O desenvolvimento de vacinas contra a raiva tão eficientes quanto as atuais ainda é considerado importante para a profilaxia dessa doença devido ao alto número de mortes por ano no mundo. Este trabalho mostra dois sistemas para a expressão recombinante do principal antígeno da raiva, a glicoproteína viral (RVGP): células Schneider 2 de Drosophila melanogaster estavelmente transfectadas (S2 rRVGP) e vírus Semliki Forest (SFV) carregando o RNA da RVGP (SFVRVGP). Ensaios de purificação por cromatografia de afinidade, da rRVGP de S2 rRVGP produzida em biorreator, demonstraram resultados promissores para o isolamento de monômeros da rRVGP. Para os estudos de imunogenicidade, camundongos foram vacinados com rRVGP de S2 rRVGP e SFV-RVGP. Dosagem de anticorpos anti-glicoproteína do vírus da raiva, neutralizantes, IgG1 e IgG2a, e citocinas demonstraram que o SFV-RVGP induziu predominantemente uma resposta imune do tipo celular e que os dois vetores foram capazes de expressar uma rRVGP imunogênica, apresentando um potencial uso clínico (veterinário e humano). / The development of new and equally efficient rabies vaccines is still considered important to the prophylaxis of the disease, which is responsible for many deaths per year worldwide. This work used two systems for recombinant expression of the major rabies antigen, the viral glycoprotein (RVGP): stably transfected Drosophila melanogaster Schneider 2 cells (S2 rRVGP) and Semliki Forest Virus (SFV) carrying the RNA of RVGP (SFV-RVGP). Purification assays of the rRVGP by metal ion affinity chromatography, from S2 rRVGP cells produced in bioreactor, showed promising results for rRVGP monomers isolation. Analysis of antibodies anti-rabies virus glycoprotein, neutralizing, IgG1 and IgG2a, and citokines showed that the SFV-RVGP induced predominantly a cellular immune response, and both vectors were capable of expressing an immunogenic rRVGP, what reinforces potential clinical applications (veterinary or human).

A glicoproteína do vírus rábico (RVGP) é o antígeno capaz de induzir a formação de anticorpos neutralizantes e a resposta imune protetora contra a infecção pelo vírus rábico. Estudamos as cinéticas de expressão da RVGP e de seu RNA mensageiro (RVGPmRNA) em dois sistemas distintos de expressão recombinante: células de Drosophila melanogaster (S2) e células BHK-21 infectadas com vírus Semliki Forest Virus (SFV). Para isso, fizemos um trabalho de padronização do tratamento de amostras de cultivos celulares, adequando-as a um teste de ELISA para dosagem da RVGP e estabelecemos um método de RT-PCR quantitativa (qRT-PCR) para a dosagem do RVGPmRNA. Desenvolvemos também um novo método de titulação de partículas SFV por qRT-PCR, aplicável a praticamente qualquer construção de SFV recombinante. Em ensaios preliminares, as preparações de RVGP recombinante utilizadas para a imunização de camundongos foram capazes de induzir a formação de anticorpos neutralizantes e de conferir um bom grau de proteção ao teste de desafio intracerebral com vírus rábico. / The rabies virus glycoprotein is the major antigen able to induce a neutralizing antibody response and survival after challenge against rabies virus infection. We have studied the kinetic expression of RVGP and its messenger RNA (RVGPmRNA) in two different recombinant expression systems: stably transfected Drosophila melanogaster cells (rS2) and BHK-21 cells infected with Semliki Forest Virus carrying RVGP genetic information (SFV-RVGP). We have done a work of standardization of the cell culture samples treatment prior to RVGP quantification by ELISA, and we have developed and standardized a quantitative RT-PCR (qRT-PCR) to quantify the RVGPmRNA. We have also developed a new method of SFV particles titration by qRT-PCR, which is applicable to other constructions of recombinant SFV. We utilized the RVGP expressed by rS2 and SFV-RVGP systems on preliminary in vivo assays. The RVGP samples used to mice immunization were able to induce neutralizing antibodies and to lead to a nice level of protection against the intracerabral rabies virus challenge.

