L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion d'articles scientifiques de niveau recherche, publiés ou non, et de thèses, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Derniers Dépôts 

Chimie

Économie et finance quantitative

Informatique

Mathématiques

Physique

Planète et Univers

Science non linéaire

Sciences cognitives

Sciences de l'environnement

Sciences de l'Homme et Société

Sciences de l'ingénieur

Sciences du Vivant

Statistiques

Monitoring of trace element in groundwater by DGT Jean Philippe Ghestem et al. DGT passive samplers, are often used for the monitoring of trace elements in inland or marine surface waters. In these contexts, they offer significant advantages over spot sampling by their performance in terms of preconcentration and integration of changes in concentration over time. At the opposite, there is little application of these tools for underground water context. The needs for underground water monitoring are not the same as surface water. However it seems interesting to test the capabilities and limitations of these passive samplers in groundwater. In the proposed study DGT-chelex have been deployed in four campaigns of about 14 days in a piezometer installed on an uncontaminated sites. Spot sampling occurred before and after deployment of samplers and also with and without purging of the piezometer. During one campaign, DGT of varying diffusive gel thickness were used to study the thickness of the diffusive boundary layer in an underground environment (with low water flow) following the method described by Kent W. Et al. (Accuracy Of The DGT-Diffusive Boundary Layer and Effective Sampling Area Considerations, Anal. Chem. (2006), 78, 3780-3787). An estimate of this boundary layer has been done and it shows that in the context of this study (and probably for most of underground water), this effect can not be neglected. Concentrations of ten trace elements (Al, Cr, Mn, Fe, Ni, Co, Cu, Zn, Cd, Pb) were followed on these 4 campaigns and over a 4 months period. The results were interpreted in terms of reproducibility of measures over 4 months and DGT concentrations were compared with spot samples concentrations.

Effects of Ca2+ on supramolecular aggregation of natural organic matter in aqueous solutions: A comparison of molecular modeling approaches A. Kalinichev et al. Natural organic matter (NOM) represents a complex molecular system that cannot be fully characterized compositionally or structurally in full atomistic detail. This makes the application of molecular modeling approaches very difficult and significantly hinders quantitative investigation of NOM properties and behavior by these otherwise very efficient computational techniques. Here we report and analyze three molecular dynamics (MD) simulations of Ca2+ complexation with NOM in aqueous solutions in an attempt to quantitatively assess possible effects of model- and system size-dependence in such simulations. Despite some obvious variations in the computed results that depend on the size of the simulated system and on the parameters of the force field models used, all three simulations are quite robust and consistent. They show Ca2+ ions associated with 35-50% of the NOM carboxylic groups at near-neutral pH and point to a strong preference for the stability of bidentate-coordinated contact ion pairs. The degree and potential mechanisms of NOM supramolecular aggregation in the presence of Ca2+ ions in solution are also assessed on a semi-quantitative level from two larger-scale MD simulations.

Matériaux de type LaFe13-xSix à fort pouvoir magnétocalorique - Synthèse et optimisation de composés massifs et hypertrempés - Caractérisations fondamentales Mariana Rosca La réfrigération magnétique à température ambiante, une alternative à la réfrigération classique, nécessite des matériaux magnétocaloriques performants. L'effet magnétocalorique se traduit par une variation d'entropie et une variation de température du matériau lorsqu'il est soumis à un champ magnétique. Les composées de type LaFe13-xSix présentent des atouts intéressants. Nous avons tout d'abord exploré cette famille de composé. Par substitutions chimiques et par insertion d'éléments interstitiels, nous avons étudié plusieurs compositions dont celles de type LaFe13-x-yCoySix et les hydrures LaFe13-xSixHy qui se sont révélées être les plus prometteuses. Les caractérisations structurales, microstructurales, thermiques et magnétiques entreprises nous ont guidés dans l'optimisation des méthodes d'élaboration : fusion à arc ou induction, des traitements thermiques, du contrôle de la stœchiométrie, etc. Dans un second temps nous avons développé la mise au point de ces matériaux à plus grande échelle en collaboration avec des partenaires industriels.

Credit Risk with asymmetric information on the default threshold Caroline Hillairet et al. We study the impact of asymmetric information in a general credit model where the default is triggered when a fundamental diff usion process of the firm passes below a random threshold. Inspired by some recent technical default events during the fi nancial crisis, we consider the role of the firm's managers who choose the level of the default threshold and have complete information. However, other investors on the market only have partial observations either on the process or on the threshold. We specify the accessible information for di fferent types of investors. Besides the framework of progressive enlargement of fi ltrations usually adopted in the credit risk modelling, we also combine the results on initial enlargement of filtrations to deal with the uncertainty on the default threshold. We consider several types of investors who have di fferent information levels and we compute the default probabilities in each case. Numerical illustrations show that the insiders who have extra information on the default threshold obtain better estimations of the default probability compared to the standard market investors.

