Bacillary angiomatosis (BA), one of the clinical manifestations resulting from infection with the facultative intracellular bacterium Bartonella henselae, is characterized by angiogenic lesions. Endothelial cells have been identified as host cells for this pathogen and are presumed important for pathogenesis as lesions contain bacteria in direct contact with the endothelium. Lesions also contain infiltrating macrophages, which contribute to the angiogenic process during B. henselae infection by secreting vascular endothelial growth factor (VEGF). While virulence factors have been characterized, and the role for macrophages in B. henselae infection has been established, endothelial cell mediators of angiogenesis have not been well defined. We investigated three potentially important pathways that are triggered by the bacterial interactions with endothelial cells. We examined the ability of endothelial cells to upregulate the chemokines monocyte-macrophage chemoattracta nt protein-1 (MCP-1) and CXCL8 and the mechanism by which B. henselae secreted proteins (BHSP) induce endothelial cell proliferation. We determined that MCP-1 production is upregulated in response to B. henselae infection, which very likely contributes to angiogenic lesion formation by recruiting the VEGF-secreting macrophage. The chemokine CXCL8 is an important mediator of angiogenesis which can cause endothelial cell survival, proliferation, and capillary tube formation. We determined that CXCL8 is secreted from B. henselae-infected cells and contributes to B. henselae-induced angiogenesis in an autocrine manner. We also investigated the role of Ca2+ signaling during B. henselae infection. We determined that BHSP induce a robust intracellular Ca2+ response in HUVEC which originates from intracellular Ca2+ pools. Additionally, endothelial cell proliferation in response to BHSP required Ca2+ activity, indicating a role for intracellular Ca2+ pools during B. henselae-induced angio genesis. Endothelial cell proliferation during B. henselae infection possibly indicates a mechanism by which a pathogen induces proliferation of its host cell in order to propagate its own survival. Numerous factors culminate in the vascular lesions that are characteristic of BA. B. henselae infection represents an important and unique model for pathogen-triggered angiogenesis, and studies into the specific mechanisms of this process may elucidate host cell-pathogen interactions as well as pathways of pathogenic angiogenesis.

La nostra proposta ha profunditzat per primera vegada en l’estudi de les lipoproteïnes d’alta densitat (HDL) en la infecció pel virus de la immunodeficiència humana (VIH), i hem determinat la seva implicació en el curs de la infecció i en el desenvolupament de l’arteriosclerosi amb la finalitat de dissenyar noves estratègies terapèutiques. Els nostres resultats mostren com l’estat inflamatori i oxidant persistent associat a la infecció té un paper central en l’aparició d’alteracions metabòliques i el desenvolupament d’aterosclerosi en aquests pacients. Concretament, utilitzant tècniques cromatogràfiques i proteòmiques, mostrem com la presència anòmala de gammaglobulines, molècules pro-inflamatories i reactants de fase aguda en el plasma d’aquests pacients indueixen la modificació del metabolisme, composició i distribució de les partícules HDL, suggerint que la monitorització de les partícules HDL en aquests pacients no ha d’estar basada únicament en la determinació de la concentració del contingut de colesterol, sinó que ha d’incloure la determinació d’altres paràmetres associats a la qualitat, i per tant, a la funcionalitat, de les partícules HDL. / Our proposal has explored for the first time the effect of human immunodeficiency virus (HIV)- infection on high-density lipoprotein (HDL) particles, and we determined their involvement in the course of infection itself and the development of atherosclerosis in order to design novel therapeutic strategies. Our results show how the chronic oxidant and inflammatory status associated with infection plays a central role in the onset of metabolic disorders and the development of subclinical atherosclerosis in these patients. Concretely, using both, chromatographic and proteomic techniques, we observed how the increased levels of gammaglobulin, pro-inflammatory molecules and acute phase reactants in the plasma of these patients induced metabolism alterations and modifications on the distribution and composition of HDL particles, suggesting that the monitoring of HDL particles in these patients should not be only based on the determination of the concentration of cholesterol content, but must include the determination of other parameters associated with particle quality, and therefore the functionality, of HDL particles

