Prandtl in 1904 discovered that the flow of a fluid over a thin obstacle can be adequately represented by an approximate set of equations, much simpler than the complex Navier-Stokes equations which govern the motion of fluid. A particularly simple for of these equations, for the two-dimensional steady flow of a fluid past a flat plate, are the Von Mises Boundary layer equations. Unfortunately the Von Mises transformation introduces a singularity at the plate and this discouraged the use of the equations as a means for obtaining numerical solutions of boundary layer problems in incompressible and compressible flow. In this thesis, we show that this difficulty can be overcome and the Von Mises equations are used as a basis for a finite difference evaluation of the velocity and temperature in the boundary layer adjacent to a flat plate, particular attention being given to conditions near the plate and more especially to the separation point. In the section on compressible flow, the calculations also yield a check on certain common simplifying assumptions.

The thesis intends to provide a better understanding of the text and the background of the Valentinian treatise, Nag Hammadi Codex, I,5. The Introduction studies the manuscript (date and provenance, purpose, scribal signs, quality), the text (an anonymous and untitled treatise, originally written in Greek, representing the Oriental branch of Valentinianism, date most likely second half of the 3rd. cent. A.D.), the language (a form of Subachmimic, with numerous orthographic and grammatical peculiar ities). A brief survey of the system is also provided, where it is regarded from three different angles. The Translation is primarily meant as an attempt to elucidate the difficult, and inadequately understood, Coptic text, and as an index to the following Commentary. The Commentary discusses the translation and relates each passage to the treatise as a whole, and to the system it contains. Valentinian themes and technical terms are pointed out and analysed systematically. The broader religious and philosophical background for the ideas contained in the treatise have also been explored. A special effort has been made to relate the system of the treatise not only to Gnostic documents, Christian literature and Late Jewish material, but also to Philosophy, and in particular to the emanationist physics of Neopythagoreanism and Neoplatonism.

In 9152, the evidence available was still insufficient to decide between various structures proposed for the rubrementinium salts. The author has therefore attempted to gather further evidence. The evidence available in the literature is discussed in Part II, and the author's work in Parts III and IV of this thesis. The substances which the present author has termed rubremetinium salts have more than one name in the literature. Karrer uses the name “dehydroemetin”. Pyman, who discovered the salts, at first used “rubremetine”. It was later found, however, that the salts are quaternary, so Pyman applied to the quaternary ion the name “rubremetinium”. This name will be used throughout the present thesis.

A linear programming cost minimization problem is presented to demonstrate the applicability of the technique to problems inherent in the forest management decision making process. A case study of an integrated forest products firm with the objective of minimizing the present value cost of wood procurement over a 20 year study is solved by a computer using the firm's present and three alternate price structures. The study is implemented with the product quality constraints and a managerial constraint of the firm and incorporates timber growth into the selection of procurement alternatives. The minimum present value cost solutions of the four price structures studied are presented and discussed. Additional information regarding the stability of the solution to changes in future expectations, easily obtained by various computer agendums, is evaluated. / M.S.

This thesis approaches the Gesta Roberti Wiscardi as a means of gaining an insight into the cultural values of its author and intended audience. A detailed study is made of the various role models within the poem: the ideal soldier, the good lord, the role of women in society, and the perception of priests and the papacy. In addition to this the text is used to establish racial stereotypes for the following groups of peoples: the Germans, Sicilians, Seljuqs, Greeks, Italians, Venetians and Normans. The significance of the characterisation of individuals who are portrayed in a manner inconsistent with their racial stereotype is also examined. The thesis re-examines the evidence in the text and in other document sources concerning the author of the poem and establishes a viable identification. A new interpretation of the role of the two patrons. Urban II and Roger Borsa, is also discussed. The question of the consistency of style in William of Apulia's poem is also addressed and set in the context of the subject matter and intent of the work. Finally a discussion is made of the evidence for the use of William as a source by three subsequent historians: Robert of Torigni, Suger of St Denis and Anna Comnena. This thesis draws attention to further use of the Gesta by Robert than previously realised and for the first time forwards a concrete case for its use by the latter two authors.