O interesse em células de inseto para a produção de proteínas complexas se deve a sua maior facilidade de cultivo e ao padrão equivalente de glicosilação quando comparado aos sistemas com células de mamíferos. O objetivo deste trabalho foi identificar fatores que limitam ou inibem a produção da glicoproteína do vírus rábico (GPV) expressa na membrana citoplasmática de células de Drosophila melanogaster transfectadas, quando cultivadas em biorreator de bancada agitado e bubble-free, operado em modo descontínuo. Avaliaram-se as influências de oxigênio dissolvido (5 < pO2 <80%), da glicose (1 < GLC0 < 15g/L) e da glutamina (0.6 < GLN0 < 7g/L). Essas variáveis afetaram de forma diferenciada o crescimento celular (produção de células e velocidades específicas-µX), o metabolismo celular (fatores de conversão - YX/GLC, YX/GLN, YLAC/GLC, YALA/GLC, YNH4/GLN, YALA/GLN), assim como a expressão da proteína recombinante (concentração, teor celular e produtividade). O aumento do pO2 reduziu em 9 vezes o crescimento celular mas aumentou o teor celular de GPV em 1,4 vezes. Baixos valores de GLC0 e GLN0, claramente, limitaram o crescimento, de modo que incrementos na concentração desses substratos, até valores intermediários, aumentaram µX,MAX em 3 vezes e 2,5 vezes, respectivamente, e a produção de células em 11 vezes e 3 vezes, respectivamente. O teor celular de GPV máximo não foi afetado pela GLC, mas aumentou em 100% para valores de GLN0 igual ou superiores a 3,5 g/L. As concentrações de lactato produzidas foram consideradas baixas (inferiores a 0,8 g/L) para exercer qualquer efeito de inibição sobre o crescimento ou a expressão da proteína. Por sua vez, as concentrações de amônio parecem inibir tanto a produção de GPV (NH4+~50mg/L) quanto o crescimento celular (NH4+~80mg/L). A condição de cultivo com de 30% de pO2, 10 g/L de GLC0 e 3,5 g/L de GLN0 resultou nos maiores valores de produtividade (9,1 µg/L.h) e de concentração de GPV (1,2 mg/L). O metabolismo de GLC e GLN apresentou grande interdependência, com alterações em GLC0 afetando o metabolismo de GLN e vice-versa. Assim, em condições de excesso de GLC0, as células apresentaram um metabolismo mais ineficiente com reduções nos fatores YX/GLC (2,3 vezes) e YX/GLN (4,6 vezes) e maior geração de subprodutos, caracterizada por incrementos nos valores de YALA/GLC (51%), YLAC/GLC (11%) e YNH4/GLN (15%). O metabolismo da GLN apresentou resposta característica de substrato em excesso para toda a faixa de valores ensaiada, com redução de 25 vezes no valor de YX/GLN e inesperadamente também uma redução na geração de subprodutos de 7 vezes para YNH4/GLN e 12 vezes para YALA/GLN. O efeito sobre o metabolismo da GLC foi mais acentuado para valores mais elevados de GLN0, com redução de 3,6 vezes para YX/GLC e incrementos de 70% para YALA/GLC e para YLAC/GLC. Os resultados sugerem ainda que a célula utiliza duas vias para metabolizar a glutamina: glutaminólise, em condição de limitação em GLC; ou glutamato sintase - NADH-GOGAT, em condição de excesso em GLC. A célula demonstrou também capacidade de sintetizar GLN, a partir de amônio ou outros aminoácidos, quando atingiu concentrações abaixo de 50 mg/L. / The interest in using insect cells to produce complex proteins is due to its ease of cultivation and its glycosylation pattern equivalent to that of mammalian cells systems. The objective of this work was to identify the limiting or inhibiting factors for the production of a rabies virus glycoprotein (RVGP), expressed in the cytoplasmatic membrane of a transfected Drosophila melanogaster S2 cells, when cultivated in a bench stirred bubble-free bioreactor, in batch mode. The influence of dissolved oxygen (5 < pO2 < 80%), of initial glucose concentration (1 < GLC0 < 15 g/L) and of initial glutamine concentration (0.6 < GLN0 < 7 g/L) was evaluated. These variables affected in a different way cell growth (cell production and specific growth rate - µX), cell metabolism (yield factors - YX/GLC, YX/GLN, YLAC/GLC, YALA/GLC, YNH4/GLN and YALA/GLN), as well as the recombinant protein expression (RVGP concentration, RVGP cell content and RVGP productivity). pO2 increase reduced 9 times cell growth, but increased 1.4 times RVGP cell content. Low initial glucose and glutamine concentrations clearly limited the cell growth, in such a way that raising these substrates concentrations up to intermediate values, increased µX,MAX 3 times and 2.5 times, respectively, and increased cell production 11 times and 3 times, respectively. The maximum RVGP cell content was not affected by GLC0, but improved 100% when GLN0 was 3.5 g/L or higher. The concentrations of produced lactate were considered low (below 0.8 g/L) to cause any inhibition effect on growth or protein expression. On the other hand, ammonium concentrations seem to inhibit RVGP production (NH4+~50 mg/L), as well as cell growth (NH4+~80 mg/L). Maximum productivity values (9.1 µg/L.h) and RVGP concentration (1.2 mg/L) were attained for 30% pO2, 10 g/L of GLC0 and 3.5 g/L of GLN0 run. The metabolism of GLC and GLN showed a great interdependence, with GLC0 changes affecting the GLN metabolism, and viceversa. Thus, in glucose excess condition, cell metabolism was less efficient. This implied in reduction of yield factors - YX/GLC (2.3 times) e YX/GLN (4.6 times) - and in higher by-products generation, characterized by augmentation in YALA/GLC (51%), YLAC/GLC (11%) and YNH4/GLN (15%). The glutamine metabolism showed a substrate excess response pattern to the whole range of concentration studied, with reduction of YX/GLN (25 times) and, unexpectedly, a reduction of by-products liberation - YNH4/GLN (7 times) and YALA/GLN (12 times). The effect on glucose metabolism was more intense when the glutamine concentration was higher, showing a 3.6 times diminution YX/GLC and a 70% augmentation for YALA/GLC and YLAC/GLC. The results suggest that cells metabolize glutamine through two different pathways glutaminolysis, under glucose limitation, or glutamate synthase - NADH-GOGAT, under glucose excess. The cell, proved also to be able to synthesize glutamine from ammonium or other amino acids, when it reached concentrations below 50 mg/L.