G-CSF is routinely used to treat neutropenia/leukopenia or to increase hematopoietic stem cell generation in bone marrow donors. G-CSF and its receptor, G-CSFR, are produced by various cell types both in the peripheral circulation and within brain. As a consequence, exogenous administration of G-CSF results in a broad spectrum of effects involving hematopoietic, immune and central nervous systems. G-CSF administration in a mouse model of Alzheimer's disease (AD) has revealed both cognitive benefits and disease modifying effects: a) decreased Aβ plaque burden, b) increased microgliosis, c) increased neurogenesis and d) improved performance in radial arm water maze (RAWM). In clinical studies, G-CSF plasma levels were found to be lower in patients with early AD in comparison to healthy age matched controls. A course of G-CSF administration in humans is known to increase levels of circulating hematopoietic stem cells (CD34 cells), monocytes and neutrophils in patients with neutropenia and when administered to patients with AD, there is also a similar increase in absolute monocyte count, CD34 cells and total neutrophils. The extent to which the beneficial effects of G-CSF in AD depend on monocyte infiltration into CNS, compared to direct neurotrophic actions of G-CSF on the CNS, is not known. The overall goal of this study was to investigate and understand the effects of G-CSF in an AD mouse model, but more specifically to distinguish the actions of G-CSF that affect the peripheral monocyte population from the direct actions on CNS. The first approach was to examine in vitro effects of G-CSF within a monocytic cell line (THP-1) and a neuronal cell line (SH-SY5Y). The second approach was to study effects of G-CSF on infiltration of bone marrow-derived cells into the brain by utilizing a chimeric GFP+ APP/PS1 AD mouse model. The third approach was to assess the effects of G-CSF on hippocampal neurogenesis in both a wild-type and AD mouse model. Comparison of the monocytic and neuronal cell lines showed a) G-CSF interacts with its cognate receptor with different binding kinetics and with a greater affinity for the monocyte G-CSFR, b) the number of G-CSF receptors in neurons is greater than in monocytes, and c) the anti-apoptotic response in neurons occurs at lower concentrations of G-CSF than in monocytes. Various concentrations of G-CSF increased proliferation of both the monocytic and neuronal cell line in vitro. G-CSF did not improve migratory properties of the monocytic cell line, either adhesiveness or migration through a membrane. In vivo G-CSF treatment (250μg/kg s.c. qod for 2 ½ weeks) in both the AD chimeric and non-chimeric AD mice resulted in increased microgliosis and decreased amyloid plaque burden in the hippocampus. In the chimeric AD mice, G-CSF treatment did not increase infiltration of GFP+ bone marrow derived cells (BMDC) into brain parenchyma and did not increase adhesion to microvasculature. In the non-chimeric AD mice there was improvement of neurogenesis to non-transgenic levels after G-CSF treatment and an increase in synaptogenesis in the CA1 region of the hippocampus. The effects of G-CSF on the endogeneous microglial population are most likely responsible for the increase in microgliosis, as no significant increase of BMDC infiltration into the brain parenchyma was found in vivo. The enhanced proliferation and improved viability of the neuronal cell line after G-CSF treatment may explain the improvement in neurogenesis and significant increase in synaptogenesis seen in the AD mouse model. The actions of G-CSF on neural stem/progenitor cells to stimulate hippocampal neurogenesis and to enhance resident microglial capacity to decrease amyloid burden are the most likely mechanisms responsible for the behavioral improvement seen in the AD mouse model.

