The main scope of this work was to investigate whether it is possible to produce free standing 3D structures by the means of spraying and using MFC as raw material. This was carried out by diluting MFC into a consistency of 2% and spraying it onto a male mould. During the trials, several different devices and procedures were investigated in order to find a possible solution. The results from the laboratory trials showed that it is possible to create trays of MFC that could suitable as a detail for packaging. The important thing was to pre-heat the mould before spraying, build the tray in layers, where spraying should be carried out in a 45° angle, with single sweeps while rotating the mould in the beginning of the process, and to use a drying method, were the drying could be focused on the wet parts at the same time as it could avoid those who already had been dried, to dry the sample between each layer of MFC until the wet surface disappeared. Exactly how many sweeps per layer that should be sprayed after the first drying does not matter much, the important thing was that the layers do not become too thick. Because then, too much moisture was trapped inside the samples which made them burst during the drying.

Cette thèse présente des résultats dans le domaine de la prédiction d’interactions protéine-ARN. C’est un domaine de recherche très actif, pour lequel la communauté internationale organise régulièrement des compétitions pour évaluer différentes techniques de prédictions in silico d’interactions protéine-protéine et protéine-ARN sur des données benchmarks (CAPRI, Critical Assessment of PRedictedInteractions), par prédiction en aveugle et en temps limité. Dans ce cadre, de nombreuses approches reposant sur des techniques d’apprentissage supervisé ont récemment obtenus de très bons résultats.Nos travaux s’inscrivent dans cette démarche.Nous avons travaillé sur des jeux de données de 120 complexes protéine-ARN extraits de la PRIDB non redondante (Protein-RNA Interface DataBase, banque de données de référence pour les interactions protéine-ARN). La méthodologie de prédiction d'interactions protéine-ARN a aussi été testée sur 40 complexes issus de benchmarks de l'état de l'art et indépendants des complexes de la PRIDB non redondante. Le faible nombre de structures natives et la difficulté de générer in silico des structures identiques à la solution in vivo nous a conduit à mettre en place une stratégie de génération de candidats par perturbation de l’ARN partenaire d’un complexe protéine-ARN natif. Les candidats ainsi obtenus sont considérés comme des conformations presque-natives si elles sont suffisamment proches du natif. Les autres candidats sont des leurres. L’objectif est de pouvoir identifier les presque natifs parmi l’ensemble des candidats potentiels, par apprentissage supervisé d'une fonction de score.Nous avons conçu pour l'évaluation des fonctions de score une méthodologie de validation croisée originale appelée le leave-"one-pdb"-out, où il existe autant de strates que de complexes protéine-ARN et où chaque strate est constituée des candidats générés à partir d'un complexe. L’une des approches présentant les meilleures performances à CAPRI est l’approche RosettaDock, optimisée pour la prédiction d’interactions protéine-protéine. Nous avons étendu la fonction de score native de RosettaDock pour résoudre la problématique protéine-ARN. Pour l'apprentissage de cette fonction de score, nous avons adapté l'algorithme évolutionnaire ROGER (ROC-based Genetic LearnER) à l'apprentissage d'une fonction logistique. Le gain obtenu par rapport à la fonction native est significatif.Nous avons aussi mis au point d'autres modèles basés sur des approches de classifieurs et de métaclassifieurs, qui montrent que des améliorations sont encore possibles.Dans un second temps, nous avons introduit et mis en oeuvre une nouvelle stratégie pour l’évaluation des candidats qui repose sur la notion de prédiction multi-échelle. Un candidat est représenté à la fois au niveau atomique, c'est-à-dire le niveau de représentation le plus détaillé, et au niveau dit “gros-grain”où nous utilisons une représentation géométrique basée sur des diagrammes de Voronoï pour regrouper ensemble plusieurs composants de la protéine ou de l’ARN. L'état de l'art montre que les diagrammes de Voronoï ont déjà permis d'obtenir de bons résultats pour la prédiction d'interactions protéine-protéine. Nous en évaluons