Cedex 03

    Exporter cet événement

  • Mercredi 23 septembre 2015 15:00-16:00 - Gilles DIDIER - I2M, Marseille

    Detecting molecular signatures of convergent evolution, followed by, Estimating diversification rates from the fossil record

    Résumé : The talk is divided into two independent parts.
    The first focuses on the detection of evolutionary traces of phenotypic convergence in genetic sequences. We consider a complex trait which is shared by a subset of species and not observed among others. In order to assess in what extent the genetic sequences of the species support the convergent evolution of the trait, we ask the question “How much positions have evolved independently to a same amino acid (or nucleotide, i.e "letter" of sequences) for the species bearing the trait ?”. This question cannot be answered without uncertainty since we don’t have access to the past of the genetic sequences (only the contemporary ones are available). We show how to compute the expectation of this number for each position under a standard model of sequences evolution. (joint work with O. Chabrol and P. Pontarotti, I2M)
    In the second part, we are interested in how fast species born and die (i.e. diversify). Diversification rates are estimated from phylogenies (i.e. trees figuring the parenthood of species) which contains extant and (sometimes) fossil taxa. By nature, rate estimations depend heavily on the time data provided in phylogenies, which are divergence times and (when used) fossil ages. Among these temporal data, fossil ages are by far the most precisely known (divergence times are inferences calibrated with fossils). We present an original method to compute the likelihood of a phylogenetic tree with fossils in which the only known time information is the fossil ages.
    Testing our approach on simulated data shows that the maximum likelihood rate estimates from the phylogenetic tree shape and the fossil dates are almost as accurate as the ones obtained by taking into account all the data, including the divergence times. Moreover they are substantially more accurate than the estimates obtained only from the exact divergence times (without taking into account the fossil record). We finally estimate the diversification and fossilization rates from a paleontological dataset (eupelycosaurs tree). (joint work with Michel Laurin, Museum d’Histoire Naturelle de Paris)

    JPEG - 20.1 ko
    Gilles DIDIER

    Lieu : FRUMAM - Aix-Marseille Université - Site St Charles
    3, place Victor Hugo - case 39
    13331 MARSEILLE Cedex 03

    Exporter cet événement

  • Mercredi 23 septembre 2015 16:00-17:00 - Michael KOPP - I2M, Marseille

    Adaptive walks following a moving optimum : (Mal)adaptation and population extinction

    Résumé : We analyze a stochastic process describing an evolving population adapting to a gradually changing environment with a moving phenotypic optimum. Our main simplifying assumption is that beneficial mutations that escape stochastic loss are fixed instantaneously. Adaptation thus resembles a jump process. Taking the limit of small jumps enables us to derive analytical approximations for the mean degree of maladaptation (how far the population lags behind the optimum) and its variance. Simulations show that this approximation is accurate as long as the mean lag is greater than the size of a typical mutation. Furthermore, in the small-jumps limit, the process converges to an Ornstein-Uhlenbeck process around the long-term mean. Using published results about first-passage times of OU processes, we derive an approximation for the time until maladaptation becomes so severe that the population is in danger of extinction. This time has an approximately exponential distribution with a mean that decreases exponentially with the inverse of the speed of environmental change relative to the rate and size of new mutations. We also derive a critical rate of environmental change, beyond which the mean degree of adaptation is to poor to allow population survival. (joint work with Elma Nassar and Etienne Pardoux, I2M)

    JPEG - 8.8 ko
    Michael KOPP

    Lieu : FRUMAM - Aix-Marseille Université - Site St Charles
    3, place Victor Hugo - case 39
    13331 MARSEILLE Cedex 03

    Exporter cet événement

  • Mercredi 30 septembre 2015 11:00-12:00 - Pierre PUDLO - I2M, Marseille

    Problème d’inférence en génétique des populations structurées sous neutralité

    Résumé : Sous neutralité, le polymorphisme génétique porte la trace de l’évolution démographique : changement de taille de populations, séparation en sous-populations ou fusion, migration entre populations, etc. Reconstruire cette histoire démographique à partir de données génétiques actuelles est une question importante. À titre d’exemple, on peut se demander : quelles sont les routes d’invasion de la coccinelle asiatique ? les pygmées dérivent-ils tous d’une population ancestrale commune ? quelle est l’histoire de l’espèce humaine (sortie de l’Afrique, etc.) ? De plus, dans une perspective de détection de zone sous sélection, la distribution du polymorphisme ainsi reconstruite peut servir d’hypothèse nulle.
    D’un point de vue statistique, tirer de l’information du polymorphisme actuel nous confronte à un problème difficile où la fonction de vraisemblance n’est pas calculable. En effet, la distribution génétique de l’échantillon s’explique par un objet latent ou caché de grande dimension (l’histoire génétique passée), incluant en particulier des structures combinatoires complexes (l’arbre généalogique des gènes de l’échantillon).
    Nous présenterons quelques méthodes d’inférences actuelles pour faire face à cette difficulté majeure, basées soit sur de l’échantillonnage préférentiel (importance sampling), soit sur des comparaisons à de nombreuses simulations (méthodes ABC pour Approximate Bayesian Computation) et quelques illustrations numériques.

    JPEG - 8.7 ko
    Pierre PUDLO

    Lieu : FRUMAM - Aix-Marseille Université - Site St Charles
    3, place Victor Hugo - case 39
    13331 MARSEILLE Cedex 03

    Exporter cet événement

  • 1 | 2 | 3 | 4 | 5

groupe de travail

Manifestation scientifique

Nature Séminaire
Intitulé Mathématiques, Évolution, Biologie (MEB)
Responsables Gilles Didier
Michael Kopp
Équipe de rattachement Mathématiques, Évolution, Biologie (MEB)
du Groupe ALEA
Fréquence Hebdomadaire
Jour-Horaire Mercredi, à 11h
Lieu FRUMAM, St Charles (accès)

Contacts :

Le séminaire MEB est la continuation du séminaire MEG, qui était le séminaire conjoint des équipes « Méthodes Mathématiques pour la Génomique » (MMG) de l’ex-IML et « Évolution Biologique et Modélisation » (EBM) de l’ex-LATP. Lors de la création de l’I2M, les deux équipes ont été réunies pour former l’équipe MEB (Mathématiques, Évolution, Biologie).
L’équipe MEB est pluri-disciplinaire, à la fois dans ses thèmes et sa composition. Elle inclut des mathématiciens, des informaticiens et des biologistes, réunis autour de la modélisation et l’étude de phénomènes biologiques, en particulier liés à l’évolution. À l’image de l’équipe, le séminaire invite des orateurs venant de différentes disciplines. Les exposés ont vocation à attirer un public plus large que les seuls membres de l’équipe, ou même que l’I2M.
Le séminaire MEB s’intéresse à la fois à la modélisation de phénomènes biologiques et au traitement de leurs données. Le thème principal du séminaire est l’évolution biologique, mais tout exposé relevant d’un des axes de l’équipe, ou plus généralement des interactions entre mathématiques et biologie, y est naturellement bienvenu.
Les domaines abordés sont principalement :
- l’évolution biologique,
- les probabilités et statistique,
- l’algorithmique,
- la génétique des populations,

Archives : séances de l’ex-séminaire "Mathématiques, Evolution, Génome" (MEG), créé en octobre 2008, dont les organisateurs étaient : E. Pardoux, P. Pontarotti, E. Rémy.