Symbiose seminars

  • TBA

    Pierre Morisse
    Thursday, December 7, 2017 - 10:30
    Room Minquiers
    Talk abstract: 

    TBA

  • TBA

    Bernardo Clavijo (Cambridge, England)
    Thursday, November 2, 2017 - 10:30 to Tuesday, November 7, 2017 - 10:30
    Room Minquiers
    Talk abstract: 

    TBA

  • Graphs as an universal modelling tool

    Nicola Yanev
    Thursday, October 5, 2017 - 10:30
    Room Aurigny
    Talk abstract: 

    The classical topics in optimization graph theory are extended by several new problems,  inspired by the practice:
    - new classes of matching problems in bipartite graphs: inspired by still actual problematic of recognition of protein folding, comparison of protein 3D structure and even the travelling salesman problem,
    - shortest walks over the vertices of the unit hypercube: inspired by needs of  optimization module of a warehouse management system, used to schedule the loading operation in a huge warehouse;
    - finding the fixed number of connected components with the minimum maximal diameter: inspired by the challenge to determine mathematical electoral districts in a single member constituency ,
    - finding the minimum clique cover in a colored graph: inspired by a new approach to the well known classification problem, that is - to recover a partially known stepwise function . 

  • Un voyage bioinformatique dans l’océan des données cytométriques de grandes dimensions

    Nicolas Tchichek
    Thursday, September 21, 2017 - 10:30
    Room Minquiers
    Talk abstract: 

    La cytométrie en flux et de masse sont des techniques expérimentales qui permettent de mesurer des protéines exprimées par des cellules à une résolution cellulaire (single-cell profiling). La cytométrie en flux peut actuellement mesurer jusqu'à 18 marqueurs cellulaires pour des millions de cellules. La cytométrie de masse, qui a été introduite plus récemment, peut mesurer jusqu'à 40 marqueurs cellulaires pour des centaines de milliers de cellules.

    L’analyse de ces données nécessite des nouvelles approches bioinformatiques pour identifier les principales populations cellulaires dans ces matrices tridimensionnelles de grandes dimensions. Ainsi, des algorithmes comme SPADE, viSNE ou Citrus ont été proposé pour identifier automatiquement les populations cellulaires présentes dans des profils cytométriques. Une fois ces populations cellulaires détectées des analyses bioinformatiques complémentaires sont nécessaire pour sélectionner les populations ayant un intérêt biologique particulier.

    Je présenterai dans ce séminaire des approches bio-informatiques que nous avons développées récemment dans notre laboratoire. Plus précisément, nous avons conçus des algorithmes qui permettent : (i) d’identifier les populations cellulaires ayant des phénotypes et comportements pertinents ; (ii) de comparer statistiquement les phénotypes des populations cellulaires ; et (iii) de combiner les informations phénotypiques de populations cellulaires obtenues à partir d’expériences différentes.

    Enfin, je conclurai en discutant des limites des approches bio-informatiques existantes et ainsi des nouveaux défis à relever dans le futur.

  • Towards a Logical Framework for Systems Biology

    Joelle Despeyroux - INRIA Sophia-Antipolis
    Thursday, September 14, 2017 - 10:30
    Room Minquiers
    Talk abstract: 

    We advocates here the use of (mathematical) logic for systems biology, as a unified framework well suited for both modeling the dynamic behaviour of biological systems, expressing properties of them, and verifying these properties.
    The potential candidate logics should have a traditional proof theoretic pedigree (including a sequent calculus presentation enjoying cut-elimination and focusing), and should come with (certified) proof tools. Beyond providing a reliable framework, this allows the adequate encodings of our biological systems.
    We have up to now two main candidate logics; both are modal extensions of linear logic.
    The examples we have considered so far are very simple ones - coming with completely formal (interactive) proofs in Coq.
    Future works includes using automatic provers, which will eventually extend existing automatic provers for linear logic. This should enable us to specify and study more realistic examples in systems biology and biomedicine.

  • Formal Language Representation and Modeling of Bio-molecular Structures by Matrix Insertion-Deletion Systems

    Lakshmanan
    Wednesday, July 5, 2017 - 10:30 to Saturday, July 29, 2017 - 10:30
    Room Aurigny
    Talk abstract: 

    David Searls initiated the representation of molecules as strings and modeled structures formed by the molecules using formal grammars. The work opened the doors to tackle many problems in computational biology like predicting pseudoknot structures of biomolecules. Subsequently, Sakakibara, Hausler, Eddy, Rivas, Umera and others introduced various grammar formalisms for modeling predominantly noticed bio-molecular structures such as stem-loop, pseduoknot, attenuator, clover-leaf, dumbbell. However, there is no unique grammar formalism that models all the structures discussed in literature. In this talk, we shall study a recently introduced grammar formalism namely matrix insertion-deletion systems. In this system, set of rules are applied in order which helps to synchronize the elements even at far. With this system, we will model several intra and inter molecular structures of DNA and RNA. We will also model parallel and anti-parallel beta sheets of protein molecules with the system. 

  • Algorithmique pour les peptides non ribosomiques

    Yoann Dufresne (univ. lille)
    Thursday, June 29, 2017 - 10:30
    Room Aurigny
    Talk abstract: 
    La composition monomérique de polymères joue un rôle essentiel dans la comparaison de structures et dans la biologie de synthèse. Cependant, la plupart des ressources moléculaires en ligne donne accès à la structure atomique des molécules et non à leur structure monomérique. C'est dans ce contexte que j'ai développé au cours de ma thèse, un outil appelé Smiles2Monomers (s2m) dédié à la création d'un pont entre ces deux structures.
     