Continuous time regime switching model applied to foreign exchange rate. Stéphane Goutte et al. Modified Cox-Ingersoll-Ross model is employed, combining with Hamilton (1989) type Markov regime switching framework, to study foreign exchange rates, where all parameter values depend on the value of a continuous time Markov chain. Basing on real data of some foreign exchange rates, the Expectation-Maximization algorithm is extended to this more general model and it is applied to calibrate all parameters. We compare the obtained results regarding to results obtained with non regime switching models and notice that our results match much better the reality than the others without Markov switching. Furthermore, we illustrate our model on various foreign exchange rate data and clarify some significant eco- nomic time periods in which financial or economic crisis appeared, thus, regime switching obtained.

Pareto optima and equilibria when preferences are incompletely known Guillaume Carlier et al. An exchange economy in which agents have convex incomplete preferences defined by families of concave utility functions is considered. Sufficient conditions for the set of efficient allocations and equilibria to coincide with the set of efficient allocations and equilibria that result when each agent has a utility in her family are provided. Welfare theorems in an incomplete preferences framework therefore hold under these conditions and efficient allocations and equilibria are characterized by first order conditions.

Predicting drug side-effect profiles: a chemical fragment-based approach. Edouard Pauwels et al. BACKGROUND: Drug side-effects, or adverse drug reactions, have become a major public health concern. It is one of the main causes of failure in the process of drug development, and of drug withdrawal once they have reached the market. Therefore, in silico prediction of potential side-effects early in the drug discovery process, before reaching the clinical stages, is of great interest to improve this long and expensive process and to provide new efficient and safe therapies for patients. RESULTS: In the present work, we propose a new method to predict potential side-effects of drug candidate molecules based on their chemical structures, applicable on large molecular databanks. A unique feature of the proposed method is its ability to extract correlated sets of chemical substructures (or chemical fragments) and side-effects. This is made possible using sparse canonical correlation analysis (SCCA). In the results, we show the usefulness of the proposed method by predicting 1385 side-effects in the SIDER database from the chemical structures of 888 approved drugs. These predictions are performed with simultaneous extraction of correlated ensembles formed by a set of chemical substructures shared by drugs that are likely to have a set of side-effects. We also conduct a comprehensive side-effect prediction for many uncharacterized drug molecules stored in DrugBank, and were able to confirm interesting predictions using independent source of information. CONCLUSIONS: The proposed method is expected to be useful in various stages of the drug development process.

Ligand scaffold hopping combining 3D maximal substructure search and molecular similarity. Flavien Quintus et al. BACKGROUND: Virtual screening methods are now well established as effective to identify hit and lead candidates and are fully integrated in most drug discovery programs. Ligand-based approaches make use of physico-chemical, structural and energetics properties of known active compounds to search large chemical libraries for related and novel chemotypes. While 2D-similarity search tools are known to be fast and efficient, the use of 3D-similarity search methods can be very valuable to many research projects as integration of "3D knowledge" can facilitate the identification of not only related molecules but also of chemicals possessing distant scaffolds as compared to the query and therefore be more inclined to scaffolds hopping. To date, very few methods performing this task are easily available to the scientific community. RESULTS: We introduce a new approach (LigCSRre) to the 3D ligand similarity search of drug candidates. It combines a 3D maximum common substructure search algorithm independent on atom order with a tunable description of atomic compatibilities to prune the search and increase its physico-chemical relevance. We show, on 47 experimentally validated active compounds across five protein targets having different specificities, that for single compound search, the approach is able to recover on average 52% of the co-actives in the top 1% of the ranked list which is better than gold standards of the field. Moreover, the combination of several runs on a single protein target using different query active compounds shows a remarkable improvement in enrichment. Such Results demonstrate LigCSRre as a valuable tool for ligand-based screening. CONCLUSION: LigCSRre constitutes a new efficient and generic approach to the 3D similarity screening of small compounds, whose flexible design opens the door to many enhancements. The program is freely available to the academics for non-profit research at: http://bioserv.rpbs.univ-paris-diderot.fr/LigCSRre.html.

Making Sense of Variations: Introducing Alternatives in Speech Synthesis Nicolas Obin et al. This paper addresses the use of speech alternatives to enrich speech synthesis systems. Speech alternatives denote the variety of strategies that a speaker can use to pronounce a sentence - depending on pragmatic constraints, speaking style, and specific strategies of the speaker. During the training, symbolic and acoustic characteristics of a unit-selection speech synthesis system are statistically modelled with context-dependent parametric models (GMMs/HMMs). During the synthesis, symbolic and acoustic alternatives are exploited using a GENERALIZED VITERBI ALGORITHM (GVA) to determine the sequence of speech units used for the synthesis. Objective and subjective evaluations supports evidence that the use of speech alternatives significantly improves speech synthesis over conventional speech synthesis systems.

Direct competition results from strong competiton for limited resource Sepideh Mirrahimi et al. We study a model of competition for resource through a chemostat-type model where species consume the common resource that is constantly supplied. We assume that the species and resources are characterized by a continuous trait. As already proved, this model, although more complicated than the usual Lotka-Volterra direct competition model, describes competitive interactions leading to concentrated distributions of species in continuous trait space. Here we assume a very fast dynamics for the supply of the resource and a fast dynamics for death and uptake rates. In this regime we show that factors that are independent of the resource competition become as important as the competition efficiency and that the direct competition model is a good approximation of the chemostat. This is performed through asymptotic analysis, introducing different scales for the resource renewal rate and the uptake rate. The mathematical difficulty relies in a possible initial layer for the resource dynamics. The chemostat model comes with a global convex Lyapunov functional. We show that the particular form of the competition kernel derived from the uptake kernel, satisfies a positivity property which is known to be necessary for the direct competition model to enjoy the related Lyapunov functional.