Until very recently, synthetic chemistry has focussed on the creation of chemical entities with desirable properties through the programmed application of isolated chemical reactions, either individually or in a cascade that afford a target compound selectively. By contrast, biological systems operate using a plethora of complex interconnected signaling and metabolic networks with multiple checkpoint controls and feedback loops allowing biological systems to adapt and respond rapidly to external stimuli. Systems chemistry attempts to capture the complexity and emergent phenomena prevalent in the life sciences within a wholly synthetic chemical framework. In this approach, complex phenomena are expressed by a group of synthetic chemical entities designed to interact and react with many partners within the ensemble in programmed ways. In this manner, it should be possible to create synthetic chemical systems whose properties are not simply the linear sum of the attributes of the individual components. Chapter 1 discusses the role of complex networks in various aspects of chemistry- related research from the origin of life to nanotechnology. Further, it introduces the concept of Systems chemistry, giving various examples of dynamic covalent networks, self-replicating systems and molecular logic gates, showing the applications of complex system research. Chapter 2 discusses the components of replicator design. Further, it introduces a network based on recognition mediated reactions that is implemented by length- segregation of the substrates and displays properties of self-sorting. Chapter 3 presents a fully addressable chemical system based on auto- and cross- catalytic properties of product templates. The system is described by Boolean logic operations with different template inputs giving different template outputs. Chapter 4 introduces a dynamic network which fate is determined by a single recognition event. The replicator is capable of exploiting and dominating the exchanging pool of reagents in order to amplify its own formation at the expense of other species through the non-linear kinetics inherent in minimal replication. Chapter 5 focuses on the development of complex dynamic systems from structurally simple molecules. The new approach allows creating multicomponent networks with many reaction pathways operating simultaneously from readily available substrates.

Quando há interesse em monitorar a variância de uma característica da qualidade de interesse através de gráfico de controle por variáveis, o gráfico S2 é a alternativa mais usual. Entretanto, há situações onde mensurar a característica da qualidade é caro, consome mais tempo por unidade de inspeção, requer maior esforço dos operadores quanto à obtenção dos dados ou envolve ensaios destrutivos. Nestes casos, a classificação da variável contínua em categorias através de um dispositivo torna-se uma alternativa interessante. A avaliação pode ser mais rápida, a análise e o equipamento utilizado podem ser mais simples, de modo que o custo final da inspeção seja menor. O objetivo do trabalho é propor um gráfico de controle por atributos para monitoramento da variabilidade. Para tanto a estatística GS2 é calculada e gráfico sinaliza se GS2 > LC, LC limite de controle determinado de modo que minimize o ARL1, fixado um valor de ARL0. Como resultado a performance do gráfico GS2 é comparada ao gráfico S2 em termos de ARL1. / In cases aiming at monitoring the variance of a products quality characteristics using a variable control chart, chart S2 is the most used alternative. However, in some situations, this solution can be expensive, demand more time per individual inspected unit, demand greater efforts from operators to acquire data or involve destructive tests. In such cases, the use of a gauge measurement tool to classify the continuous variable into categories, becomes an interesting alternative. The assessment can be faster, the analysis and the tool used can be simple, resulting in less costly final inspections. This work proposes the use of an attribute control chart to monitor variability. Statistics GS2 is calculated and control chart signalize if GS2 > CL, whereas CL is the determined control limit, minimizing ARL1 for a fixed value of ARL0. GS2 control chart performance is compared to S2 chart based on ARL1.

