This work reviews the topic of international financial linkages, including theoretical definitions and the main methodological approaches of the empirical measurement based on vector autoregressive models. One of the approaches, the Spillover Index methodology based on Diebold & Yilmaz (2009), is then used to analyze the developments of financial linkages of the Japanese stock market in the period from 1995 to 2012. The attention is paid both to the relations with western developed economies and within the region of East Asia. The main contribution of this paper is the fact that it comprises a complete review of international relations of Japanese stock market during the era of unprecedented financial liberalization. The results of the empirical study confirm the opening of Japanese stock markets towards foreign influence. Even though USA have been the major driving force behind the movements in East Asian stock markets, Japan has become a significant regional player, whose influence on East Asian countries has been growing. The developments in the Japanese stock market are on the other hand driven solely by the western developed countries, which further supports the view of Japan as the regional financial leader.

La famille des Fagacées regroupe des espèces présentant un intérêt économique, écologique et social non négligeable. Par ailleurs, ces espèces, et plus particulièrement celles du genre Quercus que l’on retrouve dans des milieux extrêmement diversifiés, constituent de bons modèles d’étude de l’adaptation des arbres à leur environnement. Pour comprendre l’architecture génétique des caractères liés à l’adaptation chez le chêne, des cartes génétiques ont été établies essentiellement sur la base de marqueurs moléculaires dominants. Le travail qui a fait l’objet de cette thèse, a consisté à développer une carte génétique de seconde génération à partir des ressources génomiques disponibles chez cette espèce. Dans un premier temps, nous avons recherché des motifs microsatellites (SSR, simple sequence repeats) au sein des séquences exprimées (EST) assemblées sous la forme d’un unigène de 28 000 éléments non redondant. Un jeu de 748 marqueurs a été développé et 255 d’entre eux ont été localisés sur la carte génétique du chêne pédonculé (Q. robur L.) en utilisant une approche dite de « bin mapping ». Leur transférabilité a été testée chez le châtaignier européen (Castanea sativa Mill.) et le hêtre commun (Fagus sylvatica L.), deux espèces phylogénétiquement proche du chêne. Un taux de transférabilité de 28% a été observé pour le hêtre et de 56,6% pour le châtaigner. Une carte génétique a alors été établie pour le châtaigner en utilisant les marqueurs SSR localisés sur la carte du chêne. La comparaison des cartes de liaison du chêne et du châtaignier a mis en évidence une bonne conservation de la macro synténie et de la macro colinéarité entre les deux espèces, ce qui ouvre des perspectives intéressantes pour le transfert d’informations génétiques (QTL par exemple) d’une espèce à l’autre. Cette étude sera prochainement enrichie par la cartographie de marqueurs orthologues dérivés de polymorphismes ponctuels (SNP), ce qui permettra de comprendre l’évolution conjointe des trois espèces majeures de la famille des Fagacées. / The Fagaceae family comprises species of economic, ecological and social importance. In addition, these species and particularly those belonging to the Quercus genus that are present in very diverse ecological niches, constitute good models to study the adaptation of forest trees to their natural environment. To understand the genetic architecture of adaptive traits in oak, genetic linkage maps have been previously established based on dominant markers. In this thesis, we developed a second generation genetic map using the genomic resources that were available in this species. First, we bioinformatically screened an expressed sequence tags catalog assembled into a 28 000 unigene elements, for simple sequence repeats (SSRs). A set of 748 markers was developed and 255 were localized on the pedunculate oak (Q. robur L.) linkage map using a bin mapping approach. Their transferability was tested in the European chestnut (Castanea sativa Mill.) and common beech (Fagus sylvatica L.), two phylogenetically related species to oak. Transferability rates of 28% and 56.6% were observed for beech and chestnut, respectively. A genetic map was then established for chestnut on the basis of orthologous SSRs already mapped in oak. The comparison between both maps clearly showed that the macro-synteny and the macro-colinearity were conserved across genus, opening interesting perspectives in respect to the transfer of genetic information (eg. QTLs, quantitative trait loci) from one species to another. This study will be soon completed by the mapping of orthologous markers derived from single nucleotide polymorphisms (SNPs). This will made it possible to better understand the evolution of the genome of these three major species of the Fagaceae family.