La question du stockage de chaleur est non négligeable dans le contexte actuel. L’une des solutions est l’utilisation de matériaux à changement de phase (MCP). Cependant leurs propriétés thermiques restent inadaptées et il est absolument nécessaire d'utiliser un substrat conducteur afin de maximiser le rendement de ces systèmes.L’objectif de cette thèse est la compréhension et la caractérisation des phénomènes physiques mis en œuvre, et l’optimisation de structures architecturés dans de tels systèmes de stockage de chaleur. Une double démarche a été adoptée à la fois expérimentale et numérique sur des structures d’accueil du MCP relativement simples (ailettes) puis plus complexes (mousses ouvertes). Nous avons pu étudier des paramètres géométriques (longueur, porosité, espacement et épaisseur des ailettes, taille de cellules des mousses) de la structure d'accueil, de son matériau constitutif et de son orientation. Les résultats expérimentaux corroborent les simulations numériques menées ce qui a permis de réaliser une étude plus systématique sur les paramètres analysés et notamment d’identifier dans quel cas il fallait prendre en compte la convection naturelle. Enfin à partir de ces résultats nous avons développé un outil permettant d’optimiser des structures pour un cahier des charges défini. / The problematic of heat storage is important in the present context. One of the solutions is to use phase change materials (PCM). Nevertheless their thermal properties are poors and a conductive substracte must absolutely be used in order to maximise the yield of theses systems.The purposes of this PhD are the physics phenomena implementation understanding and characterization, and the optimization of architectured structures for heat storage systems. A dual approach was adopted both experimental and numerical on simple PCM reception structures (fins) and on more complex ones (open foams). We analyzed influences of geometrical parameters (system lenght and porosity, thickness and space betweens fins, cellfoam size) from reception structure, its constituent material and its orientation. Experimental results support well with numerical simulations. This permits to pursue a more systematical study about analyzed parameters, and notably to identify in which cases natural convection has to be taken into account. Finally, from these results, we developped a tool which permits to optimize architectured structures for a defined bill of specifications.

Submitted by Hiolanda Rêgo (hiolandarego@gmail.com) on 2016-12-21T14:48:29Z No. of bitstreams: 1 Tese_ICS_Eliane Pereira Menezes.pdf: 18598624 bytes, checksum: e2cc868d4bb1be2b815801b4f4db02b8 (MD5) / Made available in DSpace on 2016-12-21T14:48:29Z (GMT). No. of bitstreams: 1 Tese_ICS_Eliane Pereira Menezes.pdf: 18598624 bytes, checksum: e2cc868d4bb1be2b815801b4f4db02b8 (MD5) / Estudos mais recentes mostraram uma complexidade da resposta imune e diferentes tipos celulares estão envolvidos na imunoregulação de diversas doenças, incluindo a leishmaniose. Sendo assim, o tipo de resposta imune do hospedeiro infectado é determinado pelo padrão de citocinas produzidas pelos linfócitos T CD4+. A resposta imune Th1 é principalmente representada pela produção de INF-y, IL-2, TNF- e corresponde a denominada imunidade celular que auxilia a resposta efetora de outros linfócitos T CD4+, T CD8+ citotóxicos e macrófagos. A resposta imune Th2 está relacionada com a produção de IL-4, IL-5 e IL-10 e estimulação de linfócitos B na síntese de anticorpos IgG1, IgG4 e IgE, ativação de eosinófilos e mastócitos. As células T regulatórias suprimem ambos os tipos de resposta através de contato celular e pela produção de citocinas como IL-10 e TGF-b. Diversos estudos têm demonstrado uma influência genética em muitas doenças complexas, usando escaneamento de genoma e estudos de associação. Inicialmente diversos estudos mostraram associação entre doenças infecciosas, incluindo a leishmaniose e HLA. Posteriormente, a observação de polimorfismos genéticos em genes de citocinas, quimiocinas, receptores