In recent years, deep learning based sequence modeling (neural sequence modeling) techniques have made substantial progress in many tasks, including information retrieval, question answering, information extraction, machine translation, etc. Benefiting from the highly scalable attention-based Transformer architecture and enormous open access online data, large-scale pre-trained language models have shown great modeling and generalization capacity for sequential data. However, not all domains benefit equally from the rapid development of neural sequence modeling. Domains like healthcare and software engineering have vast amounts of sequential data containing rich knowledge, yet remain under-explored due to a number of challenges: 1) the distribution of the sequences in specific domains is different from the general domain; 2) the effective comprehension of domain-specific data usually relies on domain knowledge; and 3) the labelled data is usually scarce and expensive to get in domain-specific settings. In this thesis, we focus on the research problem of applying neural sequence modeling methods to address both common and domain-specific challenges from the healthcare and software engineering domains. We systematically investigate neural-based machine learning approaches to address the above challenges in three research directions: 1) learning with long sequences, 2) learning from domain knowledge and 3) learning under limited supervision. Our work can also potentially benefit more domains with large amounts of sequential data. / Doctor of Philosophy / In the last few years, computer programs that learn and understand human languages (an area called machine learning for natural language processing) have significantly improved. These advances are visible in various areas such as retrieving information, answering questions, extracting key details from texts, and translating between languages. A key to these successes has been the use of a type of neural network structure known as a "Transformer", which can process and learn from lots of information found online. However, these successes are not uniform across all areas. Two fields, healthcare and software engineering, still present unique challenges despite having a wealth of information. Some of these challenges include the different types of information in these fields, the need for specific expertise to understand this information, and the shortage of labeled data, which is crucial for training machine learning models. In this thesis, we focus on the use of machine learning for natural language processing methods to solve these challenges in the healthcare and software engineering fields. Our research investigates learning with long documents, learning from domain-specific expertise, and learning when there's a shortage of labeled data. The insights and techniques from our work could potentially be applied to other fields that also have a lot of sequential data.

Tesis por compendio / [EN] The development of new and more complex polymeric materials involves challenging problems to basic sciences. The relationship between structure and molecular dynamics assumes great importance for the future development of novel technologies based on such polymers. Thus, the understanding of how small changes in the chemical structure affect the properties of the material is essential to progress in the technological and scientific area. An in-depth analysis of the molecular mobility leads to establish the structure-properties relationships. On this basis, the main aim of the present work is to study the molecular mobility of two different families of polymeric materials. For this purpose, the experimental techniques mainly used were Differential Scanning Calorimetry (DSC) and Dielectric Relaxation Spectroscopy (DRS). The first family of polymers characterized was a series of chemically cross-linked copolymers composed by Vinylpyrrolidone (VP) and Butyl Acrylate (BA) monomers. In the first place, the influence of the monomer molar ratio (XVP/YBA) on the copolymer properties was studied. Thus, a Fourier Transform Infrared Spectroscopy (FTIR) analysis verified dipole-dipole interactions between amide groups. The influence of these interactions on several parameters related to the molecular mobility was evidenced by the DSC, DRS and Dynamic Mechanical Analysis (DMA) techniques. Secondly, the effect of the cross-link density on the molecular dynamics of 60VP/40BA copolymers was analyzed using DSC and DRS. One single glass transition was detected by DSC measurements. The DRS analysis showed that an increase of the cross-linking produced a typical effect on the alpha process dynamics. However, the beta process, which possessed typical