Le fondaparinux (Arixtra®), anticoagulant de la classe des pentasaccharides de synthèse, est le premier inhibiteur d'origine synthétique, spécifique et indirect du facteur Xa de la coagulation. Il résulte de la synthèse chimique de l'unité pentasaccharidique des héparines, capable de se lier à l'antithrombine, une protéine endogène, inhibitrice de la coagulation. Cependant, son utilisation reste limitée par son administration uniquement possible par voie parentérale.L'objectif de ce travail de thèse est de développer une forme orale du fondaparinux en l'associant à un dérivé squalénique. Le squalène, terpénoïde naturel précurseur de la synthèse du cholestérol, possède une très bonne absorption orale (supérieure à 60 %). Dans ce contexte, deux stratégies d'association ont été développées: la première consistant à associer par liaison covalente le fondaparinux à un dérivé squalénique selon le concept de la « squalénisation » et la deuxième à associer par interactions non covalentes le fondaparinux à un dérivé squalénique cationique.Les travaux expérimentaux ont montré que la première stratégie était délicate à mettre en œuvre en raison d'une part de la difficulté à synthétiser un bioconjugué fondaparinux-squalène et d'autre part de la perte de l'activité anticoagulante du fondaparinux. En raison de ces obstacles, le concept de la « squalénisation » n'est pas adapté à ce type de molécule active. En revanche, la deuxième stratégie s'est montrée très prometteuse. Elle a consisté à formuler des nanoparticules par association non covalente du fondaparinux, chargé négativement, à un dérivé squalénique cationique. Cette approche a permis de mettre en évidence l’excellente capacité d'auto-assemblage en milieu aqueux de ces deux composés, liée à l’établissement de deux types d’interactions, électrostatiques et hydrophobes (entre les molécules de squalène). L'absorption orale du fondaparinux a été considérablement augmentée grâce à ce nouveau système nanoparticulaire. Cette nouvelle approche à base de squalène a ainsi montré son efficacité dans l'amélioration de l'administration orale du fondaparinux et pourrait représenter un système thérapeutique potentiel dans le traitement des maladies thromboemboliques. / Since its introduction in the market in 2002, fondaparinux (Arixtra®) is a drug of choice in the anticoagulant therapy. Its structure corresponds to the heparin pentasaccharide sequence that mediates its interaction with the natural plasma inhibitor of coagulation, antithrombin. However, like heparin, its application is limited due its unique administration by parenteral route. The aim of this project is to develop an efficient oral delivery system for fondaparinux by association with a squalene derivative. Squalene, a natural precursor of cholesterol in sterol biosynthesis, is well-known for its excellent oral absorption (i.e. more than 60 %). In this context, two strategies were investigated. The first consisted in achieving a covalent coupling between fondaparinux and a squalene derivative according to the concept of “squalenoylation”. The second was to associate fondaparinux to a cationic squalenoyl derivative by non-covalent association.Experimental work showed that the first strategy was delicate to implement due to the difficulty to synthesize a fondaparinux-squalene bioconjugate and, the loss of the anticoagulant properties of fondaparinux. Because of these obstacles, the concept of "squalenoylation" was not suitable for this type of active molecule. In contrast, the second strategy has been very promising. It consisted in the formulation of a nanoparticulate delivery system by ion-pairing of fondaparinux and a cationic squalenoyl derivative. This approach permitted to highlight the self-assembly of these two compounds in water as monodisperse nanoparticles thanks to electrostatic and hydrophobic interactions. Furthermore, the oral absorption of fondaparinux was significantly increased with this new nanoparticulate system. This new squalene-based approach has shown its effectiveness in improving the oral administration of fondaparinux and could be a potential delivery system in the treatment of thromboembolic diseases.

