Thèse (M.Sc.)--Université Laval, 1998. / Bibliogr.: f. 164-177. Publié aussi en version électronique.

Mestre -- Universidade de Sao Paulo. Faculdade de Saude Publica. Departamento de Pratica de Saude Publica, Sao Paulo, 2000.

Thesis (M.A.)--Texas State University-San Marcos, 2007. / Vita. Appendices: leaves 77-90. Includes bibliographical references (leaves 91-94).

Thesis (M.A.)--Texas State University-San Marcos, 2007. / Vita. Appendices: leaves 77-90. Includes bibliographical references (leaves 91-94).

Texte remanié de: Th.--Droit--Louvain--Université Catholique, 1996. / Notes bibliogr. Bibliogr. p. 519-551.

Th.--Inform.--Nancy--I.N.P.L., 1985.

Texte remanié et abrégé de: Diss.--Fachbereich für Philosophie, Wirtschafts- und Sozialwissenschaften--Technische Universität Carolo-Wilhelmina zu Braunschweig, Wintersemester 1994/95. / Titre de couv. : "Wilhelmshaven-Rüstringen im Schatten der Reichsmarine" Bibliogr. p. 320-349. Index.

Recent experimental investigation on FSS arrays forming waveguides (FSGs) and horns showed that incident electromagnetic energy can be guided and radiated at specific frequencies. This thesis aims to provide the theoretical understanding of the waves propagating inside a cylindrical FSS waveguide. With immediate applications on horn antennas the research deals with cylindrical guides, made entirely from double periodic arrays. The theoretical analysis begins as a standard electromagnetic boundary value problem. The formulated system of algebraic equations is solved either for the complex propagation constant, by an iterative procedure or, for the fields. The analysis makes use of the Floquet modal expansion, the current representation as a set of sub-domain basis functions and the Method of Moments. Initially, the thesis is concerned with single periodic structures, which is a special case to the analysis. The efficiency of the model to provide stable and valid results is examined. Next the elements are finite dipoles. The effects of the dipole resonance to the propagating and radiating characteristics of the FSS is closely investigated. Other aspects include the effects of the periodicity and the element size. The investigation concludes with an FSG with square loop elements. Validation of the results for some designs is made by comparison with measured data.

Neste trabalho, a docagem e a dinâmica molecular foram utilizadas para investigar a interação entre oligonucleotídeos e duas bases de Tröger: uma base simétrica - contendo duas proflavinas – e uma base assimétrica – contendo uma proflavina e uma fenantrolina. Nas docagens que utilizaram DNA canônico com gap como receptor, os resultados reproduziram corretamente a estereosseletividade da interação, que favorece os isômeros levorrotatórios em relação aos dextrorrotatórios, e apontou para os modos de interação sugeridos pela literatura: ligação ao sulco menor para a Tröger simétrica e intercalação com ligação ao sulco menor (modo misto) para a Tröger assimétrica. Para as simulações, foram escolhidos quatro complexos contendo os enantiômeros levorrotatórios das bases de Tröger e oligonucleotídeos com e sem gap. Embora os complexos representassem pontos de partida diferentes, houve convergência entre aqueles que continham o mesmo oligonucleotídeo como receptor. Os complexos com DNA com gap (complexos de intercalação) convergiram para um modo misto de interação, com intercalação de uma proflavina, enquanto os complexos com DNA sem gap (complexos de ligação ao sulco menor) convergiram para um modo de interação novo, em que as bases de Tröger se ligam ao sulco menor unicamente através da ponte diazocina, com os substituintes projetados para fora. Os complexos de intercalação apresentaram tempos de residência altos (20 ns) e diminuíram o ângulo de torção da dupla hélice na região do gap, levando os ângulos x e y da cadeia principal para uma região não-canônica, associada à forma A do DNA. Já nos complexos de ligação ao sulco menor, houve mudança tanto do modo como do sítio de interação e, embora tenham ocorrido eventuais curvaturas no DNA, não houve alterações significativas nos ângulos da cadeia principal, de modo que a estrutura global do DNA se manteve dentro da região considerada B-canônica. Dessa forma, os resultados sugerem que a intercalação (com adicionais contatos no sulco menor) seja o modo preferencial de interação dessas bases de Tröger com o DNA. O modo de ligação ao sulco menor, por depender quase que exclusivamente da ponte diazocina, seria independente da quiralidade e, portanto, não seria capaz de explicar enantiosselevidade das bases de Tröger. / In this work, molecular docking and molecular dynamics simulations were applied to investigate the interaction between DNA oligomers and two Tröger bases: a symmetric Tröger base derived from proflavine and a mixed Tröger base containing proflavine and phenanthroline. The dockings that used canonical DNA with an artificial intercalation gap as receptor correctly reproduced the stronger DNA binding affinity of the levorotatory isomers when compared to the dextrorotatory ones. Moreover, the results pointed out to the binding modes suggested, yet not established, by experimental data: minor groove binding for symmetric Tröger base and intercalation with additional minor groove binding for the asymmetric one. For the simulations, four complexes were selected, containing the (-)-isomers of the Tröger bases and DNA oligomers with or without intercalation gap. Interestingly, the systems containing the same DNA as receptor converged to similar binding modes: those containing DNA with gap (intercalation complexes) resulted in a mixed binding – with one proflavine intercalated and the other moiety occupying the minor groove – while those containing DNA without gap (minor groove complexes) lead to a new binding mode, in which the Tröger bases interact with the minor groove mainly through the diazocin bridge. In the intercalation complexes, the ligands presented long residence times (20 ns) and also unwound the double helix in the region of the intercalation gap, leading some backbone angles (x and y) to assume non-canonical values typical of the A form of DNA. In minor groove complexes, the ligands displayed enhanced mobility – changing both the binding mode and the binding site – and, although they promoted some bending of the double helix, the main chain torsion angles of DNA remained in the B-canonical region. Altogether, these results suggest that intercalation of proflavine (with additional contacts in the minor groove) might be the stronger DNA binding mode for these Tröger bases. Since the minor groove binding depends almost exclusively of the diazocin bridge, it appears to perform independently of the ligand chirality and, therefore, could not explain the enantioselectivity of the interaction.