Feel++: A Computational Framework for Galerkin Methods and Advanced Numerical Methods Christophe Prud'Homme et al. This paper presents an overview of a unified framework for finite element and spectral element methods in 1D, 2D and 3D in \cpp called \Feel. The article is divided in two parts. The first part provides a digression through the design of the library as well as the main abstractions handled by it, namely, meshes, function spaces, operators, linear and bilinear forms and an embedded variational language. In every case, the closeness between the language developed in \Feel and the equivalent mathematical objects is highlighted. In the second part, examples using the mortar, Schwartz (non)overlapping, three fields and two ficticious domain-like methods (the Fat Boundary Method and the Penalty Method) are presented and numerically solved in the scope of the library.

Bougerol's identity in law and extensions Stavros Vakeroudis We present a list of equivalent expressions and extensions of Bougerol's celebrated identity in law, obtained by several authors. We recall well-known results and the latest progress of the research associated with this celebrated identity in many directions, we give some new results and possible extensions and we try to point out open questions.

Photostability of dye molecules trapped in solid matrices Arnaud Dubois et al. The photostability of dye molecules trapped in transparent solid matrices synthesized by the solgel technique was studied both experimentally and theoretically using a model with numerical and approximate analytical solutions. The model is based on a one-photon photodestruction process with the creation of an absorbing bleached molecule. We give the number of photons that different trapped dye molecules can absorb on average before they are bleached. Dyes such as Perylene Red, Perylene Orange, Pyrromethenes 567 and 597, Rhodamines 6G and B, DCM, a Xanthylium salt, and Neon Red were investigated; significant differences were observed. Some dye molecules in solvents were also studied; increased stability resulted when the molecules were trapped in solid matrices.

Phase-map measurements by interferometry with sinusoidal phase modulation and four integrating-buckets Arnaud Dubois Phase-shifting interferometry based on the integrating-bucket technique with sinusoidal phase modulation is studied theoretically and demonstrated experimentally to obtain phase maps from double-beam interferometers. The method uses four frames obtained by integration of the time-varying intensity in an interference pattern during the four quarters of the modulation period. An optimum sinusoidal phase modulation is found to minimize the effect of the additive noise. The absolute accuracy of the phase measurements is discussed. Possible applications of the method are demonstrated with two interference microscopes with which the phase modulation is achieved by sinusoidal oscillation of a mirror attached to a piezoelectric transducer and by sinusoidal birefringence modulation with a photoelastic modulator. In both experimental arrangements, phase images can be produced in real time at a rate of several hertz. Noise measurements are reported and compared with theory.

DNA microarray inspection by interference microscopy Laurent Vabre et al. DNA microarrays are important new objects for biologists, offering new methods for high speed genetic studies. Many of them are manufactured by depositing small spots, containing DNA molecules, on a coated glass slide. The quality of the spots ~homogeneity, shape! is an important factor in the quality of the analysis. We have built a Linnik-type interference microscope providing three-dimensional ~3D! topographic images of surfaces with a sensitivity of a few tens of picometers in height. This microscope has been applied to obtain 3D images of DNA microarray spots. We demonstrate the ability of the system to reveal the quality of DNA spots.

Deep groundwater flow as the main pathway for chemical outputs in a small headwater watershed (Mule Hole, South India) Jean-Christophe Maréchal et al. This study examined the water and chemical cycles of a small forested head catchment. Application of Cl mass balance lead to the identification of a groundwater baseflow located into the active zone of the crystalline aquifer, below the weir. These findings indicate that groundwater contributes to a large part of chemical outputs at the catchment scale. This study shows that groundwater should not be ignored in most small watershed studies because this flow path and associated solutes are strongly significant for chemical balance assessment.

Caractériser la formation d'étoiles obscurcie à z ~ 2 dans l'Univers Laurie Riguccini Une fraction non négligeable de l'histoire de la formation stellaire a lieu dans des environnements très affectés par la poussière. Il est donc naturel de se demander si on arrive à bien contraindre cette activité de formation d'étoiles. En effet, une part importante de cette activité pourrait être manquée due à la présence de poussière. C'est dans ce contexte que s'inscrit le travail que je vais présenter.Dans la première partie de ma thèse, j'ai eu pour but de déterminer la fraction de galaxies lumineuses formant des étoiles à haut redshift (i.e. 1.5 1mJy) présentent des couleurs 100/24 et 160/24 plus faibles que les autres sources du champ COSMOS et leur luminosité semble donc provenir majoritairement d'un AGN. Les avancées technologiques et l'exploration des longueurs d'ondes en infra-rouge lointain et en submillimétrique, avec notamment Herschel, SCUBA-2, Alma, JWST, permettront de mieux comprendre la connexion AGN/ flambée de formation stellaire au sein des galaxies jusqu'à des hauts redshifts.