Quando há interesse em monitorar a variância de uma característica da qualidade de interesse através de gráfico de controle por variáveis, o gráfico S2 é a alternativa mais usual. Entretanto, há situações onde mensurar a característica da qualidade é caro, consome mais tempo por unidade de inspeção, requer maior esforço dos operadores quanto à obtenção dos dados ou envolve ensaios destrutivos. Nestes casos, a classificação da variável contínua em categorias através de um dispositivo torna-se uma alternativa interessante. A avaliação pode ser mais rápida, a análise e o equipamento utilizado podem ser mais simples, de modo que o custo final da inspeção seja menor. O objetivo do trabalho é propor um gráfico de controle por atributos para monitoramento da variabilidade. Para tanto a estatística GS2 é calculada e gráfico sinaliza se GS2 > LC, LC limite de controle determinado de modo que minimize o ARL1, fixado um valor de ARL0. Como resultado a performance do gráfico GS2 é comparada ao gráfico S2 em termos de ARL1. / In cases aiming at monitoring the variance of a products quality characteristics using a variable control chart, chart S2 is the most used alternative. However, in some situations, this solution can be expensive, demand more time per individual inspected unit, demand greater efforts from operators to acquire data or involve destructive tests. In such cases, the use of a gauge measurement tool to classify the continuous variable into categories, becomes an interesting alternative. The assessment can be faster, the analysis and the tool used can be simple, resulting in less costly final inspections. This work proposes the use of an attribute control chart to monitor variability. Statistics GS2 is calculated and control chart signalize if GS2 > CL, whereas CL is the determined control limit, minimizing ARL1 for a fixed value of ARL0. GS2 control chart performance is compared to S2 chart based on ARL1.

This work devises a novel approach for mitigating the effects of Catastrophic Forgetting in Deep Reinforcement Learning-based cognitive radio engine implementations employed in space communication applications. Previous implementations of cognitive radio space communication systems utilized a moving window- based online learning method, which discards part of its understanding of the environment each time the window is moved. This act of discarding is called Catastrophic Forgetting. This work investigated ways to control the forgetting process in a more systematic manner, both through a recursive training technique that implements forgetting in a more controlled manner and an ensemble learning technique where each member of the ensemble represents the engine's understanding over a certain period of time. Both of these techniques were integrated into a cognitive radio engine proof-of-concept, and were delivered to the SDR platform on the International Space Station. The results were then compared to the results from the original proof-of-concept. Through comparison, the ensemble learning technique showed promise when comparing performance between training techniques during different communication channel contexts.

As one of the most powerful error-correcting codes, Low-density parity check codes are widely used in digital communications. Because of the performance of LDPC codes are capable to close the shannon limited extraordinarily, LDPC codes are to be used in the new Digital Video Broadcast-Satellite-Second Generation(DVB-S2) and it is the first time that LDPC codes are included in the broadcast standard in 2003. In this thesis, a restructured parity-check matrices which can be divided into sub-matrices for LDPC code in DVB-S2 is provided. Corresponded to this restructured parity-check matrix, a reconstructed decoding table is invented. The encoding table of DVB-S2 standard only could obtain the unknown check nodes from known variable nodes, while the decoding table this thesis provided could obtain the unknown variable nodes from known check nodes what is exactly the Layered-massage passing algorithm needed. Layered-message passing algorithm which also known as "Turbo-decoding message passing" is used to reduce the decoding iterations and memory storage for messages. The thesis also investigate Bp algorithm, lambda-min algorithm, Min-sum algorithm and SISO-s algorithm, meanwhile, simulation results of these algorithms and schedules are also presented.