features of pure JG relaxation, unexpectedly lost the intermolecular character for the highest cross-linker content. The fastest gamma process was relatively unaffected. The second family of polymeric materials studied was a series of segmented polycarbonatediol polyurethane (PUPH) modified with different amounts of expanded graphite (EG) conductive filler. Scanning Electron Microscopy (SEM), X-ray diffraction measurements and FTIR analysis demonstrated a homogeneous dispersion of the EG filler in the matrix. DRS was used to study the dielectric properties of the PUPH/EG composites. The dielectric permittivity of the composites showed an insulator to conductor percolation transition with the increase of the EG content (2030 wt%). The addition of expanded graphite to the matrix caused a dramatic increase in the electrical conductivity of ten orders of magnitude, which is an indication of percolative behavior. / [ES] El desarrollo de nuevos materiales poliméricos de mayor complejidad produce un desafío cada vez mayor en el área de las ciencias básicas. La relación entre la estructura y la dinámica molecular resulta de gran importancia para el desarrollo de nuevas tecnologías basadas en estos materiales poliméricos. Así, una mayor comprensión de cómo pequeños cambios en la estructura química afectan a las propiedades de los materiales resulta esencial para el progreso científico y tecnológico. Un análisis en profundidad de la movilidad molecular permite establecer las relaciones estructura-propiedades. Partiendo de esta base, el principal objetivo del presente trabajo es el estudio de la movilidad molecular de dos familias diferentes de materiales poliméricos. Para ello, las técnicas experimentales utilizadas fueron principalmente la Calorimetría Diferencial de Barrido (DSC) y la Espectroscopia de Relajación Dieléctrica (DRS). La primera familia de polímeros caracterizada fue una serie de copolímeros entrecruzados químicamente compuestos por los monómeros Vinilpirrolidona (VP) y Acrilato de Butilo (BA). En primer lugar, se estudió la influencia de la proporción molar de monómero (XVP/YBA) en las propiedades del copolímero. A través de un análisis por Espectroscopia de Infrarrojo por Transformada de Fourier (FTIR), se verificó la existencia de interacciones dipolo-dipolo entre los grupos amida. Mediante el análisis por DSC, DRS y Análisis Dinamomecánico (DMA), se evidenció la influencia de estas interacciones en diferentes parámetros relacionados con la movilidad molecular. En segundo lugar, se analizó el efecto de la densidad de entrecruzamiento en la dinámica molecular de los copolímeros 60VP/40BA usando DSC y DRS. A través de las medidas de DSC se observó una única transición vítrea para todos los entrecruzamientos. El análisis por DRS mostró como el incremento en entrecruzante produjo el típico efecto en la dinámica del proceso alpha, pero sin embargo, el proceso beta, que tenía las características típicas de una relajación JG, perdió de forma inesperada su carácter intermolecular para el mayor contenido en entrecruzante. El proceso gamma no se vio afectado. La segunda familia de materiales poliméricos estudiada fue una serie de poliuretanos segmentados (PUPH) modificados con diferentes cantidades de grafito expandido (EG), utilizado como relleno conductivo (desde 0 a 50% en peso). El análisis de los resultados obtenidos mediante Microscopía Electrónica de Barrido (SEM), Difracción de Rayos X y FTIR demostró la homogénea dispersión del relleno de EG en la matriz de PUPH. La técnica DRS se usó para estudiar las propiedades dieléctricas de los materiales compuestos PUPH/EG. La permitividad dieléctrica de los materiales mostró una transición de percolación desde aislante a conductor al incrementarse el contenido en EG (rango de 20-30% en peso). La adición de grafito expandido a la matriz de PUPH causó un incremento significativo en la conductividad dieléctrica de diez órdenes de magnitud, lo que indica el comportamiento de percolación. / [CA] El desenvolupament de nous materials polimèrics de major complexitat produeix un desafiament cada vegada major en l'àrea de les ciències bàsiques. La relació entre l'estructura i la dinàmica molecular resulta de gran importància per al desenrotllament de noves tecnologies basades en aquests materials polimèrics. Així, una major comprensió de com petits canvis en l'estructura química afecten a les propietats dels materials, resulta essencial per al progrés científic i tecnològic. Un anàlisis en profunditat de la mobilitat molecular permet establir les relacions estructura-propietats. Partint d'aquesta base, el principal objectiu del present treball és l'estudi de la mobilitat molecular de dues famílies diferents de materials polimèrics. Per