Density imaging of volcanos with atmospheric muons F. Fehr Their long range in matter renders high-energy atmospheric muons a unique probe for geophysical explorations, permitting the cartography of density distributions which can reveal spatial and possibly also temporal variations in extended geological structures. A Collaboration between volcanologists and (astro-)particle physicists, Tomuvol, was formed in 2009 to study tomographic muon imaging of volcanos with high-resolution tracking detectors. Here we discuss preparatory work towards muon tomography as well as the first fux measurements taken at the Puy de Dome, an inactive lava dome volcano in the Massif Central.

Découplage de système lent/rapide appliqué en Economie et Econophysique Aurélien Hazan La compréhension des phénomènes économiques nécessite de prendre en compte plusieurs échelles de temps simultanément. Nous étudions le cas d'un modèle simple d'épargne, où plusieurs échelles de temps caractéristiques coexistent. Nous montrons qu'il est possible de séparer les contributions lentes et rapides confondues dans une même variable observée en nous appuyant d'une part sur une linéarisation de la dynamique (stochastique et nonlinéaire) autour d'un point d'équilibre, et d'autre part sur un découplage via la transformation de Chang, issue de la théorie de la commande. Dans un second temps, nous abordons le problème de l'agrégation du comportment d'un grand nombre d'agents, sous l'angle du découplage des composantes lentes/rapides. Nous rendons compte des critiques adressées dans la littérature aux modèles DSGE et examinons la possibilité de découpler un système multi-agents, puis un modèle éconophysique de condensation de la richesse.

Fronts between periodic patterns for bistable recursions on lattices Ricardo Coutinho et al. Bistable space-time discrete systems commonly possess a large variety of stable stationary solutions with periodic profile. In this context, it is natural to ask about the fate of trajectories composed of interfaces between steady configurations with periodic pattern and in particular, to study their propagation as traveling fronts. Here, we investigate such fronts in piecewise affine bistable recursions on the one-dimensional lattice. By introducing a definition inspired by symbolic dynamics, we prove the existence of front solutions and uniqueness of their velocity, upon the existence of their ground patterns. Moreover, the velocity dependence on parameters and the co-existence of several fronts with distinct ground patterns are also described. Finally, robustness of the results to small $C^1$-perturbations of the piecewise affine map is argued by mean continuation arguments.

Probing robustness of cellular automata through variations of asynchronous updating Olivier Bouré et al. Typically viewed as a deterministic model of spatial computing, cellular automata are here considered as a collective system subject to the noise and lack of constraints of natural computing. The updating scheme, classically synchronous, is replaced by stochastic versions which either randomly update cells, or disrupt the cell-to-cell transmission of information. Applying our panel of updating schemes to Elementary Cellular Automata, we observe a wide variety of behaviours, which we study in the scope of macroscopic statistical phenomena and microscopic analysis of the internal mechanisms. Finally we discuss on the role of updating scheme in the emergent behaviours, and the interest in using them to probe models in order to reveal robust properties.

Stratégies de collaboration pour un écrit interactif en FLE dans une communauté virtuelle d'élèves blogueurs Anna Koenig-Wiśniewska Cette thèse relevant du domaine de la Didactique des langues (français langue étrangère) concerne l'apprentissage/enseignement enrichi par les TICE. Elle définit et analyse les stratégies de collaboration pour un écrit interactif, pratiquées par une communauté virtuelle d'élèves blogueurs. Le cadre théorique s'inscrit dans la perspective actionnelle (CECRL), instrumentée des TICE, et il embrasse les réseaux sociaux et leur application dans l'éducation, les caractéristiques de l'écrit interactif en tant que nouveau cybergenre. La méthodologie appliquée est principalement qualitative et relève de la recherche-action et, en particulier, de la recherche-développement appliquée à la conception de ressources multimédia. La construction d'un blogue communautaire a permis le recueil des données pour la partie empirique. Dans cet environnement numérique d'apprentissage, des élèves de 15 à 17 ans de Pologne, Ukraine, Tchéquie et Italie, stimulés par un scénario d'apprentissage, ont réalisé des interactions écrites pendant huit mois. La recherche a montré que les extensions intégrées au blog ont facilité l'émergence d'une communauté d'apprentissage. Elle a également montré comment ces artefacts ont été exploités, en se concentrant notamment sur une analyse des stratégies mobilisées pour l'écrit, ainsi que les caractéristiques de cet écrit interactif produit.

