La réutilisation des eaux usées traitées (REUT) constitue une ressource alternative pérenne et à fort potentiel notamment pour les régions soumises à un stress hydrique. L’objectif général de cette thèse est de définir les conditions nécessaires pour garantir la fiabilité d’une filière de traitement tertiaire pour la réutilisation. La première partie s’intéresse aux limites des méthodes de dénombrement des microorganismes indicateurs couramment employées, notamment en lien avec la contamination particulaire des effluents. Les résultats indiquent qu’il existe un risque de sous-estimation par ces méthodes pour de fortes charges en matières en suspension.La désinfection par rayonnement ultraviolet (UV) est souvent employée pour garantir la qualité sanitaire de l’eau distribuée. La deuxième partie de cette thèse vise à comprendre les mécanismes pouvant affecter l’efficacité des UV. Une corrélation linéaire a pu être mise en évidence entre la contamination particulaire des effluents à traiter et la baisse d’efficacité de la désinfection UV, caractérisée d’une part par la diminution de la constante d’inactivation des microorganismes exposés aux UV et d’autre part par le risque d’apparition d’un effet de traîne (fraction de microorganismes insensibles aux UV).La mise en place d’un prétraitement par filtration s’avère donc nécessaire en amont de la désinfection UV. La troisième partie de cette thèse a permis de définir les paramètres opératoires les plus adaptés à l’effluent. Enfin, le contrôle de la qualité sur l’ensemble de la filière est envisagé, avec notamment la démonstration de faisabilité d’un système de surveillance en continu basé sur la spectrométrie UV/Visible. / Wastewater reclamation is an ongoing and promising alternative resource in a context of water stress. In particular, this is a major issue in the Mediterranean region, and its effects tend to be intensified by global warming. The main objective of this PhD thesis is to define the conditions for ensuring the reliability of a tertiary treatment chain for wastewater reclamation. The first part deals with the limits of the methods used in routine for the enumeration of indicator microorganisms, in relation with particulate contamination. The results indicate that there is a risk of underestimation by these methods for high loads of suspended matter.Ultraviolet (UV) disinfection has been used for years to ensure the biological safety of reclaimed water. The second part of this work aims to understand the mechanisms which may affect the efficiency of UV disinfection. A linear relationship has been observed between the particulate contamination of the effluents to be treated and the disinfection efficiency loss, featured on one hand by the decrease of the inactivation constant of the microorganisms and on the other hand by the risk of tailing (UV-resistant fraction among the microorganisms).The implementation of a pre-filtration is thus necessary before UV disinfection. The third part of this work enabled to identify the filtration parameters suited for the effluents to be treated. Finally, quality control on the whole chain is considered, in particular with the demonstration of feasibility of a continuous control system based on UV/Visible spectrometry.

Made available in DSpace on 2014-06-11T19:27:22Z (GMT). No. of bitstreams: 0 Previous issue date: 2009-07-28Bitstream added on 2014-06-13T19:55:57Z : No. of bitstreams: 1 stropa_kc_me_jabo.pdf: 1481358 bytes, checksum: c093689dc1f549339c3a1e758d9ae052 (MD5) / Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) / O teor protéico da semente de soja pode chegar a 42%, o que demanda uma alta quantidade de nitrogênio. Bradyrhizobium elkanii e Bradyrhizobium japonicum são bactérias fixadoras de nitrogênio que estabelecem uma simbiose com a soja convertendo o nitrogênio atmosférico em amônia, que é o composto assimilável pela planta. A maximização desta simbiose em termos de produtividade são alcançados por meio da inoculação com estirpes de bradirrizóbios, através de inoculantes comerciais. O objetivo deste trabalho consistiu em enumerar células bacterianas de inoculantes de 1, 2 e 4 anos a partir da data de fabricação e avaliar a sobrevivência das células em sementes de soja em 4, 24 e 48 h após inoculação. Os resultados de contagem das unidades formadoras de colônias (UFC) foram confrontados com a técnica de quantificação absoluta por PCR (reação em cadeia da polimerase) em tempo real (qaPCR). Os números foram coerentes em ambas as técnicas em relação à proporção nos tempos de inoculação e nos inoculantes em análise, porém a qaPCR apresentou melhor acurária e rapidez nos resultados, detectando as células incultiváveis. As células resistiram sob dessecação até t=24 h, com queda considerável em todos os inoculantes após 48 