a això, les tècniques experimentals utilitzades van ser principalment la Calorimetria Diferencial de Rastreig (DSC) i l'Espectroscòpia de Relaxació Dielèctrica (DRS). La primera família de polímers caracteritzada va ser una sèrie de copolímers entrecreuats químicament compostos pels monòmers Vinilpirrolidona (VP) i Acrilat de Butilo (BA) . En primer lloc, es va estudiar la influència de la proporció molar de monòmer (XVP/YBA) en les propietats del copolímer. A través d'una anàlisi per Espectroscòpia d'Infraroig per Transformada de Fourier (FTIR), es va verificar l'existència d'interaccions dipol-dipol entre els grups amida. Mitjançant l'anàlisi per DSC, DRS i Anàlisi Dinamomecánico (DMA), es va evidenciar la influència d'aquestes interaccions en diferents paràmetres relacionats amb la mobilitat molecular. En segon lloc, es va analitzar l'efecte de la densitat d'entrecreuament en la dinàmica molecular dels copolímers 60VP/40BA mitjançant DSC i DRS. A través de les mesures de DSC es va observar una única transició vítria per a tots els continguts d'agent entrecreuant . L'anàlisi per DRS va mostrar com l'increment en agent entrecreuant va produir l'efecte esperat en la dinàmica del procés alfa. En canvi, el procés beta, que tenia les característiques típiques d'una relaxació JG, va perdre de forma inesperada el seu caràcter intermolecular per al major contingut en agent entrecreuant. El procés més ràpid gamma no es va veure afectat. La segona família de materials polimèrics estudiada va ser una sèrie de poliuretans segmentats (PUPH) modificats amb diferents quantitats de grafit expandit (EG) , utilitzat com a farcit conductiu (des de 0 a 50% en pes). L'anàlisi dels resultats obtinguts per mitjà de Microscòpia Electrònica de Rastreig (SEM), Difracció de Rajos X i FTIR va mostrar la dispersió homogènia del EG en la matriu de PUPH. La tècnica DRS es va utilitzar per a estudiar les propietats dielèctriques dels materials compostos PUPH/EG. La permitivitat dielèctrica dels materials va mostrar una transició de percolació des d'aïllant a conductor amb l'increment de contingut en EG (20-30% en pes). L'addició d'EG a la matriu de PUPH va causar un increment significatiu en la conductivitat dielèctrica, de deu ordes de magnitud. / Redondo Foj, MB. (2015). A contribution to the study of the molecular mobility in polymeric materials by Thermal and Dielectric Analysis [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/59457 / Compendio

Submitted by Erika Demachki (erikademachki@gmail.com) on 2018-03-23T22:13:05Z No. of bitstreams: 2 Tese - Thiago Rodrigues Cavalcante - 2018.pdf: 2200622 bytes, checksum: 39118adda6b7ceff14825da442b5be57 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2018-03-26T12:16:44Z (GMT) No. of bitstreams: 2 Tese - Thiago Rodrigues Cavalcante - 2018.pdf: 2200622 bytes, checksum: 39118adda6b7ceff14825da442b5be57 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2018-03-26T12:16:44Z (GMT). No. of bitstreams: 2 Tese - Thiago Rodrigues Cavalcante - 2018.pdf: 2200622 bytes, checksum: 39118adda6b7ceff14825da442b5be57 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2018-02-27 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / In the first two chapters, we consider the following problem \begin{equation*} \left \{ \begin{array}{rcll} \alpha \Delta^{2} u + \beta \Delta u & = & f(x,u)\, & \mbox{in}\,\, \Omega \\ u = \Delta u & = & 0 \, &\mbox{on } \,\,\, \partial \Omega, \end{array} \right. \end{equation*} where $\displaystyle{\Delta^{2} u = \Delta(\Delta u)-\,\mbox{biharmonic (fourth-order operator)}}$, $\alpha > 0$ and $ \beta \in \R.$ The subset $\displaystyle{ \Omega \subset \mathbb{R}^{N}\, (N \geq 4)}$ is as somooth bounded domain and $\displaystyle{ f \in C(\overline{\Omega} \times \mathbb{R},\mathbb{R}) }.$ In each of the results obtained, we will consider different technical hypotheses and characteristics for the nonlinear function $f$ e for the value of the constant $ \beta. $ In the third chapter, we study an equation of the concave type super linear, of the form: \begin{equation} \left \{ \begin{array}{rcll} \alpha \Delta^{2} u + \beta \Delta u & = & a(x)|u|^{s-2}u + f(x,u)\, & \mbox{in}\,\, \Omega \\ u = \Delta u & = & 0 \, &\mbox{on} \,\,\, \partial \Omega, \end{array} \right. \end{equation} where $\beta \in (-\infty, \alpha \lambda_{1}).$ We consider that the function $a \in L^{\infty} (\Omega)$ and $s \in (1,2).