De la reconnaissance des expressions faciales à une perception visuelle partagée : une architecture sensori-motrice pour amorcer un référencement social d'objets, de lieux ou de comportements Sofiane Boucenna Cette thèse se concentre sur les interactions émotionnelles en robotique autonome. Le robot doit pouvoir agir et réagir dans un environnement naturel et faire face à des perturbations imprédictibles. Il est donc nécessaire que le robot puisse acquérir une autonomie comportementale à savoir la capacité d'apprentissage et d'adaptation en ligne. En particulier, nous nous proposons d'étudier quels mécanismes introduire pour que le robot ait la capacité de se constituer une perception des objets de son environnement qui puisse être partagée par celle d'un partenaire humain. Le problème sera de faire apprendre à notre robot à préférer certains objets et à éviter d'autres objets. La solution peut être trouvée en psychologie dans ce que l'on appelle "référencement social" ("social referencing") qui consiste à attribuer une valeur à un objet grâce à l'interaction avec un partenaire humain. Dans ce contexte, notre problème est de trouver comment un robot peut apprendre de manière autonome à reconnaître les expressions faciales d'un partenaire humain pour ensuite les utiliser pour donner une valence aux objets et permettre leur discrimination. Nous nous intéresserons à comprendre comment des interactions émotionnelles avec un partenaire peuvent amorcer des comportements de complexité croissante tel que le référencement social. Notre idée est que le référencement social aussi bien que la reconnaissance d'expressions faciales peut émerger d'une architecture sensori-motrice. Sans connaissance de ce que l'autre est, le robot devrait réussir à apprendre des tâches "sociales" de plus en plus complexes. Nous soutenons l'idée que le référencement social peut être amorcé par une simple cascade d'architectures sensori-motrices qui à la base ne sont pas dédiées aux interactions sociales. Cette thèse traite de plusieurs sujets qui ont comme dénominateur commun l'interaction sociale. Nous proposons tout d'abord une architecture capable d'apprendre à reconnaître de manière autonome des expressions faciales primaires grâce à un jeu d'imitation entre une tête expressive et un expérimentateur. Les interactions avec le dispositif robotique commençeraient par l'apprentissage de 5 expressions faciales prototypiques. Nous proposons ensuite une architecture capable de reproduire des mimiques faciales ainsi que leurs différents niveaux d'intensité. La tête expressive pourra reproduire des expressions secondaires par exemple une joie mêlée de colère. Nous verrons également que la discrimination de visages peut émerger de cette interaction émotionnelle à l'aide d'une rythmicité implicite qui se crée entre l'homme et le robot. Enfin, nous proposerons un modèle sensori-moteur ayant la capacité de réaliser un référencement social. Trois situations ont pu être testées: 1) un bras robotique capable d'attraper et de fuir des objets selon les interactions émotionnelles venant du partenaire humain. 2) un robot mobile capable de rejoindre ou d'éviter certaines zones de son environnement. 3) une tête expressive capable d'orienter son regard dans la même direction que l'humain tout ! en attribuant des valeurs émotionnelles aux objets via l'interaction expressive de l'expérimentateur. Nous montrons ainsi qu'une séquence développementale peut émerger d'une interaction émotionnelle de très bas niveau et que le référencement social peut s'expliquer d'abord à un niveau sensori-moteur sans nécessiter de faire appel à un modèle de théorie de l'esprit.

Dynamic QoS Management and Optimisation in Service-Based Systems Radu Calinescu et al. Service-based systems that are dynamically composed at run time to provide complex, adaptive functionality are currently one of the main development paradigms in software engineering. However, the Quality of Service (QoS) delivered by these systems remains an important concern, and needs to be managed in an equally adaptive and predictable way. To address this need, we introduce a novel, tool-supported framework for the development of adaptive service-based systems called QoSMOS (QoS Management and Optimisation of Service-based systems). QoSMOS can be used to develop service-based systems that achieve their QoS requirements through dynamically adapting to changes in the system state, environment and workload. QoSMOS service-based systems translate high-level QoS requirements specified by their administrators into probabilistic temporal logic formulae, which are then formally and automatically analysed to identify and enforce optimal system configurations. The QoSMOS self-adaptation mechanism can handle reliability- and performance-related QoS requirements, and can be integrated into newly developed solutions or legacy systems. The effectiveness and scalability of the approach are validated using simulations and a set of experiments based on an implementation of an adaptive service-based system for remote medical assistance.

Deep groundwater flow as the main pathway for chemical outputs in a small headwater watershed (Mule Hole, South India) Jean-Christophe Maréchal et al. This study examined the water and chemical cycles of a small forested head catchment. Application of Cl mass balance lead to the identification of a groundwater baseflow located into the active zone of the crystalline aquifer, below the weir. These findings indicate that groundwater contributes to a large part of chemical outputs at the catchment scale. This study shows that groundwater should not be ignored in most small watershed studies because this flow path and associated solutes are strongly significant for chemical balance assessment.

Density imaging of volcanos with atmospheric muons F. Fehr Their long range in matter renders high-energy atmospheric muons a unique probe for geophysical explorations, permitting the cartography of density distributions which can reveal spatial and possibly also temporal variations in extended geological structures. A Collaboration between volcanologists and (astro-)particle physicists, Tomuvol, was formed in 2009 to study tomographic muon imaging of volcanos with high-resolution tracking detectors. Here we discuss preparatory work towards muon tomography as well as the first fux measurements taken at the Puy de Dome, an inactive lava dome volcano in the Massif Central.