A associação não aleatória entre alelos de diferentes lócus caracteriza o que é chamado de desequilíbrio de ligação (DL) entre eles. A extensão do DL nas populações humanas pode ser influenciada por muitos fatores, tais como taxa de recombinação, características demográficas (idade, tamanho e taxa de crescimento) e fatores evolutivos (deriva genética, efeito fundador, gargalos populacionais, mutação, seleção e fluxo gênico). Portanto, o conhecimento dos padrões do DL fornecem dados que auxiliam na descrição dos eventos demográficos e evolutivos sofridos pelas populações. O objetivo deste estudo foi descrever os padrões de DL de quatro populações brasileiras e correlacioná-los com suas respectivas histórias demográficas, uma vez que estas populações experimentaram alguns dos eventos evolutivos que geram ou retardam o decréscimo do DL, como fundação por poucos indivíduos, miscigenação no momento da fundação e posterior isolamento. Foram analisadas amostras de três populações remanescentes de quilombos do Estado do Piauí, Gaucinha (GAU, n = 14), Mimbó (MIB, n = 15) e Sítio Velho (STV, n = 15) e da população urbana de Teresina, Piauí (TES, n = 15), além de sete amostras populacionais do projeto HapMap (CEU, CHB, JPT, ASW, LWK, MKK, YRI, todas com n = 15). Foram genotipados mais de 250 mil SNPs (Single Nucleotide Polymorphisms) utilizando-se o GeneChip® Human Mapping 250K Nsp I Array - Affymetrix® nas amostras das quatro populações brasileiras. Os dados brutos das populações do HapMap para este array foram obtidos na página do projeto. Os genótipos para todas as amostras foram determinados pelo algoritmo CRLMM após comparação com o algoritmo BRLMM, e as análises de DL e determinação dos blocos de haplótipos foram realizadas com o uso do programa Haploview. Considerando-se o número de blocos de haplótipos detectados em cada população estudada, padrão semelhante foi observado em todos os autossomos. Em geral, a população europeia (CEU) e as duas populações asiáticas (CHB e JPT) do HapMap apresentaram os maiores números de blocos, enquanto que os menores números foram observados nos quilombos GAU e MIB e na população TES. As populações africanas LWK, MKK e YRI e a população afro-americana ASW apresentaram os valores intermediários e a população afro-brasileira STV, apresentou um número de blocos apenas inferior a CEU, CHB e JPT. A grande contribuição africana nos quilombos GAU e MIB pode explicar o menor DL observado nestas comunidades. Por outro lado, o menor DL em TES se deve, provavelmente, à sua fundação, que envolveu um maior número de indivíduos e foi seguida por um rápido crescimento. A possível explicação para o maior DL observado em STV, em relação aos demais quilombos, consiste em sua peculiar história demográfica: esta comunidade experimentou uma miscigenação no momento de sua fundação, que foi seguida por um crescimento lento e pouca diferenciação. Assim, foi demonstrado como os eventos demográficos de cada população influenciam seus respectivos padrões de DL. / The non-random association between alleles of different loci characterizes what is called linkage disequilibrium (LD) between them. The LD extent in human populations can be influenced by many factors, such as recombination rate, demographic features (age, size and growth rate) and evolutionary events (genetic drift, founder effects, population bottlenecks, mutation, selection and gene flow). Therefore, knowledge of the LD patterns provides data that assists in describing the evolutionary and demographic events experienced by populations. The aim of this study was to describe the LD patterns of four Brazilian populations and correlate these patterns with their respective demographic histories, since these populations have experienced some of the evolutionary events that produce or retard the LD decrease, such as foundation by few individuals, admixture at the founding moment and subsequent isolation. Samples from three quilombo remnants populations of the Piauí State, Gaucinha (GAU, n = 14), Mimbó (MIB, n = 15) and Sítio Velho (STV, n = 15) and the urban population of Teresina, Piauí (TES, n = 15), and seven population samples from the HapMap Project (CEU, CHB, JPT, ASW, LWK, MKK, YRI, all with n = 15) were analyzed. More than 250 thousand SNPs (Single Nucleotide Polymorphisms) were genotyped using the GeneChip ® Human Mapping 250K Nsp Array I - Affymetrix ® in the samples of the four Brazilian populations. Raw data of the HapMap population samples for this array were obtained from the HapMap homepage. Genotypes for all samples were determined by CRLMM algorithm after comparison with the BRLMM algorithm. LD analyzes and determination of haplotype blocks were performed using the Haploview software. Considering the number of haplotype blocks detected in each population, a consistent pattern was observed for all autosomes. The European population (CEU) and the two Asian populations (CHB and JPT) of the HapMap showed the highest numbers of blocks, while the lowest numbers were observed in the GAU and MIB quilombos and in the TES population. The African populations, LWK, MKK and YRI, and the African-American ASW exhibited intermediate values and the African-Brazilian population STV, presented a number of blocks smaller than that observed for CEU, CHB and JPT. The great African contribution in the GAU and MIB quilombos may explain the lower LD observed in these communities. On the other hand, the lower LD in TES is probably due to its foundation that involved a larger number of individuals and was followed by a fast growth. A possible explanation for the higher LD observed in STV, compared to other quilombos, consists in its particular demographic history: this community experienced admixture at the time of its foundation, which was followed by slow growth and low differentiation. Thus, it was shown how the demographic events of each population influence their respective LD patterns.