Le bactériophage T5, qui infecte la bactérie Escherichia coli, est constitué d’une capside icosaédrique renfermant un ADN double brin et d’une queue permettant le transfert de son ADN dans le cytoplasme de la cellule hôte. Au cours de la morphogénèse, la capside et la queue sont assemblées séparément puis connectées pour former les particules infectieuses. La capside de T5 est d'abord assemblée sous forme d'une procapside vide constituée de 775 copies d’une protéine unique qui s'organise en hexamères et pentamères pour former respectivement les faces et les sommets de l’icosaèdre. Un des sommets est occupé par la protéine portale qui constitue un pore d’entrée pour l’ADN. L’encapsidation du génome à travers le canal portal, assurée par un moteur moléculaire, provoque l’expansion de la procapside. La capside pleine d'ADN est ensuite décorée par la protéine pb10 qui se fixe sur sa surface externe, au centre des 120 hexamères, et elle est fermée par la protéine p144 qui constitue le point d'ancrage pour la queue.Les objectifs de ma thèse étaient de déterminer la structure, la fonction et le mécanisme de fixation à la capside des deux protéines pb10 et p144. J’ai produit et purifié la protéine de fermeture p144 dont la structure et les caractéristiques de fixation sur la capside seront prochainement étudiées. J’ai résolu la structure de la protéine de décoration pb10 en solution par résonance magnétique nucléaire (RMN), ce qui a révélé une protéine formée de deux domaines. Le domaine N-terminal structuré en hélices alpha se fixe à la capside, alors que le domaine C-terminal, de type immunoglobuline, est exposé au solvant. La détermination des constantes cinétiques d'association et de dissociation par résonance plasmonique de surface (SPR) a permis de montrer que pb10 se fixe sur la capside de T5 de manière quasi-irréversible, avec une affinité de l’ordre du picomolaire. Des expériences de mutagenèse dirigée sur le domaine de fixation de pb10 indiquent que cette forte affinité repose sur un réseau d’interactions électrostatiques et hydrophobes. J’ai démontré que pb10 participe à la stabilisation de la capside de T5 par des techniques de calorimétrie différentielle à balayage (DSC) et thermo-fluorescence (FTSA). Par ailleurs, j'ai démontré qu’il était possible de fusionner des protéines de grande taille au domaine C-terminal de pb10 sans affecter sa haute affinité pour la capside. J’ai ainsi pu fonctionnaliser des capsides avec une protéine d’intérêt fusionnée à pb10. Ces résultats ouvrent la voie vers diverses applications, dont la présentation d’antigènes en vue de créer de nouveaux vaccins. / Bacteriophage T5, a lytic phage which infects the bacterium Escherichia coli, is composed of an icosahedral capsid containing a double stranded DNA and a tail responsible for DNA transfer into the host cell cytoplasm. The capsid and the tail are assembled separately and then connected to form infectious viruses. The T5 capsid is first assembled as an empty procapsid composed of 775 copies of the major head protein organized in hexamers on the faces and pentamers on the vertices of the icosahedron. A single vertex is occupied by the portal protein which forms an entry channel for DNA. DNA is packaged through the portal channel by a molecular motor thus triggering procapsid expansion. The DNA-filled capsid is then decorated by the protein pb10 which binds on the outer surface in the center of each hexamer, and closed by the connector protein p144 needed for tail attachment.The goals of my PhD were to characterize the structure, function and mechanisms of capsid binding of both pb10 and p144 proteins. I have produced and purified the head completion protein p144 for future exploration of its structure and capsid binding properties. I have solved the solution structure of the decoration protein pb10 by nuclear magnetic resonance (NMR), thus unravelling that pb10 has two domains. The alpha-helical N-terminal domain binds to the capsid and the immunoglobulin-like C-terminal domain is exposed to the environment. Surface plasmon resonance (SPR) showed that that pb10 attachment to the T5 capsid is quasi-irreversible with a picomolar affinity. Site-directed mutagenesis of the binding domain of pb10 suggested that the decoration mechanism is driven by both hydrophobic and electrostatic interactions. Differential scanning calorimetry (DSC) and fluorescence thermal shift assays (FTSA) demonstrated the role of the decoration protein in capsid stabilization. I have also demonstrated that a large protein can be fused to the C-terminal end of pb10 without affecting its high affinity for the T5 capsid. Capsids can thus be functionalized with a protein of interest fused to pb10. Possible applications of this work include antigen presentation for new vaccines.