Effects of Ca2+ on supramolecular aggregation of natural organic matter in aqueous solutions: A comparison of molecular modeling approaches A. Kalinichev et al. Natural organic matter (NOM) represents a complex molecular system that cannot be fully characterized compositionally or structurally in full atomistic detail. This makes the application of molecular modeling approaches very difficult and significantly hinders quantitative investigation of NOM properties and behavior by these otherwise very efficient computational techniques. Here we report and analyze three molecular dynamics (MD) simulations of Ca2+ complexation with NOM in aqueous solutions in an attempt to quantitatively assess possible effects of model- and system size-dependence in such simulations. Despite some obvious variations in the computed results that depend on the size of the simulated system and on the parameters of the force field models used, all three simulations are quite robust and consistent. They show Ca2+ ions associated with 35-50% of the NOM carboxylic groups at near-neutral pH and point to a strong preference for the stability of bidentate-coordinated contact ion pairs. The degree and potential mechanisms of NOM supramolecular aggregation in the presence of Ca2+ ions in solution are also assessed on a semi-quantitative level from two larger-scale MD simulations.

Making Sense of Variations: Introducing Alternatives in Speech Synthesis Nicolas Obin et al. This paper addresses the use of speech alternatives to enrich speech synthesis systems. Speech alternatives denote the variety of strategies that a speaker can use to pronounce a sentence - depending on pragmatic constraints, speaking style, and specific strategies of the speaker. During the training, symbolic and acoustic characteristics of a unit-selection speech synthesis system are statistically modelled with context-dependent parametric models (GMMs/HMMs). During the synthesis, symbolic and acoustic alternatives are exploited using a GENERALIZED VITERBI ALGORITHM (GVA) to determine the sequence of speech units used for the synthesis. Objective and subjective evaluations supports evidence that the use of speech alternatives significantly improves speech synthesis over conventional speech synthesis systems.

Professionnalisation du métier d'enseignant et informatique à l'école élémentaire : une approche par la théorie des représentations sociales et professionnelles Stéphanie Netto Depuis les années 70, l'introduction de l'informatique et d'Internet dans la formation des enseignants et dans les programmes du Primaire n'a eu de cesse de générer des questionnements, des craintes et des espoirs. Parallèlement, le recrutement et la formation des enseignants du Primaire n'ont aussi cessé d'être débattus.Dans le cadre de cette recherche, nous avons voulu comprendre la professionnalisation du métier d'enseignant en comparant les représentations de l'informatique et du métier d'enseignant au Primaire d'un groupe en formation (futurs professeurs des écoles, PE) et d'un groupe professionnel (enseignants en poste). Nous avons aussi opérationnalisé l'hypothèse de polyphasie cognitive (Moscovici, 1961) en étudiant, pour un même groupe, la représentation de l'informatique dans le contexte de la vie privée et dans le contexte de la vie professionnelle. Les résultats obtenus par questionnaire montrent, tant au niveau structural qu'au niveau du champ structuré, des différences entre les réponses de 175 futurs PE et celles de 245 enseignants sur l'informatique à l'école. Dans le cadre de leur vie privée, il y a beaucoup moins de différences notables entre leurs réponses. Concernant la représentation du métier, les quelques différences repérées sont expliquées à travers l'expérience et le professionnalisme.Cette recherche participe donc à la compréhension des processus de professionnalisation sous la composante représentationnelle. Elle confirme, par ailleurs, la nécessité de préciser le contexte d'évocation dans des recherches sur les représentations sociales.

Éléments d'analyse du marquage différentiel de l'objet dans les langues romanes Alexandru Mardale Cet article examine le marquage différentiel de l'objet (MDO) dans les trois langues romanes qui en font un usage systématique : le roumain, l'espagnol et le sarde. Notre but est de mettre en évidence les similitudes et les différences que revêt ce phénomène dans ces trois langues. À la suite de Laca (1995), Dobrovie-Sorin (1997, 2002), Bleam (2004, 2005), Cornilescu & Dobrovie-Sorin (2007), nous discuterons de façon plus approfondie une généralisation qui rend compte de son existence. Cette généralisation a pour base la corrélation entre le marquage de l'objet et le type de dénotation de ce dernier et elle est exclusivement négative. Plus précisément, nous montrerons que dans les trois langues concernées, le MDO est exclu avec les objets directs ayant une dénotation de type propriété.

Making Sense of Variations: Introducing Alternatives in Speech Synthesis Nicolas Obin et al. This paper addresses the use of speech alternatives to enrich speech synthesis systems. Speech alternatives denote the variety of strategies that a speaker can use to pronounce a sentence - depending on pragmatic constraints, speaking style, and specific strategies of the speaker. During the training, symbolic and acoustic characteristics of a unit-selection speech synthesis system are statistically modelled with context-dependent parametric models (GMMs/HMMs). During the synthesis, symbolic and acoustic alternatives are exploited using a GENERALIZED VITERBI ALGORITHM (GVA) to determine the sequence of speech units used for the synthesis. Objective and subjective evaluations supports evidence that the use of speech alternatives significantly improves speech synthesis over conventional speech synthesis systems.