donc les performances après avoir adapté le modèle à la prédiction d'interactions protéine-ARN. L’objectif est de pouvoir rapidement identifier la zone d’interaction (épitope) entre la protéine et l’ARN avant d’utiliser l’approche atomique, plus précise,mais plus coûteuse en temps de calcul. L’une des difficultés est alors de pouvoir générer des candidats suffisamment diversifiés. Les résultats obtenus sont prometteurs et ouvrent desperspectives intéressantes. Une réduction du nombre de paramètres impliqués de même qu'une adaptation du modèle de solvant explicite pourraient en améliorer les résultats. / My thesis shows results for the prediction of protein-RNA interactions with machine learning. An international community named CAPRI (Critical Assessment of PRedicted Interactions) regularly assesses in silico methods for the prediction of the interactions between macromolecules. Using blindpredictions within time constraints, protein-protein interactions and more recently protein-RNA interaction prediction techniques are assessed.In a first stage, we worked on curated protein-RNA benchmarks, including 120 3D structures extracted from the non redundant PRIDB (Protein-RNA Interface DataBase). We also tested the protein-RNA prediction method we designed using 40 protein-RNA complexes that were extracted from state-ofthe-art benchmarks and independent from the non redundant PRIDB complexes. Generating candidates identical to the in vivo solution with only a few 3D structures is an issue we tackled by modelling a candidate generation strategy using RNA structure perturbation in the protein-RNAcomplex. Such candidates are either near-native candidates – if they are close enough to the solution– or decoys – if they are too far away. We want to discriminate the near-native candidates from thedecoys. For the evaluation, we performed an original cross-validation process we called leave-”onepdb”-out, where there is one fold per protein-RNA complex and each fold contains the candidates generated using one complex. One of the gold standard approaches participating in the CAPRI experiment as to date is RosettaDock. RosettaDock is originally optimized for protein-proteincomplexes. For the learning step of our scoring function, we adapted and used an evolutionary algorithm called ROGER (ROC-based Genetic LearnER) to learn a logistic function. The results show that our scoring function performs much better than the original RosettaDock scoring function. Thus,we extend RosettaDock to the prediction of protein-RNA interactions. We also evaluated classifier based and metaclassifier-based approaches, which can lead to new improvements with further investigation.In a second stage, we introduced a new way to evaluate candidates using a multi-scale protocol. A candidate is geometrically represented on an atomic level – the most detailed scale – as well as on a coarse-grained level. The coarse-grained level is based on the construction of a Voronoi diagram over the coarse-grained atoms of the 3D structure. Voronoi diagrams already successfully modelled coarsegrained interactions for protein-protein complexes in the past. The idea behind the multi-scale protocolis to first find the interaction patch (epitope) between the protein and the RNA before using the time consuming and yet more precise atomic level. We modelled new scoring terms, as well as new scoring functions to evaluate generated candidates. Results are promising. Reducing the number of parameters involved and optimizing the explicit solvent model may improve the coarse-grained level predictions.

ESPOSITO, Roberta. Application of Magnetotelluric Method in sedimentary environments and analysis of the resistivity dispersion in presence of 3D polarizable structures. 2016. 106 f. Tese (Doutorado em Geologia)-Universidade Federal do Ceará, Fortaleza, 2016. / Submitted by Jairo Viana (jairo@ufc.br) on 2017-05-29T20:43:16Z No. of bitstreams: 1 2016_tese_resposito.pdf: 62027805 bytes, checksum: db66ec49cef3ab46aeb1ee42cfa88b82 (MD5) / Approved for entry into archive by Jairo Viana (jairo@ufc.br) on 2017-05-29T20:43:56Z (GMT) No. of bitstreams: 1 2016_tese_resposito.pdf: 62027805 bytes, checksum: db66ec49cef3ab46aeb1ee42cfa88b82 (MD5) / Approved for entry into archive by Jairo Viana (jairo@ufc.br) on 2017-05-29T20:44:04Z (GMT) No. of bitstreams: 1 2016_tese_resposito.pdf: 62027805 bytes, checksum: db66ec49cef3ab46aeb1ee42cfa88b82 (MD5) / Made