$ Finally, in the last chapter we will consider a fourth order problem in which nonlinearity is also of the convex concave type. More precisely, we study the following class of equations: \begin{equation} \left\{ \begin{aligned} \alpha \Delta^{2} u + \beta \Delta u & = \mu a(x)|u|^{q-2}u + b(x)|u|^{p-2}u&\,\,\,\,\ &\mbox{in}\,\, \Omega \\ u = \Delta u & = 0 & \,\,\,\,&\mbox{on} \,\, \partial \Omega, \end{aligned} \right. \end{equation} where the parameter $ \mu > 0 $, the powers $ 1 <q <2 <p <2 N / (N - 4) $. In addition we assume that the functions $ \displaystyle {a, b: \Omega \rightarrow \mathbb {R}}$ are continuous that can change signal and, $ a ^{+}, b ^{+} \neq 0. $ / Nos dois primeiros Capítulos, consideramos a seguinte classe de problemas: \begin{equation*} \left \{ \begin{array}{rcll} \alpha \Delta^{2} u + \beta \Delta u & = & f(x,u)\, & \mbox{em}\,\, \Omega \\ u = \Delta u & = & 0 \, &\mbox{sobre } \,\,\, \partial \Omega, \end{array} \right. \end{equation*} onde $\displaystyle{\Delta^{2} u = \Delta(\Delta u)-\,\mbox{biharmônico},}$ $\alpha > 0$ e $ \beta \in \R.$ O subconjunto $\displaystyle{ \Omega \subset \mathbb{R}^{N}\,(N \geq 4)}$ será um domínio limitado e a não linearidade $\displaystyle{ f \in C(\overline{\Omega} \times \mathbb{R},\mathbb{R}) }.$ Em cada um dos resultados obtidos, consideraremos hipóteses técnicas e características diferentes para a função não linear $f$ e para o valor da constante $\beta.$ No terceiro Capítulo, estudamos uma equação do tipo côncavo super linear, da forma: \begin{equation*} \left \{ \begin{array}{rcll} \alpha \Delta^{2} u + \beta \Delta u & = & a(x)|u|^{s-2}u + f(x,u)\, & \mbox{em}\,\, \Omega \\ u = \Delta u & = & 0 \, &\mbox{sobre } \,\,\, \partial \Omega, \end{array} \right. \end{equation*} onde $\alpha > 0$ e $\beta \in (-\infty, \alpha \lambda_{1})$. Consideramos que a função $a \in L^{\infty}(\Omega)$ e que $s \in (1,2).$ Por fim, no último Capítulo vamos considerar um problema de quarta ordem no qual a não linearidade é do tipo côncavo-convexa. Mais precisamente, estudamos a seguinte classe de equações: \begin{equation*} \left\{ \begin{aligned} \alpha \Delta^{2} u + \beta \Delta u & = \mu a(x)|u|^{q-2}u + b(x)|u|^{p-2}u&\,\,\,\,\ &\mbox{em}\,\, \Omega \\ u = \Delta u & = 0 & \,\,\,\,&\mbox{sobre} \,\, \partial \Omega, \end{aligned} \right. \end{equation*} onde o parâmetro $\mu > 0$ e as potências $ 1 < q < 2 < p < 2 N /(N - 4)$. Adicionalmente supomos que as funções $\displaystyle{a, b : \Omega \rightarrow \mathbb{R} }$ sejam contínuas podendo trocar de sinal em $\Omega$ e que $a^{+},b^{+} \neq 0.$

<p>The rotator cuff is a group of four muscles and tendons in the shoulder that function to lift and rotate the arm. Rotator cuff tendon tears are increasingly common: more than 545,000 rotator cuff surgeries occur annually in the US. However, treatment is often complicated by disorganized collagen matrix formed via fibrosis and results in high re-tear rates. Tendon tissue engineering seeks to solve the problem using biomaterials to promote neo-tendon formation to augment repair or regenerate tendon. However, while current biomaterials provide the opportunity to improve tendon healing, they frequently still exhibit fibrosis in preclinical studies. Therefore, a critical need exists to understand the mechanisms of aligned collagen formation when designing biomaterials for tendon tissue engineering. Matrix architecture and transient receptor potential cation channel subfamily V member 4 (TRPV4) regulate aligned collagen formation during tenogenesis in vitro, but the mechanism remains to be determined. Recently, TRPV4 stimulation was found to induce nuclear localization and activation of transcriptional co-activators Yes-associated protein (YAP). YAP expression is upregulated during tendon development, a process characterized by aligned collagen formation, and in response to physiological mechanical stimulation, suggesting it could play an important role in tendon. The objective of this work is to improve tissue engineering strategies and progress toward making a device that regenerate tendon after injury. Aim 1 incorporates tendon-derived matrix into synthetic polymer scaffolds to add biological signaling cues to induce tenogenesis. Aim 2 uses a 2D photolithography system (microphotopatterning) to optimize architectural and structural cues to promote stem cell differentiation toward tenogenic, chondrogenic, and osteogenic lineages. Aim 3 investigates dynamic tensile loading protocols to promote collagen matrix synthesis and improve engineered tendon mechanical function. Aim 4 investigates the role of TRPV4 and YAP in collagen alignment during engineered tendon development.</p>