Modélisation et conception optimale d'un moteur linéaire à induction pour système de traction ferroviaire Jinlin Gong Cette thèse porte sur l'étude des performances du moteur linéaire de référence par la méthode d'analyse éléments finis, et la conception optimale sur un modèle lourd en temps de calcul. La méthode éléments finis est utilisée pour étudier les performances du moteur linéaire de référence, car le modèle analytique d'un moteur linéaire est difficile à construire dû aux effets d'extrémités. Le modèle éléments finis (MEF) 2D permet de prendre en compte l'effet d'extrémité de longueur finie. L'effet d'extrémité de largeur finie est intégré au modèle 2D en faisant varier la conductivité du secondaire et en ajoutant une inductance de tête de bobines. Ensuite, le couplage entre le MEF 3D magnétique et thermique permet de prendre en compte tous les effets d'extrémité et de l'influence de la température. Un banc d'essais est construit pour valider les modélisations éléments finis. La comparaison entre les différents modèles montre l'importance du modèle couplé. L'optimisation directe sur un MEF est très couteuse en temps de calcul. Les stratégies d'intégrer un modèle de substitution au lieu d'un MEF sont étudiées. L'optimisation directe sur un modèle de substitution et l'algorithme Efficient Global Optimisation (EGO) sont comparés. Un algorithme Space Mapping (SM) 3 niveaux est proposé. Les résultats des cas tests montrent que le SM 3 niveaux est plus efficace par rapport à l'algorithme SM 2 niveaux. Une nouvelle stratégie d'optimisation avec un faible budget d'évaluation du MEF est proposée et testée dans le contexte d'une modélisation difficile. La stratégie proposée permet d'évaluer les MEF en parallèle, ainsi permet un gain considérable de réduction du temps d'optimisation.

Localization of extended brain sources from EEG/MEG: The ExSo-MUSIC approach. Gwénaël Birot et al. We propose a new MUSIC-like method, called 2q-ExSo-MUSIC (q≥1). This method is an extension of the 2q-MUSIC (q≥1) approach for solving the EEG/MEG inverse problem, when spatially-extended neocortical sources ("ExSo") are considered. It introduces a novel ExSo-MUSIC principle. The novelty is two-fold: i) the parameterization of the spatial source distribution that leads to an appropriate metric in the context of distributed brain sources and ii) the introduction of an original, efficient and low-cost way of optimizing this metric. In 2q-ExSo-MUSIC, the possible use of higher order statistics (q≥2) offers a better robustness with respect to Gaussian noise of unknown spatial coherence and modeling errors. As a result we reduced the penalizing effects of both the background cerebral activity that can be seen as a Gaussian and spatially correlated noise, and the modeling errors induced by the non-exact resolution of the forward problem. Computer results on simulated EEG signals obtained with physiologically-relevant models of both the sources and the volume conductor show a highly increased performance of our 2q-ExSo-MUSIC method as compared to the classical 2q-MUSIC algorithms.

Predicting drug side-effect profiles: a chemical fragment-based approach. Edouard Pauwels et al. BACKGROUND: Drug side-effects, or adverse drug reactions, have become a major public health concern. It is one of the main causes of failure in the process of drug development, and of drug withdrawal once they have reached the market. Therefore, in silico prediction of potential side-effects early in the drug discovery process, before reaching the clinical stages, is of great interest to improve this long and expensive process and to provide new efficient and safe therapies for patients. RESULTS: In the present work, we propose a new method to predict potential side-effects of drug candidate molecules based on their chemical structures, applicable on large molecular databanks. A unique feature of the proposed method is its ability to extract correlated sets of chemical substructures (or chemical fragments) and side-effects. This is made possible using sparse canonical correlation analysis (SCCA). In the results, we show the usefulness of the proposed method by predicting 1385 side-effects in the SIDER database from the chemical structures of 888 approved drugs. These predictions are performed with simultaneous extraction of correlated ensembles formed by a set of chemical substructures shared by drugs that are likely to have a set of side-effects. We also conduct a comprehensive side-effect prediction for many uncharacterized drug molecules stored in DrugBank, and were able to confirm interesting predictions using independent source of information. CONCLUSIONS: The proposed method is expected to be useful in various stages of the drug development process.

The CRE1 carbon catabolite repressor of the fungus Trichoderma reesei: a master regulator of carbon assimilation. Thomas Portnoy et al. BACKGROUND: The identification and characterization of the transcriptional regulatory networks governing the physiology and adaptation of microbial cells is a key step in understanding their behaviour. One such wide-domain regulatory circuit, essential to all cells, is carbon catabolite repression (CCR): it allows the cell to prefer some carbon sources, whose assimilation is of high nutritional value, over less profitable ones. In lower multicellular fungi, the C2H2 zinc finger CreA/CRE1 protein has been shown to act as the transcriptional repressor in this process. However, the complete list of its gene targets is not known. RESULTS: Here, we deciphered the CRE1 regulatory range in the model cellulose and hemicellulose-degrading fungus Trichoderma reesei (anamorph of Hypocrea jecorina) by profiling transcription in a wild-type and a delta-cre1 mutant strain on glucose at constant growth rates known to repress and de-repress CCR-affected genes. Analysis of genome-wide microarrays reveals 2.8% of transcripts whose expression was regulated in at least one of the four experimental conditions: 47.3% of which were repressed by CRE1, whereas 29.0% were actually induced by CRE1, and 17.2% only affected by the growth rate but CRE1 independent. Among CRE1 repressed transcripts, genes encoding unknown proteins and transport proteins were overrepresented. In addition, we found CRE1-repression of nitrogenous substances uptake, components of chromatin remodeling and the transcriptional mediator complex, as well as developmental processes. CONCLUSIONS: Our study provides the first global insight into the molecular physiological response of a multicellular fungus to carbon catabolite regulation and identifies several not yet known targets in a growth-controlled environment.