h sob dessecação. Provavelmente esta queda é resultado de alteração bioquímica e fisiológica de seu metabolismo, dispondo de mecanismos defensivos às condições adversas para sua sobrevivência. A enumeração por qaPCR pode ser usada como prova, em casos de variação encontrada na quantificação por diferentes laboratórios, considerando que o método por UFC apresenta muitas variáveis e não incluem células viáveis não cultiváveis (VBNC). O fato do estado fisiológico dos rizóbios em inoculantes ser bastante variável ao longo do período de armazenamento, a enumeração celular seja pelo método de plaqueamento... / Soybean grains contain up 42% of protein, so this crop requires high amounts of nitrogen for plant development. Bradyrhizobium elkanii and Bradyrhizobium japonicum are nitrogen fixing bacteria that establish symbiosis with soybean, converting the atmospheric nitrogen into ammonia that is the assimilable inorganic nitrogen compound for the plant. The optimization of this symbiosis and the success of biological nitrogen fixation (BNF) are reached by inoculating the seeds with strains of bradyrhizobia, using commercial inoculants. The aim of this work consisted in the enumeration of bacterial cells of inoculants 1, 2, and 4 years old counting from the date of manufacture and in the evaluation of surviving cells under desiccation in soybean seeds after 4h, 24h, and 48 h. The results of counting of the colony forming colonies (CFU) were correlated with those of obtained using absolute quantification by PCR (realtime PCR qaPCR). The values were coherent in both the techniques in relation to the ratio in times of inoculation and the inoculants properly. However, qaPCR was quicker and more occurat; and also allowed detection of the non-culturable cells. The cells resisted under desiccation until t=24 h, with considerable fall in all the inoculants after 48h under desiccation. This was probably due to by biochemical and physiological changes in its metabolism, making use of defensive mechanisms to the adverse conditions for its survival. qaPCR enumeration could be used as a proof when variation in cell counting by different laboratories occurs. Considering that CFU based method present a lot of variables and do not account live cells in viable but non culturable (VBNC). The fact of the physiological state in rhizobia inoculants be sufficiently changeable throughout the storage period, the cellular enumeration for both methods do not reveal real state of the biological system in question... (Complete abstract click electronic access below)

Resumo: O teor protéico da semente de soja pode chegar a 42%, o que demanda uma alta quantidade de nitrogênio. Bradyrhizobium elkanii e Bradyrhizobium japonicum são bactérias fixadoras de nitrogênio que estabelecem uma simbiose com a soja convertendo o nitrogênio atmosférico em amônia, que é o composto assimilável pela planta. A maximização desta simbiose em termos de produtividade são alcançados por meio da inoculação com estirpes de bradirrizóbios, através de inoculantes comerciais. O objetivo deste trabalho consistiu em enumerar células bacterianas de inoculantes de 1, 2 e 4 anos a partir da data de fabricação e avaliar a sobrevivência das células em sementes de soja em 4, 24 e 48 h após inoculação. Os resultados de contagem das unidades formadoras de colônias (UFC) foram confrontados com a técnica de quantificação absoluta por PCR (reação em cadeia da polimerase) em tempo real (qaPCR). Os números foram coerentes em ambas as técnicas em relação à proporção nos tempos de inoculação e nos inoculantes em análise, porém a qaPCR apresentou melhor acurária e rapidez nos resultados, detectando as células incultiváveis. As células resistiram sob dessecação até t=24 h, com queda considerável em todos os inoculantes após 48 h sob dessecação. Provavelmente esta queda é resultado de alteração bioquímica e fisiológica de seu metabolismo, dispondo de mecanismos defensivos às condições adversas para sua sobrevivência. A enumeração por qaPCR pode ser usada como prova, em casos de variação encontrada na quantificação por diferentes laboratórios, considerando que o método por UFC apresenta muitas variáveis e não incluem células viáveis não cultiváveis (VBNC). O fato do estado fisiológico dos rizóbios em inoculantes ser bastante variável ao longo do período de armazenamento, a enumeração celular seja pelo método de plaqueamento... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Soybean grains contain up 42% of protein, so this crop requires high amounts of nitrogen for plant development. Bradyrhizobium elkanii and Bradyrhizobium japonicum are nitrogen fixing bacteria that establish symbiosis with soybean, converting the atmospheric nitrogen into ammonia that is the assimilable inorganic nitrogen compound for the plant. The optimization of this symbiosis and the success of biological nitrogen fixation (BNF) are reached by inoculating the seeds with strains of bradyrhizobia, using commercial inoculants. The aim of this work consisted in the enumeration of bacterial cells of inoculants 1, 2, and 4 years old counting from the date of manufacture and in the evaluation of surviving cells under desiccation in soybean seeds after 4h, 24h, and 48 h. The results of counting of the colony forming colonies (CFU) were correlated with those of obtained using absolute quantification by PCR (realtime PCR qaPCR). The values were coherent in both the techniques in relation to the ratio in times of inoculation and the inoculants properly. However, qaPCR was quicker and more occurat; and also allowed detection of the non-culturable cells. The cells resisted under desiccation until t=24 h, with considerable fall in all the inoculants after 48h under desiccation. This was probably due to by biochemical and physiological changes in its metabolism, making use of defensive mechanisms to the adverse conditions for its survival. qaPCR enumeration could be used as a proof when variation in cell counting by different laboratories occurs. Considering that CFU based method present a lot of variables and do not account live cells in viable but non culturable (VBNC). The fact of the physiological state in rhizobia inoculants be sufficiently changeable throughout the storage period, the cellular enumeration for both methods do not reveal real state of the biological system in question... (Complete abstract click electronic access below) / Orientador: Jackson Antônio Marcondes de Souza / Coorientadora: Eliana Gertrudes Macedo Lemos / Banca: Manoel Victor Franco Lemos / Banca: Emanuel Maltempi de Souza / Mestre

Esta dissertação introduz o conceito de cortes KL, o que permite controlar tanto o número K de entradas como o número L de saídas em uma região de um circuito. O projeto de um circuito digital pode ser dividido em duas fases: síntese lógica e síntese física. Dentro de síntese lógica, um dos principais passos é o mapeamento tecnológico. Tradicionalmente, o processo de mapeamento tecnológico somente lida com funções de saída única, para a construção de circuitos. O objetivo deste método é explorar o uso de blocos de múltiplas saídas no mapeamento tecnológico. Para prover escalabilidade, o conceito de fatoração de cortes é estendido para os cortes KL. Algoritmos para enumerar esses cortes e também para enumerar alguns subconjuntos de cortes com características específicas são apresentados e os resultados são mostrados. Como exemplos de aplicações práticas, diferentes algoritmos de cobertura são propostos. O algoritmo guloso é uma alternativa simples e produz bons resultados em área, mas é muito restritivo, pois não é factível em mapeamento orientado à atraso. Outro algoritmo de cobertura apresentado é uma extensão do algoritmo de fluxo de área e permite a utilização de cortes com várias saídas, mantendo possível a consideração de outros custos. Um algoritmo de correspondência Booleana que é capaz de lidar com blocos com múltiplas saídas também é descrito. Isso permite a utilização de uma biblioteca padrão com células com mais de uma saída no mapeamento tecnológico. Os resultados mostram a viabilidade e utilidade do método. / This thesis introduces the concept of KL-feasible cuts, which allows controlling both the number K of inputs and the number L of outputs in a circuit region. The design of a digital circuit can roughly be divided in two phases: logic synthesis and physical synthesis. Within logic synthesis, one of the main steps is the technology mapping. Traditionally, the technology mapping process only handles single output functions, in order to construct circuits. The objective of this method is to explore the use of multiple output blocks on technology mapping. To provide scalability, the concept of factor cuts is extended to KL-cuts. Algorithms for enumerating these cuts and also for enumerating some subsets of cuts with some special characteristics are presented and results are shown. As examples of practical applications, different covering algorithms are proposed. The greedy algorithm is a simple alternative and produces good results in area, but it is too restrictive, as it is not practical in timing oriented mapping. The other covering algorithm presented is an extension to the area flow algorithm and allows cuts with multiple outputs to be used while making possible the control of some other costs. A Boolean matching algorithm that is able to handle multiple output blocks is also described, which permits the use of a standard cell library with more than one output on technology mapping. The results show the viability and usefulness of the method.