celulares, fatores de transcrição e diversos produtos que influenciam direta ou indiretamente a resposta imune, gerou diversos estudos associando esses polimorfismos com doenças. Por exemplo, um polimorfismo na posição -308, com mutação de G por A, no gene de TNF- , foi demonstrado que induzia a produção de níveis mais elevados dessa citocina e tem sido associado com várias doenças, incluindo leishmaniose cutânea (LC), mucosa (LM) e visceral (LV). No presente trabalho, avaliamos alguns polimorfismos funcionais em genes que codificam produtos relacionados à resposta imune com o objetivo de investigar se fatores genéticos relacionados com o controle da resposta imune influenciam o curso da infecção por Leishmania. Dois tipos de análises foram realizados: 1) Um estudo tipo caso-controle com objetivo de comparar as freqüências alélicas dos genes de IL- 6, linfotoxina- e MCP-1 em quatro grupos de 60 indivíduos cada: LM, LC, controle de vizinhança sem doença (“aparentemente normal”) (CN) e controle DTH+; 2) Um estudo baseado em família (FBAT) para confirmar as associações encontradas no estudo caso-controle. O estudo caso-controle demonstrou: 1) Alta freqüência do alelo C de IL-6 -174 G/C em LM quando comparado com LC e CN. A análise de FBAT confirmou a associação entre o alelo C e LM tanto no modelo aditivo (z=4.295, p=0.000017) como no modelo dominante (z=4.325, p=0.000015); 2) Uma freqüência alta do alelo G de linfotoxina- +252 G/A em LM quando comparado com LC, não sendo essa associação confirmada pelo estudo de famílias: 3) Uma freqüência alta do alelo G de CCL2/MCP-1 -2518 A/G em LM quando comparado com CN (OR=2.3; p=0.0078; 95% CI 1.3-4.1). O teste de FBAT confirmou a associação do alelo G de MCP-1 com LM no modelo recessivo (z=2.69, p=0.007). Esse estudo demonstrou uma associação entre os polimorfismos de genes que induzem a resposta inflamatória com a LM, sugerindo que eles podem influenciar o curso clínico da leishmaniose. Esta tese inclui 3 artigos originais e um capítulo de livro.

Submitted by Reginaldo Soares de Freitas (reginaldo.freitas@ufv.br) on 2017-06-01T16:11:21Z No. of bitstreams: 1 texto completo.pdf: 487232 bytes, checksum: 4fd53655f2412899670b34bd2a57487e (MD5) / Made available in DSpace on 2017-06-01T16:11:21Z (GMT). No. of bitstreams: 1 texto completo.pdf: 487232 bytes, checksum: 4fd53655f2412899670b34bd2a57487e (MD5) Previous issue date: 2004-08-16 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Este trabalho teve como objetivos: - estudar a sensibilidade ao etileno de duas variedades de gerânio em vaso durante sua fase pós- produção; - investigar a eficiência do 1-metilciclopropeno (1-MCP) na prevenção dos danos causados pelo etileno e determinar seus melhores níveis para a aplicação em vasos de gerânio visando o aumento da vida de prateleira e diminuição da queda de pétalas. O experimento sensibilidade ao etileno exógeno foi instalado em parcelas subdivididas, tendo nas parcelas 5 avaliações, uma antes (0 h) e 4 após o início da exposição ao etileno (3, 24, 48 e 72 h) e nas subparcelas um esquema fatorial 3x2, com 3 concentrações de etileno (0, 10 e 100 μL/L) e 2 variedades, no DIC com 3 repetições e 1 vaso de gerânio por repetição. As plantas, das variedades Pulsar Red (VR) e Pulsar Salmon (VS), foram expostas ao etileno em câmaras herméticas, por 3 h. A queda de pétalas (QP) foi avaliada logo após o tratamento e a cada 24 h até todas as plantas atingirem 100% de QP. No experimento com 1- MCP utilizou-se um esquema de parcelas subdivididas, com 4 tempos de exposição ao 1-MCP nas parcelas (3, 6, 9 e 12 h) e, nas subparcelas, um esquema fatorial 5x2, com 5 concentrações de Ethylbloc® (0; 0,1; 0,5; 1,0 e 1,5 g/m3) e 2 variedades, com 3 repetições e 1 vaso de gerânio por repetição. Após o tratamento com 1-MCP as plantas foram expostas a 1 μL/L de etileno, durante 3 h. Avaliou-se o número de dias para queda de 50% das pétalas (D50) e a vida de prateleira da planta. No primeiro experimento, 3 horas após o início da exposição a 10 μL/L de etileno, observou-se diferença (P < 0,05) entre as duas variedades (VR = 97,3% e VS = 76,0%). Após 24 h, plantas da VR e VS sem exposição ao etileno, apresentaram diferença (P < 0,05) (VR = 76% e VS = 42,7%). A % QP foi afetada significativamente pelo aumento da concentração de etileno, sendo mais precoce em plantas tratadas com o gás. Já no segundo experimento a vida de prateleira teve média maior (P < 0,05) na VR (30,3 dias) que VS (12,7 dias) em 12 h e 1 g/m3 de Ethylbloc®. O número de dias para 50% de queda de pétalas atingiu as maiores médias na VR, sendo diferente da VS, com 3 h e 1 μL/L (7,7 e 4,3 dias, respectivamente). Analisando-se os dados nas equações de regressão, foram obtidos os valores máximos de vida de prateleira para VR (23,98 dias) com 3 h e 0,84 g/m3 de Ethylbloc®. A equação de regressão de D50 para VR tem valores máximos com 3,56 h e 1,5 g/m3 de Ethylbloc® (6,9 dias). D50 da VS não foi influenciado pelos níveis de concentração de 1-MCP e pelo tempo de exposição, apresentando média de 4,6 dias. Com o experimento exposição a etileno exógeno, conclui- se que as duas variedades respondem à presença de 10 μL/L de etileno exógeno, por 3 h, mas a VR é mais sensível que a VS. Com o experimento exposição ao 1-MCP, conclui-se que a melhor combinação entre concentração e tempo de exposição para inibir a queda precoce de pétalas e aumentar a vida de prateleira da VR é 1,0 g/m3 e 3 h, e que para a VS, o 1- MCP não foi eficiente. / The objectives of this dissertation were to study the ethylene sensitivity of two varieties of potted geranium during post-production, to evaluate the efficacy of 1-MCP in preventing ethylene-damages, and the optimum exposure time and concentration of 1-MCP to increase shelf life and overcome petal abscission. The first experiment was installed in split-plot pattern with 5 evaluations after the beginning of ethylene treatment (0, 3, 24, 48 and 72 h) in the main plot and the sub plot with a factorial pattern 3x2 in completely randomized factorial - 3 ethylene concentrations (0, 10 and 100 μL/L) and 2 varieties (Pulsar Red, VR, and Pulsar Salmon, VS), with 3 replicates each. The plants were exposed to ethylene in closed chambers, for 3 h. Petal abscission (PA) was evaluated just after the treatment with 24 h intervals until reached 100% PA. The second experiment had the experimental design in split-plots, with 4 exposure times (3, 6, 9 e 12 h) in the main plot and a factorial pattern 5x2 in completely randomized factorial in the sub plot - 5 Ethylbloc® concentrations (0, 0.1; 0.5; 1.0 e 1.5 g/m3) and 2 varieties (VR e VS), with 3 replicates each. After the 1-MCP treatment, the plants were exposed to 1 μL/L of ethylene, for 3 h. The number of days to 50% petal abscission (D50) and the shelf life (C30) were evaluated. Three hours after beginning the first experiment, 10 μL/L of ethylene caused different (P < 0.05) % PA between the two varieties (VR = 97.3% and VS = 76.0%). After 24 h, plants of VR and VS without exposition to ethylene presented difference (VR = 76% and VS = 42.7%) in petal abscission. The % PA was affected significantly by the increase of the ethylene concentration, being higher in plants exposed to ethylene. In the second experiment, shelf life was higher (P < 0.05) to VR (30.3 days) than VS (12.7 days) with 12 h and 1 g/m3 of Ethylbloc®. D50 reached higher value for VR, being different from VS, with 7.7 e 4.3 days, respectively. Based on data regression analysis, maximum values for C30 in VR (23.98 days) were found with 3 h and 0.84 g/m3 of Ethylbloc. The D50 regression for VR has maximum values with 3.56 h and 1.5 g/m3 of Ethylbloc (6.91 days). D50 of VS was not influenced by the treatment with 1-MCP, and presented average of 4.55 days. It can be concluded that VR is more sensitive to ethylene than VS, but the two varieties respond to the presence of 10 μL/L of ethylene, for 3 h; the best combination between concentration and treatment length to inhibit petal abscission and to increase the shelf life of VR was 1.0 g/m3 for 3 h and the treatment with 1-MCP in VS was not efficient.