In this thesis I give a response to external world scepticism. I first argue that scepticism arises when we accept that it is an empirical question whether I am in a sceptical scenario, that is, a scenario in which my beliefs are coherent, and yet my empirical beliefs are false. The idea that it is an empirical question whether I am in a sceptical scenario gets its plausibility from the realist claim that our empirical beliefs have an objective subject matter. I then attempt to give a response to scepticism that is compatible with this realist claim. Three promising responses to scepticism are considered, but are found to be inadequate. Seeing why these responses are inadequate helps us to appreciate some of the conditions on an adequate response to scepticism. By drawing on the work of Donald Davidson I develop a response to scepticism that is compatible with the realist claim, and that meets these conditions. According to this response, when we get clear about the concept of belief we see that sceptical scenarios are a conceptual impossibility. Thus, just as it is not an empirical question whether I am a married bachelor, it is not an empirical question whether I am in a sceptical scenario, and the argument for scepticism breaks down.

Deep learning models are dominating performance across a wide variety of tasks. From protein folding to computer vision to voice recognition, deep learning is changing the way we interact with data. The field of natural products, and more specifically genomic mining, has been slow to adapt to these new technological innovations. As we are in the midst of a data explosion, it is not for lack of training data. Instead, it is due to the lack of a blueprint demonstrating how to correctly integrate these models to maximise performance and inference. During my PhD, I showcase the use of large language models across a variety of data domains to improve common workflows in the field of natural product drug discovery. I improved natural product scaffold comparison by representing molecules as sentences. I developed a series of deep learning models to replace archaic technologies and create a more scalable genomic mining pipeline decreasing running times by 8X. I integrated deep learning-based genomic and enzymatic inference into legacy tooling to improve the quality of short-read assemblies. I also demonstrate how intelligent querying of multi-omic datasets can be used to facilitate the gene signature prediction of encoded microbial metabolites. The models and workflows I developed are wide in scope with the hopes of blueprinting how these industry standard tools can be applied across the entirety of natural product drug discovery. / Thesis / Doctor of Philosophy (PhD)