available in DSpace on 2017-05-29T20:44:04Z (GMT). No. of bitstreams: 1 2016_tese_resposito.pdf: 62027805 bytes, checksum: db66ec49cef3ab46aeb1ee42cfa88b82 (MD5) Previous issue date: 2016 / O Método magnetotelúrico permite a determinação da distribuição de resistividade elétrica na subsuperfície, a fim de diferenciar as estruturas presentes. Com os procedimentos de inversão de dados é possível obter valores de resistividade de algumas dezenas de metros até centenas de quilómetros. Como acontece com todos os métodos de prospecção geofísica, o método MT está sujeito a ambiguidades por causa do fenômeno da dispersão da resistividade. Este fenómeno pode deformar as curvas de resposta MT e pode conduzir a uma interpretação errónea. Uma conclusão geral é que os efeitos de dispersão pode influenciar a resposta MT em forma reconhecível ou sutil. Em ambos os casos, sem considerar os efeitos de distorção a interpretação pode levar a conclusões enganosas sobre as propriedades físicas das estruturas pesquisadas. Este trabalho trata com diferentes aspectos do Método Magnetotelúrico. Em primeiro lugar a dispersão da resistividade é estudada teoricamente realizando, pela primeira vez, simulações de um modelo 3-D da Terra (caso 1-D é apresentado a partir de um trabalho anterior de Esposito e Patella, 2009 e o caso 2-D é reconstruído baseando a simulação sobre um trabalho de Mauriello et al., 1996), com um estudo de caso em uma área geotérmica (Snake River Plain, Idaho, USA). Mostra-se uma aplicação do MT na área do Pecém (Ceará, Brasil), desconsiderando a dispersão da resistividade, a fim de mostrar a eficiência do método para resolver contrastes de resitividade na subsuperfície e, finalmente, são realizadas as simulações teóricas que mostram a aplicação do método MT para pesquisa de Oléo e Gás na porção emersa da Bacia Potiguar (Rio Grande do Norte, Brasil), a fim de considerar a MT como um apoio em áreas onde a sísmica, usada principalmente para esses fins, pode ser logisticamente difícil de aplicar ou precisa de uma comparação com outros métodos de prospecção geofísica. / O Método magnetotelúrico permite a determinação da distribuição de resistividade elétrica na subsuperfície, a fim de diferenciar as estruturas presentes. Com os procedimentos de inversão de dados é possível obter valores de resistividade de algumas dezenas de metros até centenas de quilómetros. Como acontece com todos os métodos de prospecção geofísica, o método MT está sujeito a ambiguidades por causa do fenômeno da dispersão da resistividade. Este fenómeno pode deformar as curvas de resposta MT e pode conduzir a uma interpretação errónea. Uma conclusão geral é que os efeitos de dispersão pode influenciar a resposta MT em forma reconhecível ou sutil. Em ambos os casos, sem considerar os efeitos de distorção a interpretação pode levar a conclusões enganosas sobre as propriedades físicas das estruturas pesquisadas. Este trabalho trata com diferentes aspectos do Método Magnetotelúrico. Em primeiro lugar a dispersão da resistividade é estudada teoricamente realizando, pela primeira vez, simulações de um modelo 3-D da Terra (caso 1-D é apresentado a partir de um trabalho anterior de Esposito e Patella, 2009 e o caso 2-D é reconstruído baseando a simulação sobre um trabalho de Mauriello et al., 1996), com um estudo de caso em uma área geotérmica (Snake River Plain, Idaho, USA). Mostra-se uma aplicação do MT na área do Pecém (Ceará, Brasil), desconsiderando a dispersão da resistividade, a fim de mostrar a eficiência do método para resolver contrastes de resitividade na subsuperfície e, finalmente, são realizadas as simulações teóricas que mostram a aplicação do método MT para pesquisa de Oléo e Gás na porção emersa da Bacia Potiguar (Rio Grande do Norte, Brasil), a fim de considerar a MT como um apoio em áreas onde a sísmica, usada principalmente para esses fins, pode ser logisticamente difícil de aplicar ou precisa de uma comparação com outros métodos de prospecção geofísica.

This project consists of three parts. The first part is aimed to summarize list of actual packaging systems and especially systems using 3D construction. Project continues in the second part, which is more practical and contains design and production of organic and inorganic testing substrates for lead-free soldered 3D structures. Last experimental part is about tests performed on soldered substrates and evaluation of results of these practical tests.