No Brasil e internacionalmente, a mortalidade infantil é conhecido problema de saúde pública. Embora suas taxas estejam diminuindo, a maior parte desses óbitos ocorre por causas evitáveis, sendo sua redução uma questão de justiça social. Este estudo trata do risco de morte infantil evitável nos últimos seis anos e foi realizado em uma coorte retrospectiva de recém nascidos de mães residentes no município gaúcho de Novo Hamburgo de 2008 a 2013, por meio da vinculação dos bancos de dados do Sistema de Informações sobre Nascidos Vivos (SINASC), Sistema de Informações sobre Mortalidade (SIM) e dos dados das investigações de óbitos infantis dos residentes naquele município. O município onde se deu o estudo obteve uma taxa de mortalidade infantil de 14,0 por mil nascidos vivos para o período, sendo em sua totalidade, investigados; desta taxa de óbitos, 10,4 ocorreram por causas evitáveis. Houve predomínio de ocorrência no período neonatal precoce. Conforme a classificação de evitabilidade SUS, os óbitos seriam reduzíveis principalmente por ações adequadas de promoção à saúde vinculadas a ações adequadas de atenção à saúde, por adequada atenção à mulher na gestação e por ações adequadas de diagnóstico e tratamento. Este estudo identificou grupos expostos a maior risco de morte infantil evitável, onde os seguintes fatores mostraram associação com a mortalidade infantil evitável: Condições socioeconômicas e reprodutivas maternas, nas variáveis raça/cor, escolaridade, situação conjugal, situação reprodutiva e ausência de trabalho remunerado; Pré-natal e nascimento, onde o número de consultas e o local de nascimento e finalmente, a Saúde da criança ao nascimento, onde a idade gestacional, peso ao nascimento, vitalidade do recém-nascido, malformação congênita e gemelaridade, foram preponderantes; todos com χ2; p ≤ 0,01, denotando associação estatística significativa com o óbito infantil. Conferir visibilidade aos determinantes da mortalidade infantil é essencial à tomada de decisões e o conhecimento produzido será potencialmente útil para informar políticas e programas voltados à redução da mortalidade e à promoção da saúde populacional, especialmente a infantil. / Infant mortality is a known public health problem in Brazil and worldwide. Although its rates are decreasing, most of these deaths could be avoided, and its reduction is a matter of social justice. This study talks about the risk of avoidable infant mortality in the last six years in a retrospective cohort of infants from mother who lived in the city of Novo Hamburgo, Rio Grande do Sul, Brazil, from 2008 to 2013. Government data regarding births and infant mortality were analyzed. The city where the study was conducted presented an infant mortality rate of 14.0 every 1000 infants born alive for the period, and mostly occurring in the early neonatal period. They were all investigated, and 10.4 every 1000 infants born alive were avoidable. According to SUS's classification of avoidability, these rates could be lowered mainly with appropriate actions to promote health, linked to appropriate actions of health care by providing specialized care to women during pregnancy and adequate measures for the diagnosis and treatments. This study identified high-risk groups. The main factors associated with infant mortality were: Socio-economical and reproductive conditions of the mother, race/color of the skin, schooling, marital status, absence of paid work; Prenatal and birth, with the number of queries and the birthplace, and, finally, the Health of the child at the moment of birth, where gestational age, birth weight, vitality of the newborn, congenital malformation and twin pregnancy were the most important, all at χ2; p ≤ 0,004, a meaningful statistical association. It is important that the circumstances that determine infant mortality be made visible. This is essential so that the necessary measures will be taken to control it. The knowledge produced here will be potentially useful in programs aimed at improving the quality of health care and decreasing infant mortality rates.