Die Planung und Ausführung von Bauwerken basiert auf Informationsprozesse, in denen verknüpfte Fachmodelle verschiedener Baudomänen als fachübergreifende Informationsräume verwendet werden. Dabei führen erhöhte Anforderungen an spezialisierte Arbeitsschritte sowie die wachsende Komplexität der Bauprojekte zu einem Anwachsen der Menge, des Umfangs und der Komplexität der ausgetauschten Informationsräume. Bei der Betrachtung des Informationsbedarfs der Bauinformationsprozesse lässt sich eine Kontextabhängigkeit erkennen, in der verschiedene Aspekte des Bearbeitungskontextes sowohl die Menge und Qualität als auch die Ausschnitte und Verknüpfungstiefe der erforderlichen Informationsräume determinieren. Die vorliegende Arbeit befasst sich mit der kontextgerechten Informationsversorgung von Informationsprozessen im Bauwesen. Auf der Grundlage multimodellbasierter Informationsräume wird ein Ansatz vorgestellt, der die Kontextabhängigkeit des Informationsbedarfs durch kontextadaptive Multimodellvorlagen formalisiert und entsprechende kontextgerechte Informationsräume erzeugt. Dafür werden in einem ersten Schritt der Bearbeitungskontext von Bauinformationsprozessen betrachtet sowie informationslogistisch relevante Kontextaspekte identifiziert und durch ein Kontextmodell abgebildet. Für die Formalisierung der unterschiedlichen Einflüsse verschiedener Kontextaspekte auf die Ausgestaltung des Informationsbedarfs wird ein Regelsystem entwickelt, mit dem kontextadaptive Multimodellvorlagen definiert werden können. Durch Auswertung dieser Vorlagen zum Anwendungszeitpunkt lässt sich ein situativer Informationsbedarf antizipieren, auf dessen Basis ein kontextgerechtes Multimodell erzeugt werden kann. Dieser Ansatz ermöglicht die Realisierung einer kontextbewussten Informationslogistik, die den Projektpartnern im Bauwesen genau die Informationsräume bereitstellt, die in einer konkreten Bearbeitungssituation benötigt werden. Für die Bearbeitung regelbasierter Kontextwirkrelationen wird ein Editor vorgestellt, der die Erzeugung kontextadaptiver Multimodellvorlagen unterstützt. Außerdem wird anhand einer Architektur zur Erzeugung kontextgerechter Informationsräume die Vorgehensweise der informationslogistischen Kontextintegration beschrieben, mit der ein kontextbasierter Informationsbedarf antizipiert und ein entsprechendes kontextgerechtes Multimodell erzeugt werden kann. Der Einsatz des vorgestellten Ansatzes wird abschließend anhand eines Beispielszenarios aus der Planungsphase evaluiert, in der im Rahmen verschiedener asynchroner Bauinformationsprozesse Informationsräume gemeinsam bearbeitet werden. / The planning and creating of structures and buildings is based on building information processes, in which linked specialized models of different domains are used as multidisciplinary information spaces. Thereby increased requirements for specialized work processes, and the growing complexity of construction projects lead to an increase in the amount, scope and complexity of the exchanged information spaces. When considering the information requirements of building information processes, a contextdependence is revealed that determines the quantity and quality as well as the cutouts and linking depth of the required information spaces depending on various aspects of the processing context. This thesis addresses contextoriented information supply for collaborative information processes in the construction industry. Based on multimodel information spaces, an approach is presented that formalizes the context dependency of information requirements by contextadaptive multimodel templates and generates corresponding contextoriented information spaces. In a first step different aspects of the process context of building information processes are considered and logistically relevant information are identified and mapped by a context model. For the description of the different influences of various context aspects on the configuration of information needs a regulating system is developed, which can be used to define contextadaptive multi model templates. By evaluating these templates at time of use, situative information requirements can be anticipated and an adequate contextoriented multimodel can be generated. This approach enables the implementation of a contextaware information logistics, which accurately provides the information spaces for the project partners in the construction industry, which are needed in a concrete working situation. For the processing of rulebased context active relations an editor, which supports the generation of contextadaptive multi model templates, is presented. Based on architecture for generating contextappropriate information spaces, the approach of information logistics context integration is described, which allows to anticipate contextbased information needs and to generate a corresponding contextoriented multimodel. The use of the approach is finally evaluated using an example scenario from the planning phase, in which various asynchronous building information processes jointly process an information space.