Ligand scaffold hopping combining 3D maximal substructure search and molecular similarity. Flavien Quintus et al. BACKGROUND: Virtual screening methods are now well established as effective to identify hit and lead candidates and are fully integrated in most drug discovery programs. Ligand-based approaches make use of physico-chemical, structural and energetics properties of known active compounds to search large chemical libraries for related and novel chemotypes. While 2D-similarity search tools are known to be fast and efficient, the use of 3D-similarity search methods can be very valuable to many research projects as integration of "3D knowledge" can facilitate the identification of not only related molecules but also of chemicals possessing distant scaffolds as compared to the query and therefore be more inclined to scaffolds hopping. To date, very few methods performing this task are easily available to the scientific community. RESULTS: We introduce a new approach (LigCSRre) to the 3D ligand similarity search of drug candidates. It combines a 3D maximum common substructure search algorithm independent on atom order with a tunable description of atomic compatibilities to prune the search and increase its physico-chemical relevance. We show, on 47 experimentally validated active compounds across five protein targets having different specificities, that for single compound search, the approach is able to recover on average 52% of the co-actives in the top 1% of the ranked list which is better than gold standards of the field. Moreover, the combination of several runs on a single protein target using different query active compounds shows a remarkable improvement in enrichment. Such Results demonstrate LigCSRre as a valuable tool for ligand-based screening. CONCLUSION: LigCSRre constitutes a new efficient and generic approach to the 3D similarity screening of small compounds, whose flexible design opens the door to many enhancements. The program is freely available to the academics for non-profit research at: http://bioserv.rpbs.univ-paris-diderot.fr/LigCSRre.html.

Making Sense of Variations: Introducing Alternatives in Speech Synthesis Nicolas Obin et al. This paper addresses the use of speech alternatives to enrich speech synthesis systems. Speech alternatives denote the variety of strategies that a speaker can use to pronounce a sentence - depending on pragmatic constraints, speaking style, and specific strategies of the speaker. During the training, symbolic and acoustic characteristics of a unit-selection speech synthesis system are statistically modelled with context-dependent parametric models (GMMs/HMMs). During the synthesis, symbolic and acoustic alternatives are exploited using a GENERALIZED VITERBI ALGORITHM (GVA) to determine the sequence of speech units used for the synthesis. Objective and subjective evaluations supports evidence that the use of speech alternatives significantly improves speech synthesis over conventional speech synthesis systems.

A generalization of Kaplan-Meier estimator for analyzing bivariate mortality under right-censoring and left-truncation with applications to model-checking for survival copula models Olivier Lopez In this paper, we provide a new nonparametric estimator of the joint distribution of two lifetimes under random right censoring and left truncation, which can be seen as a bivariate extension of the Kaplan-Meier estimator. We derive asymptotic results for this estimator, including uniform $n^{1/2}-$consistency, and develop a general methodology to study bivariate lifetime modelling, which is a critical issue in the study of pensions with a reversion condition. Application to goodness-of-fit for survival copula models is discussed. We show that the procedure that we use are consistent, and propose a bootstrap procedure based on our estimator to compute the critical values. All the new techniques that we propose are experimented on the Canadian data-set initially studied by Frees et al. (1996).

Minimax risks for sparse regressions: Ultra-high-dimensional phenomenons Nicolas Verzelen Consider the standard Gaussian linear regression model $Y=X\theta+\epsilon$, where $Y\in R^n$ is a response vector and $ X\in R^{n*p}$ is a design matrix. Numerous work have been devoted to building efficient estimators of $\theta$ when $p$ is much larger than $n$. In such a situation, a classical approach amounts to assume that $\theta_0$ is approximately sparse. This paper studies the minimax risks of estimation and testing over classes of $k$-sparse vectors $\theta$. These bounds shed light on the limitations due to high-dimensionality. The results encompass the problem of prediction (estimation of $X\theta$), the inverse problem (estimation of $\theta_0$) and linear testing (testing $X\theta=0$). Interestingly, an elbow effect occurs when the number of variables $k\log(p/k)$ becomes large compared to $n$. Indeed, the minimax risks and hypothesis separation distances blow up in this ultra-high dimensional setting. We also prove that even dimension reduction techniques cannot provide satisfying results in an ultra-high dimensional setting. Moreover, we compute the minimax risks when the variance of the noise is unknown. The knowledge of this variance is shown to play a significant role in the optimal rates of estimation and testing. All these minimax bounds provide a characterization of statistical problems that are so difficult so that no procedure can provide satisfying results.