Les retours et les études dont on dispose sur les réseaux IP par satellite ne permettent pas d'apprécier les performances dont ils sont capables. Pourtant, les difficultés de transmettre de l'IP par satellite persistent encore. L'inadaptation du protocole IP, initialement conçu pour des réseaux terrestres, au large produit délai-bande du média satellite est une raison. Le fonctionnement souvent dé-corrélé entre les niveaux supérieurs de la pile TCP/IP et les couches physique et MAC du média satellite, est une autre. Dans le cadre de ce travail de thèse nous adoptons une démarche expérimentale basée sur l'observation, l'analyse et l'évaluation de systèmes implantant des technologies IP par satellite tels que le standard DVB-RCS, la technologie propriétaire iDirect ou la nouvelle norme DVB-S2. Nous étudions l'impact des règles de qualité de service IP sur les performances des applications dans un contexte de bande limitée. Nous nous penchons notamment sur l'évaluation des efficacités de l'encapsulation IP en termes de consommation de bande. Notre premier objectif est de déceler les niveaux auxquels un opérateur peut agir en vue d'optimiser la configuration d'un système IP par satellite et en accroître les performances

Ένα από τα βασικά μειονεκτήματα που παρουσιάζει ο σχεδιασμός και η υλοποίηση LDPC αποκωδικοποιητών είναι η μεγάλη πολυπλοκότητα που παρουσιάζεται σε επίπεδο υλικού εξαιτίας της εσωτερικής διασύνδεσης των μονάδων επεξεργασίας δεδομένων.H αρχιτεκτονική που επιτυγχάνει το μέγιστο επίπεδο παραλληλότητας και κατά συνέπεια είναι πολύ αποδοτική όσον αφορά την ταχύτητα αποκωδικοποίησης, δεν χρησιμοποιείται συχνά εξαιτίας της πολυπλοκότητας του υλικού λόγω των πολλαπλών κυκλωμάτων διασύνδεσης που απαιτεί. Στην παρούσα διπλωματική εργασία προτείνεται μια νέα αρχιτεκτονική για το δίκτυο διασύνδεσης ενώ παράλληλα έχει υλοποιηθεί και ένας αλγόριθμος για την αποδοτική τοποθέτηση των επεξεργαστικών μονάδων σε αυτό το δίκτυο. Επίσης έχει μελετηθεί και η επίδραση μειωμένης μετάδοσης πληροφορίας σε κάθε επανάληψη του αλγορίθμου αποκωδικοποίησης.Το περιβάλλον που χρησιμοποιήθηκε για την εξομοίωση και την παραγωγή των αποτελεσμάτων είναι η πλατφόρμα της Matlab. Η προτεινόμενη αρχιτεκτονική υλοποιήθηκε και εξομοιώθηκε σε κώδικες LDPC που αποτελούν μέρος του προτύπου DVB - S2 (Digital Video Broadcasting).Το συγκεκριμένο πρότυπο, εκτός των άλλων, καθορίζει και τις προδιαγραφές των κωδίκων LDPC που χρησιμοποιούνται κατά την κωδικοποίηση και αποκωδικοποίηση δεδομένων σε συστήματα ψηφιακής δορυφορικής μετάδοσης. Τα αποτελέσματα των εξομοιώσεων σχετίζονται με την πολυπλοκότητα της προτεινόμενης αρχιτεκτονικής σε υλικό αλλά και της απόδοσης (ταχύτητα αποκωδικοποίησης) και συγκρίνονται με την βασική πλήρως παράλληλη αρχιτεκτονική. / One of the main disadvantages of the design and implementation of LDPC decoders is the great complexity presented at the hardware level because of the internal interconnection of processing units. The fully parallel architecture that achieves the maximum level of parallelism and hence is very efficient in terms of speed decoding is not used often because of the hardware complexity due to the multiple interface circuits required. This MSc thesis proposes a new architecture for the network interface and also introduces an algorithm for the efficient placement of the processing units in this network. In addition to that, a modified version of the decoding algorithm has been implemented. The relative advantage of this algorithm is that in each iteration only a percentage of the processing units exchange information with each other. That approach further reduces the hardware complexity and power usage. The environment used to simulate and produce the results is Matlab. The proposed architecture is implemented and simulated in LDPC codes that are part of the standard DVB - S2 (Digital Video Broadcasting). This standard, among other things, determines the specifications of the LDPC codes used in the channel encoding and decoding process in digital satellite transmission systems. The results of the simulations related to the complexity of the proposed architecture in hardware and performance (decoding speed) are compared with the fully parallel architecture.