Historiquement, les Modèles Graphiques Probabilistes (PGMs) sont une solution d’apprentissage à partir des données incertaines et plates, appelées aussi données propositionnelles ou représentations attribut-valeur. Au début des années 2000, un grand intérêt a été adressé au traitement des données relationnelles présentant un grand nombre d’objets participant à des différentes relations. Les Modèles Probabilistes Relationnels (PRMs) présentent une extension des PGMs pour le contexte relationnel. Avec l’évolution rapide issue de l’internet, des innovations technologiques et des applications web, les données sont devenues de plus en plus variées et complexes. D’où l’essor du Big Data. Plusieurs types de bases de données ont été créés pour s’adapter aux nouvelles caractéristiques des données, dont les plus utilisés sont les bases de données graphe. Toutefois, tous les travaux d’apprentissage des PRMs sont consacrés à apprendre à partir des données bien structurées et stockées dans des bases de données relationnelles. Les bases de données graphe sont non structurées et n’obéissent pas à un schéma bien défini. Les arcs entre les noeuds peuvent avoir des différentes signatures. En effet, les relations qui ne correspondent pas à un modèle ER peuvent exister dans l'instance de base de données. Ces relations sont considérées comme des exceptions. Dans ce travail de thèse, nous nous intéressons à ce type de bases de données. Nous étudions aussi deux types de PRMs à savoir, Direct Acyclic Probabilistic Entity Relationship (DAPER) et chaines de markov logiques (MLNs). Nous proposons deux contributions majeures. Premièrement, Une approche d’apprentissage des DAPERs à partir des bases de données graphe partiellement structurées. Une deuxième approche consiste à exploiter la logique de premier ordre pour apprendre les DAPERs en utilisant les MLNs pour prendre en considération les exceptions qui peuvent parvenir lors de l’apprentissage. Nous menons une étude expérimentale permettant de comparer nos méthodes proposées avec les approches déjà existantes. / Historically, Probabilistic Graphical Models (PGMs) are a solution for learning from uncertain and flat data, also called propositional data or attributevalue representations. In the early 2000s, great interest was addressed to the processing of relational data which includes a large number of objects participating in different relations. Probabilistic Relational Models (PRMs) present an extension of PGMs to the relational context. With the rise of the internet, numerous technological innovations and web applications are driving the dramatic increase of various and complex data. Consequently, Big Data has emerged. Several types of data stores have been created to manage this new data, including the graph databases. Recently there has been an increasing interest in graph databases to model objects and interactions. However, all PRMs structure learning use wellstructured data that are stored in relational databases. Graph databases are unstructured and schema-free data stores. Edges between nodes can have various signatures. Since, relationships that do not correspond to an ER model could be depicted in the database instance. These relationships are considered as exceptions. In this thesis, we are interested by this type of data stores. Also, we study two kinds of PRMs namely, Direct Acyclic Probabilistic Entity Relationship (DAPER) and Markov Logic Networks (MLNs). We propose two significant contributions. First, an approach to learn DAPERs from partially structured graph databases. A second approach consists to benefit from first-order logic to learn DAPERs using MLN framework to take into account the exceptions that are dropped during DAPER learning. We are conducting experimental studies to compare our proposed methods with existing approaches.