Os caracteres quantitativos normalmente têm elevada importância agronômica e econômica, sendo geralmente os mais importantes nos programas de melhoramento das mais diversas espécies, como é o caso do milho (Zea mays L.). Dentre os vários caracteres considerados, destaca-se a produção de grãos e seus componentes. Dessa forma, o objeto de estudo do presente trabalho foi mapear QTL´s relacionados à vários caracteres de importância agronômica, estimar seus efeitos genéticos e entender as causas da correlação genética (pleiotropia ou ligação), em uma população de milho tropical. Para tanto, foi utilizada uma população com 400 progênies F2:3, foram avaliadas em quatro delineamento látice 10 x 10 em cinco ambientes. Os métodos de mapeamento utilizados foram o Mapeamento por Intervalo Composto (CIM), de forma univariada e multivariada considerando múltiplos caracteres (mCIM). O mapa de ligação previamente construído possui 117 locos marcadores microssatélites, com distância média de 14 cM entre eles em média. Os caracteres avaliados foram: produção de grãos (PG), peso da espiga (PE), prolificidade (PROL), número de espigas (NE), número de ramificações do pendão (NRP), rendimento (REND); altura de planta (AP), altura da espiga (AE), comprimento da espiga (CE), diâmetro da espiga (DE), número de fileiras da espiga (NFI), número de grãos por fileira (NGFI), número de folhas acima da primeira espiga (NFO), posição relativa da espiga (PR), porcentagem de acamamento de plantas (ACP) e porcentagem de quebramento do colmo (QUE). Em geral poucos QTL´s foram mapeados devida à alta interação G x A, e esses resultados foram consistentes com os apresentados na literatura. Usando o mCIM, foi possível separar QTL´s ligados de QTL´s com efeito pleiotrópico, permitindo melhor entendimento das causas genéticas da correlação. De forma geral, caracteres mais correlacionados como PG e AP tiveram predomínio de QTL´s pleiotrópicos, enquanto que caracteres menos correlacionados (como por exemplo, CE e NGFI) tiveram QTL´s segregando de forma independente ou com ligação entre si, ou seja, com baixa presença de efeitos pleiotrópicos. Caracteres correlacionados negativamente com os demais em geral apresentaram efeitos aditivos com sinais opostos aos dos demais caracteres. Dessa forma, foi possível identificar regiões que podem ser manipuladas para realizar seleção assistida de forma mais eficiente. De forma geral, foi difícil localizar QTL´s de grande efeito, principalmente com uso do mCIM, dada a presença de elevada interação entre genótipos e ambientes, que fez que que apenas os QTL´s mais estáveis fossem mapeados. / Quantitative traits normally are the most important ones in plant breeding programs for several species, such as maize (Zea mays L.). Several traits are commonly evaluated and grain yield and its components are normally the major focus of selection. The objective of this study was to map QTL related to the several traits of agronomic importance, estimating their genetic effects and genomic locations, aiming to understand the genetic causes of correlation (pleiotropy or linkage) in tropical maize population. A population with 400 F 2:3 inbred lines was used and the progenies were evaluated in five different environments in Piracicaba, São Paulo, Brazil. QTL were mapped using Composite Interval Mapping (CIM) for several traits in a univariate way, and also using an extension of CIM allowing QTL mapping for several traits simultaneously (multivariate CIM, or mCIM). The genetic map was previously estimated and had 117 microsatelite loci, with average distance of 14 cM between them. The traits considered were: grain yield (GY), ear weight (EW), prolificacy (PROL), ear number (EN), tassel branch number (TBN), plant height (PH), ear height (EH), ear length (EL), ear diameter (ED), kernel row number (KRN), kernels per row (KR), leaf number (LN), ear position (EP) and stalk lodging (SL). In general, few stable QTL were mapped due to high G x E interaction, and the results were consistent with previous ones reported on the literature. Using mCIM, it was possible to separate linked QTL from QTL with pleiotropic effects, allowing a better understand of the genetic causes of the correlation. In general, traits with higher correlation such as GY and PH tend to have more pleiotropic QTL than low correlated traits, such as EL and KR, which have some linked QTL. Negative correlated traits in general had QTL with additive effects with opposite signs. Based on the results, it was possible to identify regions that can be manipulated to do marker assisted selection in a more efficient way, combining QTL alleles in order to build favorable genotypes.