Orientador: José Plínio de Oliveira Santos / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Matemática Estatística e Computação Científica / Made available in DSpace on 2018-08-25T04:32:08Z (GMT). No. of bitstreams: 1 BrasilJunior_NelsonGomes_M.pdf: 980636 bytes, checksum: dd8d61baeb633d5f598abc3523def800 (MD5) Previous issue date: 2014 / Resumo: Neste trabalho, estudamos a sequência dos Números de Catalan, uma sequência que aparece como solução de vários problemas de contagem envolvendo árvores, palavras, grafos e outras estruturas combinatórias. Atualmente, são conhecidas cerca de 200 interpretações combinatórias distintas para os Números de Catalan, o que motiva o estudo de relações entre estas interpretações, isto é, entre conjuntos cuja cardinalidade é dada pelos termos desta sequência. O principal objetivo do nosso trabalho é, portanto, mostrar bijeções entre esses conjuntos. No início do texto fazemos uma pequena introdução histórica aos números de Catalan, assim como definimos algumas formas de representar a sequência estudada. Depois mostramos algumas bijeções clássicas entre conjuntos contados pela sequência de Catalan. Além disso, apresentamos outras bijeções entre conjuntos envolvendo diversos objetos combinatórios. No total, são exibidas 29 bijeções / Abstract: In this work, we study the sequence of Catalan Numbers, which appears as a solution of many counting problems involving trees, words, graphs and other combinatorial structures. Nowadays, about 200 different combinatorial interpretations of the Catalan Numbers are known and that motivates the study between them, i. e., the study between sets whose cardinality is given by the terms of this sequence. The main objective of our work is therefore to show bijections between these sets. In the beginning, we make a short historical introduction of the Catalan Numbers and define some ways to represent the sequence. After that, we show some classical bijections between sets counted by the Catalan Numbers. Additionally, we exhibit other bijections between sets involving several combinatorial objects. Altogether, 29 bijections are presented / Mestrado / Matematica Aplicada / Mestre em Matemática Aplicada

Dans cette thèse, nous abordons le problème de l'identification et de la quantification de variants (épissage alternatif et polymorphisme génomique) dans des données de RNA-seq sans génome de référence, et sans faire un assemblage complet des transcripts. Basé sur l'idée que chaque variant correspond à un motif reconnaissable, qu'on appelle une bulle, dans un graphe de Bruijn construit à partir des lectures de RNA-seq, nous proposons un modèle pour les variants dans de tels graphes. Nous introduisons ensuite une méthode, appelé KisSplice, pour extraire les événements d'épissage alternatif, et nous montrons qu'il trouve plus d'événements corrects que les assembleurs de transcriptome traditionnels. Afin d'améliorer son temps d'exécution, nous proposons un nouvel algorithme polynomial pour énumérer les bulles. On montre qu'il est plusieurs ordres de grandeur plus rapide que les approches précédentes. Afin de réduire sa consommation en mémoire, nous proposons une nouvelle façon de représenter un graphe de Bruijn. Nous montrons que notre approche utilise 30% à 40% moins de mémoire que l'état de l'art. Nous appliquons les techniques développées pour énumérer les bulles à deux problémes classiques. Nous donnons le premier algorithme optimal pour énumérer les cycles dans des graphes non orientés. Il s'agit de la première amélioration à ce probléme en près de 40 ans. Nous considérons ensuite une variante du problème des K chemins plus courts: au lieu de limiter le nombre des chemins, nous limitons leurs poids. Nous présentons de nouveaux algorithmes qui utilisent exponentiellement moins mémoire que les approches précédentes / In this thesis, we address the problem of identifying and quantifying variants (alternative splicing and genomic polymorphism) in RNA-seq data when no reference genome is available, without assembling the full transcripts. Based on the idea that each variant corresponds to a recognizable pattern, a bubble, in a de Bruijn graph constructed from the RNA-seq reads, we propose a general model for all variants in such graphs. We then introduce an exact method, called KisSplice, to extract alternative splicing events and show that it outperforms general purpose transcriptome assemblers. We put an extra effort to make KisSplice as scalable