L’efficacité des chimiothérapies est limitée par la surexpression de pompes d’efflux adressées à la membrane plasmique des cellules cibles. En effet, celles-ci réduisent le taux intracellulaire des médicaments anticancéreux, antiviraux, antifongiques et antibactériens. La P-gp/ABCB1 est la plus impliquée dans ce phénomène, suivie de MRP1/ABCC1 et de BCRP/ABCG2. Une approche pour surmonter ce phénomène est de développer des médicaments qui ne soient pas expulsés par ces pompes. Dans ce contexte, nous avons développé une nouvelle classe d’inhibiteurs de la protéase du VIH-1 qui ne sont ni transportés par P-gp ni par BCRP. Ils sont ainsi des candidats intéressants aux trithérapies contre le SIDA. Un point clé pour comprendre comment ces transporteurs font traverser les médicaments à travers la membrane est d’identifier des nouvelles structures. Ainsi, nous avons résolu trois structures de P-gp de souris. Une d’entre-elles est complexée à un nano-anticorps lié au premier NBD («nucleotide-binding domain»), qui fige la P-gp dans une nouvelle conformation ouverte vers l'intérieur. Pour finir, nous avons localisé deux sites de liaison de P-gp en caractérisant les modes d'inhibition de deux inhibiteurs précédemment cocristallises avec la pompe. Ceci permet de mieux comprendre le mécanisme de translocation et offre la possibilité de cibler plus précisément ces sites pour développer des modulateurs de cette pompe / Resistance to chemotherapy is partly due to efflux pumps expressed in the plasma membrane which prevents the accumulation of anticancer, antiviral, antifungal and antibacterial drugs in target cells. Three human ABC transporters are particularly involved in MDR phenotype: P-gp/ABCB1, MRP1/ABCC1 and BCRP/ABCG2. Among the different approaches used to overcome the resistance linked to these transporters, the development of non-substrate drugs MDR-ABC transporters has been described. Here, new class of HIV-1 protease inhibitors not recognized by P-gp/BCRP were identified, promising to be attractive candidates to HAART therapy. Since the determination of the X-ray structures in different conformations is a key point to understand how MDR-ABC transporters translocate drugs across the plasma membrane, the crystal structures of three inward-facing conformations of mouse P-gp were resolved. One structure has a camel nanobody bound to the C-terminal side of the first nucleotide-binding domain, revealing a unique epitope on P-gp and freezing a new open-inward conformation. Finally, the enzymatic characterization of two inhibitors co-crystallized with the mouse P-gp has allowed to localize two main binding sites by which drugs efflux occurs. These results bring new findings of the drug-efflux mechanism and offer the possibility to target more precisely those sites to develop modulators of this pump

The automatic prediction of protein three dimensional structures from its amino acid sequence has become one of the most important and researched fields in bioinformatics. As models are not experimental structures determined with known accuracy but rather with prediction it’s vital to determine estimates of models quality. We attempt to solve this problem using machine learning techniques and information from both the sequence and structure of the protein. The goal is to generate a machine that understands structures from PDB and when given a new model, predicts whether it belongs to the same class as the PDB structures (correct or incorrect protein models). Different subsets of PDB (protein data bank) are considered for evaluating the prediction potential of the machine learning methods. Here we show two such machines, one using SVM (support vector machines) and another using fuzzy decision trees (FDT). First using a preliminary encoding style SVM could get around 70% in protein model quality assessment accuracy, and improved Fuzzy Decision Tree (IFDT) could reach above 80% accuracy. For the purpose of reducing computational overhead multiprocessor environment and basic feature selection method is used in machine learning algorithm using SVM. Next an enhanced scheme is introduced using new encoding style. In the new style, information like amino acid substitution matrix, polarity, secondary structure information and relative distance between alpha carbon atoms etc is collected through spatial traversing of the 3D structure to form training vectors. This guarantees that the properties of alpha carbon atoms that are close together in 3D space and thus interacting are used in vector formation. With the use of fuzzy decision tree, we obtained a training accuracy around 90%. There is significant improvement compared to previous encoding technique in prediction accuracy and execution time. This outcome motivates to continue to explore effective machine learning algorithms for accurate protein model quality assessment. Finally these machines are tested using CASP8 and CASP9 templates and compared with other CASP competitors, with promising results. We further discuss the importance of model quality assessment and other information from proteins that could be considered for the same.

This work deals with evolutionary design of 3D structures. The work brings the summary of the previous works in this area and brings autor's suggested solution of evolutionary design of 3D structures. This paper seeks to the ability of easy fitness function definition in the systems for evolutionary design of structures. The author tries to make one of the first steps to the future systems for evolution design of any universal structures in contrast with the evolution systems for design of a concrete type of structure. The result of this work is the basic system for evolutionary design of 3D structures with the ability of external fitness function definition via the XML file. This paper offers also the  simple advices and observations for the potential future work in this area.