Les études longitudinales sur une longue période permettent d’échantillonner plusieurs fois le phénomène étudié et ainsi, avec des mesures répétées, dégager une tendance confirmée. Mais, dès lors, elles produisent de très larges bases de données épidémiologiques accompagnées de plus de sources de bruit par rapport aux études à observation unique ; et souvent, contiennent de la corrélation dans les mesures. Ici, nous avons présenté à travers cette thèse une étude de long terme des facteurs épidémiologiques et génétiques du paludisme menée dans deux cohortes familiales du Sénégal, l’une dans le village de Dielmo suivi pendant 19 années consécutives (1990 – 2008) et l’autre dans le village de Ndiop suivi pendant 16 années consécutives (1993 – 2008). L’objectif de ce travail de thèse a été de développer des méthodes d’analyse statistique pour identifier des gênes de susceptibilité / résistance au paludisme prenant en compte les relations familiales, les mesures répétées et des potentielles interactions génotypes – environnement dans l’évaluation des phénotypes. Par la suite, de tels phénotypes corrigés des facteurs identifiés comme potentielles sources de confusion et/ou de bruit ont été alors utilisés pour les tests de liaison et d’association génétique. Le phénotype principal étudié chez chaque volontaire a été la survenue ou non d’accès palustre, attribué à une infection au parasite Plasmodium falciparum, durant chaque trimestre de présence (PFA). Les études ont été menées de manière indépendante dans chacun des deux villages, de même que les analyses descriptives, l’estimation de la contribution génétique humaine et des effets individuels. Les tests de liaison et d’association génétique ont été réalisés par des méthodes familiales basées sur l’analyse de la transmission d’allèles des parents aux enfants (Transmission Disequilibrium Test). Ces méthodes sont connues pour être robustes par rapport au problème de la stratification de population et donc nous permettent d’augmenter la taille de notre échantillon dans les études de liaison et d’association génétique en analysant les deux villages en même temps. / Long term longitudinal surveys have the advantage to enable several sampling of the studied phenomena and then, with the repeated measures obtained, find a confirmed tendency. However, these long term surveys generate large epidemiological datasets including more sources of noise than normal datasets (e.g. one single measure per observation unit) and potential correlation in the measured values. Here, we studied data from a long-term epidemiological and genetic survey of malaria disease in two family-based cohorts in Senegal, followed for 19 years (1990–2008) in Dielmo and for 16 years (1993–2008) in Ndiop. The main objectives of this work were to take into account familial relationships, repeated measures as well as effect of covariates to measure both environmental and host genetic (heritability) impacts on the outcome of infection with the malaria parasite Plasmodium falciparum, and then use findings from such analyses for linkage and association studies. The outcome of interest was the occurrence of a P. falciparum malaria attack during each trimester (PFA). The two villages were studied independently; epidemiological analyses, estimation of heritability and individual effects were then performed in each village separately. Linkage and association analyses used family-based methods (based on the original Transmission Disequilibrium Test) known to be immune from population stratification problems. Then to increase sample size for linkage and association analyses, data from the two villages were used together.