Orientador: Michael John Brennan / Resumo: A independência funcional para indivíduos com deficiência motora é indispensável para a qualidade de vida, principalmente quando se considera tarefas como o autocuidado (incluindo a higiene pessoal), as transferências e locomoção. Essa pesquisa apresenta um triciclo inovador para pessoas com paraplegia ou alguma deficiência motora dos membros inferiores, desenvolvido com duas forma de acionamento: cíclico manual (handbike) e elétrico. A estrutura tubular foi elaborada com liga de alumínio 6063 - T5. A plataforma ANSYS Academic versão 19.2 foi empregada para uma análise estrutural do chassi do equipamento. O chassi foi submetido num ensaio de carregamento estático progressivo de até 1200 N aproximadamente. O deslocamento onde apoia os pés do usuário foi medido por um relógio comparador de precisão de 0,01 mm. O deslocamento da plataforma inferior de apoio dos pés apresentou uma divergência de aproximadamente 4% se comparado com os dados obtidos pela plataforma ANSYS. Assim, validando a análise teórica pela plataforma ANSYS e confirmando que a estrutura fabricada por dobramento e solda TIG suporta com segurança um usuário de até 100 kg. Parâmetros relacionados ao sistema elétrico de acionamento do motor brushless de 1000 W foram obtidos por meio de sensores de corrente e voltagem, assim como parâmetros de desempenho mecânico foram obtidos empregando uma célula de carga de fundo de escala de 2kN para análise de força de tração e um sensor de velocidade. Os resultados indicam uma... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Functional independence for individuals with motor disabilities is essential for quality of life, especially when considering tasks such as self-care (including personal hygiene), transfers and locomotion. This research presents an innovative tricycle for people with motor deficiency of the lower limbs, developed with two forms of drive: manual cyclic (handbike) and electric. The tubular structure was made with 6063 - T5 aluminum alloy. The ANSYS Academic version 19.2 platform was used for a structural analysis of the equipment chassis. A static loading test of up to 1200 N was performed on the chassis. The deformation of support the user's feet was measured by a dial gauge with a 0,01 mm precision. The deformation of the lower footrest platform showed a divergence of approximately 4% when compared to the data obtained by the ANSYS platform. Thus, validating the theoretical analysis by the ANSYS platform and confirming that the structure manufactured by TIG welding safely supports a user of up to 100 kg. Parameters related to the electric drive system of the 1000 W brushless motor were obtained by means of current and voltage sensors, as well as mechanical performance parameters were obtained employing a load cell of 2kN for force analysis and a speed sensor. The results indicate an energy efficiency of the equipment of up to 92%. The maximum strength of the equipment is 500 N, which makes it possible to perform a climb with slope of 20º, approximately considering a user of 8... (Complete abstract click electronic access below) / Doutor