as possible. In order to improve the running time, we propose a new polynomial delay algorithm to enumerate bubbles. We show that it is several orders of magnitude faster than previous approaches. In order to reduce its memory consumption, we propose a new compact way to build and represent a de Bruijn graph. We show that our approach uses 30% to 40% less memory than the state of the art, with an insignificant impact on the construction time. Additionally, we apply the techniques developed to list bubbles in two classical problems: cycle enumeration and the K-shortest paths problem. We give the first optimal algorithm to list cycles in undirected graphs, improving over Johnson’s algorithm. This is the first improvement to this problem in almost 40 years. We then consider a different parameterization of the K-shortest (simple) paths problem: instead of bounding the number of st-paths, we bound the weight of the st-paths. We present new algorithms using exponentially less memory than previous approaches

The thesis considers a representation of a discrete multidimensional probability distribution using an apparatus of compositional models, and focuses on the theoretical background and structure of search space for structure learning algorithms in the framework of such models and particularly focuses on the subclass of decomposable models. Based on the theoretical results, proposals of basic learning techniques are introduced and compared.

Orientador: Fernando José Von Zuben / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Elétrica e de Computação / Made available in DSpace on 2018-08-27T03:28:44Z (GMT). No. of bitstreams: 1 Oliveira_SaulloHaniellGalvaode_M.pdf: 1171322 bytes, checksum: 5488cfc9b843dbab6d7a5745af1e3d4b (MD5) Previous issue date: 2015 / Resumo: Biclusterização envolve a clusterização simultânea de objetos e seus atributos, definindo mo- delos locais de relacionamento entre os objetos e seus atributos. Assim como a clusterização, a biclusterização tem uma vasta gama de aplicações, desde suporte a sistemas de recomendação, até análise de dados de expressão gênica. Inicialmente, diversas heurísticas foram propostas para encontrar biclusters numa base de dados numérica. No entanto, tais heurísticas apresen- tam alguns inconvenientes, como não encontrar biclusters relevantes na base de dados e não maximizar o volume dos biclusters encontrados. Algoritmos enumerativos são uma proposta recente, especialmente no caso de bases numéricas, cuja solução é um conjunto de biclusters maximais e não redundantes. Contudo, a habilidade de enumerar biclusters trouxe mais um cenário desafiador: em bases de dados ruidosas, cada bicluster original se fragmenta em vá- rios outros biclusters com alto nível de sobreposição, o que impede uma análise direta dos resultados obtidos. Essa fragmentação irá ocorrer independente da definição escolhida de co- erência interna no bicluster, sendo mais relacionada com o próprio nível de ruído. Buscando reverter essa fragmentação, nesse trabalho propomos duas formas de agregação de biclusters a partir de resultados que apresentem alto grau de sobreposição: uma baseada na clusteriza- ção hierárquica com single linkage, e outra explorando diretamente a taxa de sobreposição dos biclusters. Em seguida, um passo de poda é executado para remover objetos ou atributos indesejados que podem ter sido incluídos como resultado da agregação. As duas propostas foram comparadas entre si e com o estado da arte, em diversos experimentos, incluindo bases de dados artificiais e reais. Essas duas novas formas de agregação não só reduziram significa- tivamente a quantidade de biclusters, essencialmente defragmentando os biclusters originais, mas também aumentaram consistentemente a qualidade da solução, medida em termos de precisão e recuperação, quando os biclusters são conhecidos previamente / Abstract: Biclustering involves the simultaneous clustering of objects and their attributes, thus defin- ing local models for the two-way relationship of objects and attributes. Just like clustering, biclustering has a broad set of applications, ranging from an advanced support for recom- mender systems of practical relevance to a decisive role in data mining techniques devoted to gene expression data analysis. Initially, heuristics have been proposed to find biclusters, and their main drawbacks are the possibility of losing some existing biclusters and the inca- pability of maximizing the volume of the obtained