Made available in DSpace on 2014-06-12T15:51:29Z (GMT). No. of bitstreams: 2 arquivo2776_1.pdf: 6833514 bytes, checksum: de6f82bea549e57112eae403a110a117 (MD5) license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2010 / O estado nutricional tem influências importantes no desenvolvimento fetal e a alteração na qualidade e/ou quantidade da alimentação materna durante a gravidez tem consequências sobre a saúde posterior da prole, mudando suas respostas aos desafios ambientais e, portanto, sua predisposição a doenças. Ácidos graxos ômega-6 e ômega-9 são importantes nutrientes para o crescimento e desenvolvimento e parecem desempenhar um importante papel na modulação da doença inflamatória cardiovascular e hepática. O objetivo deste estudo foi investigar o efeito de dietas hiperlipidêmicas sobre o endotélio e o fígado de matrizes, bem como, avaliar em matrizes hipercolesterolêmicas o efeito do consumo de ômega-6 e ômega-9 sobre o fígado e o endotélio aórtico da prole. Na primeira etapa, avaliamos, em camundongas Swiss, o efeito de uma dieta hiperlipidêmica, rica em ômega-6 e colesterol, sobre o endotélio e o fígado. Análises bioquímicas mostraram que as dietas experimentais causaram dislipidemia, inflamação e peroxidação lipídica; a análise ultraestrutural, por sua vez, evidenciou que as dietas hiperlipidêmicas promoveram patogênese endotelial e acúmulo de lipídeo hepático. Entretanto, a dieta se mostrou mais hepatotóxica quando suplementada com colesterol. Em uma segunda etapa deste trabalho, com o intuito de se caracterizar a influência da dieta de matrizes sobre o endotélio e o fígado da prole, camundongas da espécie C57BL/6J, deficientes de receptor LDL, foram submetidas a uma dieta rica em ômega-6 ou ômega-9 durante 45 dias antes do acasalamento até o nascimento da prole. Matrizes alimentadas com dieta rica em ômega-6 apresentaram aumento sérico de colesterol total (CT) e seus filhotes, por sua vez, apresentaram elevação sérica de CT, triglicérides (TG) e proteína quimiotática de monócito-1 (MCP-1) circulante e diminuição de lipoproteína de alta densidade (HDL). Por outro lado, filhotes de matrizes alimentadas com dieta rica em ômega-9 apresentaram diminuição de TG e aumento de lipoproteína de baixa densidade (LDL). A análise morfológica do fígado de filhotes de matrizes alimentadas com dieta rica em ômega-6 mostrou esteatose, infiltrado leucocitário e aumento da expressão de MCP-1, enquanto que na análise ultraestrutural foram observadas gotículas de lipídeos e miofibroblastos. Tais alterações não foram observadas nos filhotes de matrizes alimentadas com ômega-9. Em análises morfológicas da aorta ascendente, filhotes de matrizes alimentadas com dieta rica em ômega-6 apresentaram um aumento da expressão de VCAM-1 e MCP-1 e alterações ultraestruturais severas como lâmina elástica descontínua, desprendimento de células endoteliais e presença de mitocôndrias degeneradas. De forma semelhante, estas alterações não foram observadas nos filhotes de matrizes alimentadas com dieta rica em ômega-9. Nosso estudo sugere que dietas hiperlipidêmicas promovem dano endotelial e hepático em matrizes. E que matrizes hipercolesterolêmicas alimentadas com dieta rica em ômega-6 predispõem seus filhotes à disfunção endotelial e hepática. Por outro lado, uma dieta rica em ômega-9 tem um efeito não lesivo

The importance of interactions between transmembrane domains of integral membrane proteins has been well-established in a range of essential cellular functions. Most integral membrane proteins also possess regions that lie on the exterior of the membrane that may influence the ability of these transmembrane domains to interact. We sought to test this hypothesis by quantifying the energetics of transmembrane helix self-association in the absence and presence of an amphipathic helix that can bind to the membrane surface. The model chosen for this study was the major coat protein (MCP) of M13 bacteriophage, which has an N-terminal amphipathic helix linked to its single transmembrane segment via a flexible linker. Dimerization of both full-length MCP and a peptide containing only the transmembrane domain (MCPTM) was studied by solution NMR in SDS micelles. We found that there was an increase in the apparent dimerization affinity in the absence of the N-terminal helix. However, this increase in apparent affinity could be attributed to differences in detergent-binding properties of the two polypeptides in monomeric versus dimeric states when the empty micelle was considered to be a participant in the dimer dissociation. Preliminary results from the integral membrane protein, p7 of the hepatitis C virus are also presented in this thesis. It has been demonstrated that p7 enhances viral infectivity and accumulation, and that this function may require oligomerization in the membrane. While we encountered limitations due to challenges in the generation of sufficient quantities of pure p7 samples, we were able to perform circular dichroism spectroscopy under conditions that may favor different oligomeric states. These studies suggest that there is a change in the degree of helicity upon oligomerization, and suggest that SDS could be a suitable system to characterize the interactions of the p7 oligomer in the future.