Les bactériophages neutralisent les systèmes de défense et détournent les fonctions vitales de leur hôte pour favoriser leur multiplication. Les gènes de phages qui gouvernent cette prise de contrôle de l’hôte restent mal connus, pourtant leur caractérisation présente un intérêt majeur pour mettre à jour des fonctions bactériennes spécifiquement ciblées par les phages et pour concevoir de nouveaux agents antibactériens.Le phage T5 injecte son ADN dans la bactérie Escherichia coli en deux étapes. Seuls les gènes précoces codés par 8% du génome entrent dans la cellule et le transfert s’arrête. Leur expression induit la dégradation du chromosome de l’hôte et l’inactivation de ses systèmes de restriction et de réparation de l’ADN. Après quelques minutes, le reste de la molécule d’ADN est injecté, ce qui permet la production de nouveaux phages. Deux gènes précoces A1 et A2 ont été identifiés comme essentiels pour la reprise du transfert de l’ADN et A1 est également nécessaire pour induire la dégradation de l’ADN de l’hôte. A1 et A2 sont les deux seuls gènes connus pour être impliqués dans la régulation de ce système original d’infection, mais leur fonction n’a jamais été identifiée.Ma thèse porte sur la caractérisation fonctionnelle et structurale des protéines A1 et A2. J’ai purifié A1 et démontré in vitro qu’elle avait une activité DNase dépendante du manganèse. Sa structure atomique a été résolue par cryomicroscopie électronique à 3.01 Å de résolution, montrant une organisation octamérique de symétrie D4 inédite pour une DNase. Chaque monomère (61kDa) contient un domaine exonuclease dont le site actif lie deux ions Mn2+ et qui s’apparente au site catalytique des domaines exonucléases de la DNA polymerase II et des DNAses associées aux systèmes de recombinaison homologue et de réparation de l’ADN comme Mre11. En construisant différents mutants de A1, j’ai identifié certains acides aminés essentiels pour l’activité catalytique et, par des expériences de complémentation fonctionnelle, j’ai montré que cette activité était indispensable pour l’infection. L’ensemble de ces résultats suggèrent que A1 est la DNase, jusqu’ici inconnue, responsable de la dégradation massive du génome de l’hôte au tout début de l’infection. Enfin, j’ai observé que la production de A1 pendant l’infection induit une forte activité recombinase. De nombreux autres bactériophages qui n’appartiennent pas à la famille des T5virus produisent également une protéine similaire à A1 dont la fonction n’a jamais été identifiée. Ce travail est un premier pas vers la compréhension de son rôle dans le mécanisme général d‘infection par les phages. Une deuxième partie de cette thèse porte sur la caractérisation structurale de A2. Des recherches de similarité indiquent la présence d’un domaine Helix-Turn-Helix typique des régulateurs transcriptionnels. J’ai purifié A2 et montré que cette protéine de 14 kDa est un dimère en solution. La caractérisation des propriétés biochimiques de A2 a permis de débuter l’étude de sa structure par RMN.Les résultats de ma thèse ont révélé la structure originale d’une DNase de bactériophage qui contrôle la dégradation du génome bactérien et la régulation du transport de l’ADN viral au début du cycle infectieux. Ces résultats soulèvent des questions intrigantes : comment l’ADN de T5 est-il protégé de l’activité DNase de A1 ? Comment A1 et A2 interagissent-elles lors des étapes de prise de contrôle de l’hôte ? / Bacteriophages defeat bacterial defences and hijack host cell machineries to establish a favourable environment for their multiplication. Early-expressed viral genes that govern host takeover are highly diverse from one phage to another and most of them have no assigned function. They thus represent a pool of novel genes whose products potentially subvert bacterial cell vital functions and could help in designing new antibacterial strategies.T5 phage uses a unique 2-step mechanism to deliver its DNA into its host Escherichia coli. At the onset of the infection, only 8 % of the genome enter the cell before the transfer temporarily stops. Expression of the genes encoded by this DNA portion leads to host chromosome degradation and inactivation of host restriction and DNA mending systems. After a few minutes, T5 DNA transfer resumes, allowing further phage multiplication. A1 and A2 are early genes required for DNA transfer completion and A1 is also necessary to trigger host DNA degradation. A1 and A2 are the only two genes known to be involved in the regulation of this original infection system, but their function yet remains to be characterized.The objectives of this work were to characterize the function and structure of A1 and A2 proteins. I have purified the A1 protein and shown that it has a manganese-dependent DNase activity in vitro. Cryo Electron Microscopy at 3.01 Å resolution unravelled its structure, showing an octameric organization with a D4 symmetry, which is unprecedented for a DNase. Each monomer (61 kDa) carries an exonuclease domain harbouring an active site with two Mn2+ ions. This site is similar to those from the exonuclease domain of the DNA polymerase II and from DNases involved in DNA mending and recombination events like Mre11. I identified essential catalytic residues for the DNase activity and demonstrated that this activity is crucial for infection by engineering A1 mutant proteins and by doing functional complementation assays. Taken together, my results suggest that A1 could then be the elusive DNase responsible for the massive host genome degradation observed during T5 phage infection. Eventually, I uncovered a recombinase activity associated to A1 production during infection. Similar proteins to A1 with unknown functions are produced in several other bacteriophages outside of the T5virus family. This work is a first step towards understanding the role of this protein in the general mechanism of infection by bacteriophages. In a second part, I worked on the structural characterisation of A2 protein. Similarity searches revealed a helix-turn-helix domain typically found in transcriptional regulators. I purified and demonstrated the dimeric organisation of this 14-kDa protein in solution. This initial characterization of A2 has opened avenues for further NMR studies.During my Ph.D., I uncovered the structure of an original bacteriophage DNase that controls bacterial genome degradation and that regulates viral DNA transport at the beginning of the infectious cycle. These results open the intriguing question about the mechanism for T5 DNA protection from A1 DNase activity as well as about the interplay between A1 and A2 during the host takeover.