biclusters. Recently efficient algorithms were conceived to enumerate all the biclusters, particularly in numerical datasets, so that they compose a complete set of maximal and non-redundant biclusters. However, the ability to enumerate biclusters revealed a challenging scenario: in noisy datasets, each true bicluster becomes highly fragmented and with a high degree of overlapping, thus preventing a direct analysis of the obtained results. Fragmentation will happen no matter the boundary condi- tion adopted to specify the internal coherence of the valid biclusters, though the degree of fragmentation will be associated with the noise level. Aiming at reverting the fragmentation, we propose here two approaches for properly aggregating a set of biclusters exhibiting a high degree of overlapping: one based on single linkage and the other directly exploring the rate of overlapping. A pruning step is then employed to filter intruder objects and/or attributes that were added as a side effect of aggregation. Both proposals were compared with each other and also with the actual state-of-the-art in several experiments, including real and artificial datasets. The two newly-conceived aggregation mechanisms not only significantly reduced the number of biclusters, essentially defragmenting true biclusters, but also consistently in- creased the quality of the whole solution, measured in terms of Precision and Recall when the composition of the dataset is known a priori / Mestrado / Engenharia de Computação / Mestre em Engenharia Elétrica

La quantification des infirmités en France au XIXe siècle est guidée par deux objectifs principaux. Il s'agit, d'une part, d'établir la répartition géographique de ces infirmités afin d'en évaluer les causes et d'ouvrir des structures d'accueil adaptées (asiles, institutions d'éducation). D'autre part, cela consiste à mesurer les évolutions du nombre des infirmes. Cette thèse étudie ces pratiques de quantification à partir de deux sources : les comptes rendus sur le recrutement de l'armée (publiés à partir de 1818) et qui ne traitent que de générations de jeunes hommes, et les recensements quinquennaux (de 1851 à 1876) qui concernent l'ensemble de la population. Cette thèse examine la construction du chiffre d'infirmes, de sa mise en place à sa remise en cause. Différents éléments sont articulés dans l'analyse. Tout d'abord, nous étudions la place des infirmes dans la société et le rôle de l'État dans leur prise en charge. Nous nous intéressons aussi aux savants (réunis dans des sociétés ou en congrès pour partager leurs recherches) et aux instances de gouvernance (maires, préfets, ministères). Les premiers étudient la population pour des raisons scientifiques, les seconds pour des motifs administratifs. Cette mise en parallèle d'intérêts divergents fait pourtant ressortir une inquiétude commune: celle d'observer une dégénérescence de la population. Des statisticiens et médecins s'attèlent donc à analyser ces nouvelles productions statistiques pour démontrer l'exagération de ces craintes. / The enumeration of the infirm in France in the 19th century had several objectives: to map their geographical distribution, in order to better understand the causes of these disabilities and to provide suitable facilities (asylums, educational institutions), but also to track trends in view to potentially control and decrease the size of this group. This study addresses the quantification of disability through two data sources: the five-year censuses (1851-1876) and the reports of army recruitment (published from 1818 onwards).We study how these figures were constructed, from their initial implementation until they started being challenged. To do so, we also explore the place of the disabled in society, and the role of the State in their care. We deal with the interactions between scientists (gathered in learned societies or in conferences to share their research) and various governance bodies (mayors, prefects, ministries). The former studied this population for scientific reasons, while the latter for administrative purposes. Comparing these divergent interests highlights however a common anxiety: that they were observing a degeneration of the population. Because of the regularity of data outputs, time trends, which until then had not been available, could now be constructed. Statisticians and doctors could therefore use these new statistical figures to demonstrate the exaggeration of these fears.