Corrosion of metallic materials is a natural process, and our study shows that even in an alkaline environment severe corrosion may occur on a carbon steel surface. While corrosion cannot be stopped it can be retarded. Many of the traditional anti-corrosion approaches such as the chromate process are effective but hazardous to the environment and human health. Mefp-1, a protein derived from blue mussel byssus, is well known for its extraordinary adhesion and film forming properties. Moreover, it has been reported that Mefp-1 confers a certain corrosion protection for stainless steel. All these facts indicate that this protein may be developed into corrosion inhibitors with ‘green’, ‘effective’ and ‘smart’ properties. In this study, a range of surface-sensitive techniques have been used to investigate adsorption kinetics, film forming and film compaction mechanisms of Mefp-1. In situ atomic force microscopy (AFM) enables the protein adsorption on substrates to be visualized, whereas the ex situ AFM facilitates the characterization of micro- and nano-structures of the protein films. In situ Peak Force AFM can be used to determine nano-mechanical properties of the surface layers. The quartz crystal microbalance with dissipation monitoring (QCM-D) was used to reveal the build-up of the Mefp-1 film on substrates and measure the viscoelastic properties of the adsorbed film. Analytical techniques and theoretical calculations were applied to gain insights into the formation and compaction processes such as oxidation and complexation of pre-formed Mefp-1 films. The electron probe micro analyzer (EPMA) and X-ray photoelectron spectroscopy (XPS) were utilized to obtain the chemical composition of the surface layer. Electrochemical impedance spectroscopy (EIS) measurements were performed to evaluate the corrosion inhibition efficiency of different forms of Mefp-1 on carbon steel substrates. The results demonstrate that Mefp-1 adsorbs on carbon steel surfaces across a broad pH interval, and it forms a continuous film covering the substrate providing a certain extent of corrosion protection. At a higher pH, the adsorption is faster and the formed film is more compact. At neutral pH, results on the iron substrate suggest an initially fast adsorption, with the molecules oriented preferentially parallel to the surface, followed by a structural change within the film leading to molecules extending towards solution. Both oxidation and complexation of the Mefp-1 can lead to the compaction of the protein films. Addition of Fe3+ induces a transition from an extended and soft protein layer to a denser and stiffer one by enhancing the formation of tri-Fe3+/catechol complexes in the surface film, leading to water removal and film compaction. Exposure to a NaIO4 solution results in the cross-linking of Mefp-1, which also results in a significant compaction of the pre-formed protein film. Mefp-1 is an effective corrosion inhibitor for carbon steel when added to an acidic solution, and the inhibition efficiency increases with time. As a film-forming corrosion inhibitor, the pre-formed Mefp-1 film provides a certain level of corrosion protection for short term applications, and the protection efficiency can be significantly enhanced by the film compaction processes. For the long term applications, a thin film composed of Mefp-1 and ceria nanoparticles was developed. The deposited Mefp-1/ceria composite film contains micro-sized aggregates of Mefp-1/Fe3+ complexes and CeO2 particles. The Mefp-1/ceria film may promote the further oxidation of ferrous oxides, and the corrosion resistance increases with time. Moreover, phosphate ions react with Fe ions released from the surface and form deposits preferentially at the surface defect sites. The deposits incorporate into the Mefp-1/ceria composite film and heal the surface defects, which result in a significantly improved corrosion inhibition effect for the Mefp-1/ceria composite film in both initial and prolonged exposure situations / <p>QC 20130610</p>

The envelope glycoprotein (Env) of Human Immunodeficiency Virus 1 (HIV-1) has developed several immune-evasion mechanisms to avoid the induction of neutralising antibodies, including immunodominant non-neutralising epitopes, conformational flexibility of conserved epitopes, and spontaneous subunit dissociation, thus impeding vaccine development. Here, chemical modification of Env-based vaccine antigens is explored to overcome these obstacles. Firstly, covalent fixation of Env by chemical cross-linking was used to stabilise the conformationally flexible structure and prevent subunit dissociation. Cross-linked Env constructs showed reduced binding of many non-neutralising antibodies whilst largely maintaining antibody recognition by broadly neutralising antibodies. Compared to unmodified material, immunisation with some of these cross-linked proteins led to the induction of significantly increased antibody titres targeting the conserved CD4 binding site of Env despite similar overall antibody titres. These refocused antibody responses resulted in increased serum neutralising titres compared to animals receiving unmodified protein. Secondly, an epitope masking strategy was developed to reduce or eliminate the immunogenicity of neutralisation-irrelevant surfaces. This was achieved using site-selective addition of theoretically immunosilent glycoconjugates to lysine residues. Masking of model protein hen egg lysozyme (HEL) led to site-selective loss of antibody binding to the modification sites in vitro, which translated into refocusing of antibody responses from masked to unmasked epitopes in vivo. Mutant HIV-1 and influenza virus surface glycoproteins were designed that had lysine residues removed from close proximity to the respective broadly neutralising epitopes, but added throughout the remaining surface. Masking of these mutant proteins with second-generation glycoconjugates led to predictable perturbations of antibody binding in vitro. However, administration of these modified glycoproteins revealed unexpectedly that the masking glycans were highly immunogenic in vivo. Thus, this strategy may well prove useful if truly non-immunogenic glycoconjugates can be identified. Taken together, these chemical modifications of vaccine antigens may allow focused targeting of specific antigenic regions for increased B cell recognition, and may thus be a valuable tool for vaccine antigen design.