Two organophosphorus compounds, namely diethyloxymethyl-9-oxa-10- phosphaphenanthrene-10-oxide (DOPAC) and diethyl (2,2-diethoxyethyl) phosphonate (DPAC) were applied on cotton cellulose to impart non-carcinogenic and durable (in alkaline washing) flame retardant property to it. Some acidic catalysts, sodium dihydrogen phosphate (NaH2PO4), ammonium dihydrogen phosphate (NH4H2PO4) and phosphoric acid (H3PO4), were successfully used to settle acetal linkage between cellulose and flame retardant (FR) compound. Appreciable limiting oxygen index (LOI) values of 24% and 23.9% were achieved in case of the samples treated with FR compound DPAC along with the combined acidic catalyzing effect of NaH2PO4+H3PO4 and NaH2PO4+NH4H2PO4. A distinguishing outcome of total heat of combustion (THC) 3.27 KJ/g was revealed during pyrolysis combustion flow calorimetry (PCFC) test of the treated sample. In respect of thermal degradation, low temperature dehydration in conjugation with sufficient amount of char residue (30.5%) was obtained in case of DOPAC treated sample. Consistently, the temperature of peak heat release rate (TPHRR) (325°C) of DPAC treated sample supported the expected low temperature pyrolysis in condensed phase mechanism. Subsequent thermogravimetric analysis (TGA) also reported inspiring weight retention% of the treated samples. Furthermore, for both of the flame retardant compounds, effect of different catalysts, considering both individual and combined, effect of solvents, and overall the optimization of the process parameters were studied in detail. / Program: Magisterutbildning i textilteknologi

Na atual conjuntura, a competitividade das empresas passou a ser condição sine qua non para a permanência no mercado. Ademais, a competitividade deixou de depender da empresa por si só e passou a ser resultado da competitividade de toda a cadeia de valor em que a empresa está inserida. Neste contexto, surge no Brasil o Programa Nacional de Encadeamento Produtivo que visa, dentre outros objetivos, desenvolver a cadeia de valor de empresas para aumentar a competitividade de pequenos, médios e grandes negócios. O Programa Nacional de Encadeamento Produtivo atua em cinco focos estratégicos, sendo um deles o foco em Rede de Aprendizagem. O foco Rede de Aprendizagem, por fortalecer as ligações e interações entre os membros de uma cadeia de valor, ajuda no desenvolvimento dos outros focos estratégicos do programa. Possui, assim, importância singular em todo o processo de encadeamento produtivo. Mesmo com tamanha importância, as redes ainda estão aquém do potencial de desenvolvimento que elas podem gerar. Deste modo, o trabalho teve como objetivo geral propor referências para auxiliar o desenvolvimento de redes organizacionais. Para tanto, realizou-se pesquisa qualitativa baseada em entrevistas em profundidade. As entrevistas foram realizadas em oito redes organizacionais. Com base nas boas práticas existentes, nas diferentes formas de atuação, na cultura de cooperação entre os membros das redes e no impacto efetivo que cada uma das redes visitadas produz nas empresas, foram propostas referências para nortear o desenvolvimento das redes de aprendizagem. Os principais resultados foram referências sobre estrutura de gestão, processos internos, funções dos membros, métricas de desenvolvimento, dentre outros norteadores de atuação. / Nowadays, the competitiveness of enterprises has become an indispensable condition for surviving in the market. Furthermore, competitiveness no longer depends on the company itself, but rather became a result of the competitiveness of the whole value chain. In this context, the National Program for Productive Linkage was established in Brazil. This program aims, among other objectives, at developing the value chain of companies in order to increase the competitiveness of small, medium and large businesses. The National Program for Productive Linkage act on five strategic foci, one of them being the focus on Learning Network. The Learning Network focus, by strengthening the connections and interactions among members of a chain value, enhances the development of the other strategic foci of the program and, thus, has a singular importance in the whole process of the productive linkage. Despite being so important, the networks are still below the potential for development that they can generate. Therefore, the main objective of this study was to propose references to support the development of entrepreneurial Learning Networks. To this end, a qualitative research was done, based on in-depth interviews. The interviews were conducted in eight organizational networks. Based on: (i) the existing good practices; (ii) the various ways of working; (iii) the cooperational culture among their members; and (iv) the effective impacts that Organizational Networks produce in the enterprises; references have been proposed to guide the development of Learning Network. The main results were references to management structure, internal processes, functions of members, development metrics, among other guides for better performance.