    Au cours de la présentation, nous reviendrons sur les aspects algorithmiques de graphes liés aux recherches de structures. Nous verrons, entre autre, comment il est possible d'effectuer des recherches exactes de sous graphes en optimisant le temps dépensé.
    Au delà de l'aspect algorithmique, je présenterai également les applications concrètes de ces algorithmes dans l'optique d'améliorer la base de donnée Norine. Nous pointerons les avantages de l'utilisation de s2m autant dans la limitation du nombre d'erreurs ajoutées à la base par les utilisateurs, la correction des erreurs déjà présentes ainsi que l'entrée automatique de nouvelles données.
  • An eco-systems view of complex natural microbiomes

    Samuel Chaffron (Universite Nantes)
    Thursday, June 15, 2017 - 10:30
    Room Aurigny
    Talk abstract: 

    Microbial communities play crucial ecological roles on our planet, impacting all ecosystems at various levels. Notably, they sustain Earth’s biogeochemical cycles in the oceans, protect plants from pathogens, and also influence host nutrition, immunity and development in humans. Next-Generation Sequencing (NGS) and functional genomics technologies (so called omics approaches) are revolutionizing the field of environmental microbiology and are reshaping our view on microbial ecosystems. These approaches enable the study of ecological systems at the molecular scale and are revealing the under-explored diversity and complexity of microbial ecosystems. Yet our ability to understand and predict the structures and functions of these complex and dynamic microbial communities is very limited. Notwithstanding, the genomic content of naturally co-occurring microbes can now be investigated in situ, and their metabolic networks can be reconstructed and integrated to gain insights into their physiology (from an environmental genotype to its phenotype). Past and on-going work presented here will ultimately lead to the design of integrative models through a microbial metabolic modeling platform integrating omics data, ecological information and metabolic networks reconstruction to further understand and predict microbial communities and ecosystems in and around us. This Microbial Systems Ecology (ECOSYSMIC) platform will enable to go beyond the simple description of natural microbial communities by building computational models to gain a predictive understanding of community function and dynamics, and acquire a mechanistic understanding of microbial species interactions and ecosystem functioning.

  • Filling Scaffolds

    Laurent Bulteau ( LIGM)
    Thursday, June 8, 2017 - 10:30
    Room Aurigny
    Talk abstract: 

     

    The Scaffold Filling problem was introduced by Muñoz et al. with the objective
    of using, for genomic distance purposes, not only perfectly sequenced genomes but
    also unfinished drafts. Indeed, with the development of NGS technologies, it has
    become much faster and cheaper to produce a first draft of any genome. However,
    the cost of “polishing” the draft to a complete sequence has not decreased with the
    same rate, thus many species are left with a genome in scaffold form. In such a form,
    a genome is only known as a series of contigs (i.e., contiguous segments of genes),
    separated by unknown gaps, sometimes with an indication on the length of the gap.
    With the help of a reference genome, that is, the complete genome of a close-
    enough species, one can hope to fill the gaps. Indeed, Muñoz et al. proposed a
    polynomial-time algorithm computing a most parsimonious rearrangement scenario
    between a scaffold and a reference genome, thereby completing the scaffold. However,
    this approach can only be applied in the absence of duplications – the problem
    becomes computationaly hard otherwise. From then on, several algorithms have been
    proposed to deal with gene duplications in order to compute simplified rearrangement
    distances, using both approximation and parameterized techniques. We will
    review these methods, as well as possible extensions of the problem.

  • A Combinatorial Algorithm for Microbial Consortia Synthetic Design

    Alice Julien Laferrière (Lyon)
    Thursday, April 27, 2017 - 10:30
    Room Aurigny
    Talk abstract: 
    Résumé en français plus bas. 
     
     
    -------------------------------------------------------------------------------------
     
    MULTIPUS (MULTIple species for the synthetic Production of Useful biochemical Substances) is a method to infer microbial communities and metabolic pathways to product specific target compounds from a set of defined substrates.
     
    We use a weighted directed hypergraph and we merge several metabolic networks, including possible transports and exogenous reactions. The problem translates into the enumeration of Directed Steiner Hypertrees. 
     
    We proposed to solve this enumeration problem using an dynamic programming algorithm or an answer-set programming (ASP) solver. Finally, we illustrated MULTIPUS with two case study, the production of antibiotics using a synthetic consortia and the artificial association of a methanogenic archae with Klebsiella pneumoniae.
     
     
    -------------------------------------------------------------------------------------------------------------
    Nous avons réfléchi à une méthode permettant d'inférer des voies métaboliques d’intérêt pour la production de composés chimiques au sein d'une communauté de micro-organismes.
     Dans cette méthode, nous représentons les réseaux métaboliques sous la forme d'hypergraphes dirigés et pondérés et permettons de prendre en compte  différentes souches ou espèces ainsi que des transports de composés ou encore l'insertion de réactions par ingénierie génétique.
     
    Le problème correspond alors à un problème d'énumération d'hyper-arbres de Steiner dirigés. 
     
    Nous avons proposé de résoudre cela soit avec un algorithme de programmation dynamique paramétré, soit par une approche de programmation par ensemble réponse (ASP). 
    Ces deux approches sont implémentées dans  MULTIPUS (MULTIple species for the synthetic Production of Useful biochemical Substances). Enfin nous avons illustré l’intérêt de MULTIPUS dans deux cas d'applications, une communauté synthétique pour la production d'antibiotiques et un consortium artificiel.

Pages