Marquée par la mise en place et par le développement graduel d’importantes transformations de type socioéconomique et démographique, la deuxième moitié du 19e siècle constitue le scénario à partir duquel nous analysons et comparons le phénomène de la mobilité sociale intergénérationnelle au Québec et en Ontario, plus précisément dans la période 1852-1881. Grâce à la disponibilité de bases de microdonnées censitaires largement représentatives de la population qui habitait dans chacune de ces deux provinces en 1852 et en 1881 ainsi qu’au développement récent d’une technique de jumelage automatique, nous avons réussi à obtenir un échantillon de 4226 individus jumelés entre les recensements canadiens de 1852 et de 1881. Ces individus sont les garçons âgés de 0 à 15 ans en 1852, qui habitaient majoritairement en milieu rural au Québec ou en Ontario et qui se trouvent dans l’échantillon de 20% du recensement canadien de 1852. Cet échantillon jumelé nous a permis d’observer les caractéristiques de la famille d’origine de ces garçons en 1852 – par exemple, le statut socioprofessionnel du père et la fréquentation scolaire – ainsi que leur propre statut socioprofessionnel (en tant qu’adultes) en 1881. Malgré certains défis posés par la disponibilité et le type de données ainsi que par la procédure de jumelage, cet échantillon illustre bien les changements majeurs qui ont eu lieu durant la période étudiée dans le marché du travail, soit le déclin du groupe des cultivateurs au profit des travailleurs non-manuels et des travailleurs manuels (surtout les qualifiés). De plus, cet échantillon nous a permis d’identifier que malgré le déclin du groupe des cultivateurs entre les pères (en 1852) et les fils (en 1881), l’agriculture aurait continué à être importante durant cette période et aurait même été ouverte à des individus ayant des origines socioprofessionnelles ou socioéconomiques différentes, c'est-à-dire, à des fils de non-cultivateurs. Cette importance soutenue et cette ouverture de l’agriculture semble avoir été plus importante en Ontario qu’au Québec, ce qui pourrait être associé aux différences entre les provinces en ce qui a trait aux caractéristiques et au développement du secteur agricole entre 1852 et 1881. / Marked by the gradual development of important socioeconomic and demographic transformations, the second half of the 19th century constitutes the context of our analyses of the intergenerational social mobility in the provinces of Ontario and Quebec, specially focusing on the period 1852-1881. Taking advantage of the availability of census microdata databases – which are to a great extent representative of the population residing in those two provinces in 1852 and in 1881 – as well as of the development of a recent technique of automatic linkage, we used a sample of 4226 individuals who were linked between the Canadian censuses of 1852 and 1881. Those individuals are boys aged 0 to 15 years in 1852, who lived mainly in a rural area in Ontario or in Quebec and who are included in the 20% Canadian census sample of 1852. From this linked sample we could observe the characteristics of the family of origin in 1852 – e.g. the occupational status of the father and the boy’s school attendance – as well as the subject’s own occupational status as an adult in 1881. Despite some challenges posed by the availability and the type of the data as well as by the linking procedure, this sample illustrates quite well the transformations of the labour market that took place during the period of the study, notably the decline of the occupational group of the farmers in favour of the non-manual and manual skilled workers. Nevertheless, despite the decline in the proportion of farmers among sons (in 1881) compared to their fathers (in 1852), using this linked sample we discovered that agriculture continued to play an important role in economic activity and that this sector was even open to individuals with different occupational or socioeconomic backgrounds, i.e. to sons of non-farmer fathers. The sustained importance and openness of the agricultural sector seems to have been more important in Ontario than in Quebec. This difference could be associated with the contrasts between those two provinces regarding the characteristics and the development of the agricultural sector during the second half of the 19th century.