Sorgo sacarino e cana-de-açúcar são duas importantes gramíneas com fins potencialmente bioenergéticos. No entanto, apesar do conhecido relacionamento evolutivo, os genomas dessas espécies diferem em complexidade e tamanho. O sorgo, Sorghum bicolor, é diploide, com número básico de cromossomos igual a dez, os quais totalizam ~ 726 Mb já sequenciadas. Já a cana cultivada, Saccharum × officinarum, é um autopoliploide com frequente aneuploidia, e apresenta genoma monoploide estimado em ~ 1 Gb. Provavelmente, decorre deste fato, e dos cruzamentos interespecíficos que originaram as variedades atuais, a relativa dificuldade em se realizar estudos genéticos em cana, e, como consequência, em se incrementar os trabalhos de melhoramento na espécie. Nesse contexto, a possibilidade de integrar estudos de mapeamento entre sorgo e cana torna-se viável dado o emprego de metodologias apropriadas. O presente trabalho objetivou mapear e comparar QTLs para caracteres agro-industriais nos genomas de ambas as espécies, baseando-se no relacionamento evolutivo existente entre elas. Para tanto, foram utilizadas duas populações de mapeamento. A população de sorgo sacarino foi constituída por 223 RILs genotipadas por mais de cem mil marcadores baseados em GBS fisicamente mapeados contra o genoma da espécie. A população de cana-de-açúcar constituiu-se de uma progênie F1 segregante com 153 indivíduos genotipados por 500 marcadores baseados em géis (SSR e TRAP) e 7.049 marcadores baseados em GBS, segregando em dose única. Esses marcadores possibilitaram a construção de um mapa genético informativo e saturado, com 993 marcadores distribuídos ao longo de 223 grupos de ligação, totalizando 3.682,05 cM. Ambas as populações foram avaliadas para quatro caracteres de interesse bioenergético: altura de colmos, toneladas de colmos ou de massa verde por hectare, e porcentagens de pol de caldo e de fibra. Modelos mistos foram utilizados para a análise dos dados fenotípicos, evidenciando a existência de interação genótipo-ambiente a partir da estruturação de matrizes de variâncias-covariâncias genéticas. As médias ajustadas conjunta e marginalmente foram utilizadas na descoberta de QTLs. Para este fim, modelos de mapeamento de múltiplos intervalos uni- e multivariados foram utilizados e determinaram a descoberta de 53 e 36 regiões contendo QTLs para as populações de sorgo e cana, respectivamente, para o conjunto dos quatro caracteres. Os genomas foram comparados utilizando os marcadores baseados em GBS de cana com informação posicional em relação ao genoma do sorgo. Um total de 16 regiões sintênicas identificadas entre as espécies possibilitaram inferências a respeito do controle evolutivamente conservado dos caracteres relacionados. Mais oito regiões foram adicionadas a estas após análise de marcadores individualmente para a população de cana. A descoberta dessas regiões subjacente à variação de caracteres bioenergéticos sugere aplicações na clonagem de genes e na seleção assistida por marcadores, beneficiando os programas de melhoramento de ambas as espécies. / Sweet sorghum and sugarcane are two important grasses for bioenergy purposes. However, despite their known evolutionary relationship, the genomes of these species differ in complexity and size. Sorghum bicolor is a diploid species, with basic chromosome number of ten and ~ 726 Mb completely sequenced, whereas Saccharum × officinarum has a autopolyploid genome with frequent aneuploidy and monoploid size estimated at ~ 1 Gb. Therefore, genetic studies and breeding in sugarcane is challenging. In this context, the possibility of integrating mapping studies between sorghum and sugarcane becomes feasible given the recent development of appropriate methodologies. In this work, we aimed to map and compare QTLs for bioenergy traits in both species. To do this, two mapping populations were used. The population of sorghum consisted of 223 RILs genotyped by more than one hundred thousand GBS-based markers, which were physically mapped against the species genome. The population of sugarcane is an F1 segregating progeny with 153 individuals genotyped by 500 gel-based (SSR and TRAP) and 7,049 GBS-based single-dose markers. These markers allowed the construction of an informative and dense genetic map with 993 markers belonging to 223 linkage groups and spanning 3,682.05 cM. Both populations were evaluated for four bioenergy traits: stalk height, prodution in tons per hectare, and percentages of pol and fiber. Mixed models were used to analyze phenotypic data and showed genotype-by-environment interaction on their genetic variance-covariance structures. Joint and marginal adjusted means were used for QTL discovery. Toward this end, univariate and multivariate multiple interval mapping models were used, and a total of 53 and 36 QTLs were found for sorghum and sugarcane, respectively. Comparison of the genomes were based on GBS markers in sugarcane with relative sorghum chromosome information. A total of 16 syntenic regions were identified between the species, allowing inferences in relation to evolutionary conserved control of the related traits. In addition, eight regions were also identified by considering single marker analyses. The discovery of QTLs underlying such bioenergy traits may suggest further applications in gene cloning and marker assisted selection for both sweet sorghum and sugarcane species.