A family of concave penalties, including the smoothly clipped absolute deviation (SCAD) and minimax concave penalties (MCP), has been shown to have attractive properties in variable selection. The computation of concave penalized solutions, however, is a difficult task. We propose a majorization minimization by coordinate descent (MMCD) algorithm to compute the solutions of concave penalized generalized linear models (GLM). In contrast to the existing algorithms that uses local quadratic or local linear approximation of the penalty, the MMCD majorizes the negative log-likelihood by a quadratic loss, but does not use any approximation to the penalty. This strategy avoids the computation of scaling factors in iterative steps, hence improves the efficiency of coordinate descent. Under certain regularity conditions, we establish the theoretical convergence property of the MMCD algorithm. We implement this algorithm in a penalized logistic regression model using the SCAD and MCP penalties. Simulation studies and a data example demonstrate that the MMCD works sufficiently fast for the penalized logistic regression in high-dimensional settings where the number of covariates is much larger than the sample size. Grouping structure among predictors exists in many regression applications. We first propose an l2 grouped concave penalty to incorporate such group information in a regression model. The l2 grouped concave penalty performs group selection and includes group Lasso as a special case. An efficient algorithm is developed and its theoretical convergence property is established under certain regularity conditions. The group selection property of the l2 grouped concave penalty is desirable in some applications; while in other applications selection at both group and individual levels is needed. Hence, we propose an l1 grouped concave penalty for variable selection at both individual and group levels. An efficient algorithm is also developed for the l1 grouped concave penalty. Simulation studies are performed to evaluate the finite-sample performance of the two grouped concave selection methods. The new grouped penalties are also used in analyzing two motivation datasets. The results from both the simulation and real data analyses demonstrate certain benefits of using grouped penalties. Therefore, the proposed concave group penalties are valuable alternatives to the standard concave penalties.

During combustion of biomass, particulate matter is emitted, which has severe health impacts on humans. The company ITK Envifront has developed a scrubber technology that cleans the flue gas while also recovering the flue gas energy, increasing the efficiency of the combustion plant. In this thesis, a simulation model was built in MATLAB according to the Finite Element Method. Validation of the model against 3 different facilities showed reasonable accuracy with a tendency to overestimate the scrubber heat recovery and a mean prediction deviation of approximately 7 %. The model was then used to make suggestions for process optimization. An increase of funnel height, and number of spray nozzles could increase the scrubbers heat recovery with up to 7 % and 8 %, respectively. Addition of moisture to the flue gas through evaporation of water droplets had the potential to increase scrubber efficiency with 10 %, and usage of the highest setting of the adjustable nozzle bank showed the potential to increase the efficiency with up to 5 % compared to the mid-setting. Furthermore, the process parameters of a scrubber with optimized running conditions, was compared to a scrubber with the current running conditions, through running of the developed model. The optimized running conditions showed an increase in scrubber efficiency with up to 14 %, resulting in an increase in scrubber heat recovery of approx. 90 kW at a boiler load of 3 MW. As a final conclusion, the developed model shows great potential to be used to as a toolbox to further investigate and optimize the scrubber design and operation. As a future work, it would be interesting to further model its performance regarding particle removal.