Les protéines extraites des végétaux sont des matériaux relativement peu coûteux, non toxiques, biocompatibles et biodégradables. Elles représentent une bonne alternative aux protéines d’origine animale et aux polymères dérivés du pétrole. Dans le cadre de cette étude, les protéines extraites de graines de soja et de tournesol ont été utilisées en tant que matériaux enrobants pour la microencapsulation de la matière active hydrophobe (α-tocophérol) ou hydrophile (acide ascorbique) par le procédé d’atomisation. Les protéines de soja sont largement utilisées dans les applications alimentaires et non-alimentaires, notamment en microencapsulation. Elles sont donc étudiées dans ce travail comme matériau enrobant de référence. Les protéines de tournesol n’ont quant à elles pas d’application industrielle concrète, si ce n’est sous la forme de tourteaux dans l’alimentation animale. C’est pourquoi il nous semble pertinent de trouver des nouvelles voies de valorisation pour ce coproduit d’origine agricole. Plusieurs modifications des protéines, telles que l’hydrolyse enzymatique, l’acylation, la réticulation enzymatique et la cationisation ont été étudiées dans le but d’améliorer les propriétés encapsulantes du matériau enrobant. Dans le contexte de la chimie verte, toutes les modifications ont été effectuées sans utilisation de solvants organiques ni de catalyseurs chimiques. L’influence des modifications chimiques et enzymatiques des protéines, et des paramètres du procédé (pression d’homogénéisation, ratio matériau enrobant/matière active et concentration en protéines) sur les différentes caractéristiques des préparations liquides et des microparticules (viscosité, taille des gouttelettes dans le cas des émulsions, morphologie et taille des microparticules), ainsi que sur les paramètres liés au procédé d’atomisation (rendement et efficacité de microencapsulation) a été particulièrement étudiée au cours de ce travail. Les résultats obtenus confirment que l’extrait protéique de tournesol est tout à fait pertinent comme matériau enrobant et permet d’obtenir des efficacités de microencapsulation significativement plus élevées par rapport à celles obtenues avec l’extrait protéique de soja. / Proteins extracted from vegetables are relatively low-cost, non-toxic, biocompatible and biodegradable raw materials. They represent a good alternative to animal-based proteins and petroleum-extracted polymers. In this study, proteins derived from soybean and sunflower seeds were used as wall materials for microencapsulation of hydrophobic (-tocopherol) or hydrophilic (ascorbic acid) active material by spray-drying technique. Soybean proteins are widely used in food and non-food applications, especially in microencapsulation. They were studied in this work as wall material of reference. Sunflower proteins are not actually used in industrial application, but only in the form of oil-cake for animal feeding. That’s why new ways of valorization of this agricultural by-product should be investigated. Several proteins’ modifications such as enzymatic hydrolysis, acylation, cross-linking and cationization were studied in order to improve encapsulating properties of wall material. In the context of green chemistry, all the modifications and preparations were performed without use of organic solvents and chemical catalysts. The effect of protein chemical and enzymatic modifications, and process parameters (homogenization pressure, wall/core ratio and protein concentration) on different characteristics of liquid preparations and microparticles (viscosity, emulsion droplet size, microparticle size and morphology) and on parameters related to the spray-drying process (yield and efficiency of microencapsulation) was particularly investigated in this study. The obtained results confirmed that sunflower proteins are quite suitable as encapsulating agent and provide the microencapsulation efficiencies significantly higher compared to those obtained with soy proteins.

[pt] Arnold no seu trabalho The asymptotic Hopf Invariant and its applications de 1986, considerou sobre um domínio (ômega maiúsculo) compacto de R3 com bordo suave e homología trivial campos X e Y de divergência nula e tangentes ao bordo de (ômega maiúsculo) e definiu o índice de enlaçamento assintótico lk(X; Y ) e o invariante de Hopf associados a X e Y pela integral I(X; Y ) igual a (integral em ômega maiúsculo de alfa produto d-beta), onde (d-alfa) igual a iX-vol e (d-beta) igual a iy-vol, e mostrou que I(X; Y ) igual a lk(X; Y ). Agora, no presente trabalho estenderemos estas definições de índices de enlaçamento assintótico lk(fi maiúsculo,xi maiúsculo) e de invariante de Hopf I(fi maiúsculo,xi maiúsculo), onde (fi maiúsculo) e (xi maiúsculo) são ações de Rk e de Rs, k mais s igual a n-1, respectivamente de difeomorfismos que preservam volume em (ômega maiúsculo n) a bola unitária fechada em Rn e mostraremos que lk (fi maiúsculo, xi maiúsculo) igual a I(fi maiúsculo,xi maiúsculo). / [en] V.I. Arnold, in his paper The algebraic Hopf invariant and its applications published in 1986, considered a compact domain (ômega maiúsculo) in R3 with a smooth boundary and trivial homology and two divergence free vector fields X and Y in (ômega maiúsculo) tangent to the boundary. He defined an asymptotic linking invariant lk(X; Y ) and a Hopf invariant associated to X and Y by the integral I(X; Y ) equal (integral em ômega maiúsculo de alfa produto d-beta) where (d-alfa) equal iX-vol e (d-beta) equal iy- vol. He showed that que I(X; Y ) equal lk(X; Y ). In the present work we extend these definitions of the asymptotic linking invariant lk(fi capital letter,xi capital letter) and the Hopf invariant I(fi maiúsculo,xi capital letter) where (fi capital letter) and (xi capital letter) are actions Rk and Rs, k plus s equal n-1 by volume preserving diffeomorphisms, on the closed unit ball (ômega capital letter n) in and we show lk (fi capital letter, xi capital lette r equal I(ficapital letter ,xi capital letter).