Cette thèse s'inscrit dans le cadre de la biologie des systèmes et porte plus particulièrement sur un problème relatif aux réseaux biologiques hétérogènes. Elle se concentre sur les relations entre le métabolisme et le contexte génomique, en utilisant une approche de fouille de graphes.Il est communément admis que des étapes enzymatiques successives impliquant des produits de gènes situés à proximité sur le chromosome traduisent un avantage évolutif du maintien de cette relation de voisinage au niveau métabolique ainsi que génomique. En conséquence, nous choisissons de nous concentrer sur la détection de réactions voisines catalysées par des produits de gènes voisins, où la notion de voisinage peut être modulée en autorisant que certaines réactions et/ou gènes soient omis. Plus spécifiquement, les motifs recherchés sont des trails de réactions (c'est-à-dire des séquences de réactions pouvant répéter des réactions, mais pas les liens entre elles) catalysées par des produits de gènes voisins. De tels motifs de voisinage sont appelés des motifs métaboliques et génomiques.De plus, on s'intéresse aux motifs de voisinage métabolique et génomique conservés, c'est-à-dire à des motifs similaires pour plusieurs espèces. Parmi les variations considérées pour un motif conservé, on considère l'absence/présence de réactions et/ou de gènes, ou leur ordre différent.Dans un premier temps, nous proposons des algorithmes et des méthodes afin d'identifier des motifs de voisinage métabolique et génomique conservés. Ces méthodes sont implémentées dans le pipeline libre CoMetGeNe (COnserved METabolic and GEnomic NEighborhoods). À l'aide de CoMetGeNe, on analyse une sélection de 50 espèces bactériennes, en utilisant des données issues de la base de connaissances KEGG.Dans un second temps, un développement de la détection de motifs conservés est exploré en prenant en compte la similarité chimique entre réactions. Il permet de mettre en évidence une classe de modules métaboliques conservés, caractérisée par le voisinage des gènes intervenants. / This thesis fits within the field of systems biology and addresses a problem related to heterogeneous biological networks. It focuses on the relationship between metabolism and genomic context through a graph mining approach.It is well-known that succeeding enzymatic steps involving products of genes in close proximity on the chromosome translate an evolutionary advantage in maintaining this neighborhood relationship at both the metabolic and genomic levels. We therefore choose to focus on the detection of neighboring reactions being catalyzed by products of neighboring genes, where the notion of neighborhood may be modulated by allowing the omission of several reactions and/or genes. More specifically, the sought motifs are trails of reactions (meaning reaction sequences in which reactions may be repeated, but not the links between them). Such neighborhood motifs are referred to as metabolic and genomic patterns.In addition, we are also interested in detecting conserved metabolic and genomic patterns, meaning similar patterns across multiple species. Among the possible variations for a conserved pattern, the presence/absence of reactions and/or genes may be considered, or the different order of reactions and/or genes.A first development proposes algorithms and methods for the identification of conserved metabolic and genomic patterns. These methods are implemented in an open-source pipeline called CoMetGeNe (COnserved METabolic and GEnomic NEighborhoods). By means of this pipeline, we analyze a data set of 50 bacterial species, using data extracted from the KEGG knowledge base.A second development explores the detection of conserved patterns by taking into account the chemical similarity between reactions. This allows for the detection of a class of conserved metabolic modules in which neighboring genes are involved.