The doctoral thesis is focused on research on application possibilities of soldered interconnection structures especially for the use with innovative 3D structures assembled using the LTCC ceramic materials, PVD deposited and galvanically modified pads, solid core solder balls. It consists of several main parts. Introduction is followed by theoretical survey of current situation and technologies. Thesis continues with introduction of main goals and with summary of successively performed experiments and their results. End of the work belongs to summary of results and gained knowledge.

Les structures protéiques peuvent être divisées en répétitives et apériodiques, les structures apériodiques correspondant pour la plupart à des protéines globulaires. Les protéines répétitives (PRs) contiennent des unités de répétitions adjacentes, appelées séquences répétées en tandem (TRs). Les PRs sont abondantes et ont une importance fonctionnelle fondamentale. De plus de nombreuses études ont démontré l'implication des TRs dans les pathologies humaines. Ainsi, la découverte des PRs et la compréhension de leur relation séquence-structure-fonction, offrent des perspectives de recherche prometteuses.Le développement d’initiatives en génomique structurale, combiné à une meilleure adaptation des techniques de cristallographie et de RMN à l’étude des protéines non globulaires, a permis d’élucider la structure d’un nombre croissant de PRs, d’où la nécessité de mettre en place un système de classification. Les structures répétitives ont été réparties en cinq classes, principalement fondées sur la longueur des TRs: Classe I - agrégats cristallins; Classe II - structures fibreuses; Classe III - structures allongées, dont la stabilité dépend des interactions qui s’établissent entre les motifs répétés. Classe IV - structures répétitives fermées ; Classe V - structures en collier de perles. Les efforts de ces dernières années ont abouti au développement d’outils bioinformatiques utiles à la détection et l'analyse d'éléments répétitifs présents au sein des structures protéiques (3D TRs). En fonction des caractéristiques des répétitions, certaines méthodes fonctionnent mieux que d'autres, mais, jusqu’à présent, aucune ne permettait de couvrir toute la gamme des répétitions. Ce constat nous a incités à développer une nouvelle méthode, appelée détecteur de protéines en tandem (TAPO). TAPO exploite les périodicités des coordonnées atomiques ainsi que d'autres types de représentation structurale, comprenant les chaînes générées par un alphabet conformationnel, les cartes de contact entre résidus, et les arrangements en vecteurs d'éléments de structure secondaire. Actuellement, sept scores, issus des caractéristiques analysées par TAPO, sont combinés à l’aide d’une Machine à Vecteur Support pour produire un score final permettant de différencier les protéines renfermant ou non des 3D TRs. En atteignant 94% de sensibilité et 97% de spécificité pour la référence actuelle, TAPO présente des performances améliorées par rapport aux autres méthodes de pointe. Le développement de TAPO offre de nouvelles opportunités pour l’analyse à grande échelle des protéines renfermant des 3D TRs. Ainsi, notre analyse de la base de données PDB, à l’aide de TAPO, a montré que 19% des protéines contiennent des 3D TRs. L'analyse à grande échelle des structures 3D TRs dans PDB nous a également permis de découvrir plusieurs nouveaux types de structures répétitives, absents de la classification existante et dont certains sont décrits ici.Nous avons entrepris une analyse complète des 3D TRs constitutifs du Rossmann Fold (RF). Notre intérêt pour les RFs a été suscité par le fait que de nombreuses protéines RFs représentent un cas ambigüe vis à vis des structures répétitives et non répétitives. A priori, les unités hélice α - feuillet β des RFs devraient avoir une forte tendance à s’empiler et donc, à former des structures répétitives. Afin de déterminer la fréquence à laquelle les RFs forment de longues unités de répétition empilées, nous avons sélectionné, à l’aide de TAPO, des structures contenant des RFs et les avons classées. Notre analyse montre que les RFs typiques ne peuvent pas être clairement définis comme des structures répétitives mais plutôt comme des unités de structures globulaires, comptant au plus trois répétitions α-β. Des éléments de discussion seront proposés pour tenter d’expliquer cette observation surprenante. / In general, protein structures can be divided into: repetitive and aperiodic structures. Most of the aperiodic structures are globular proteins. The repetitive proteins contain arrays of repeats that are adjacent to each other, called Tandem Repeats (TRs). Proteins containing TRs are abundant and have fundamental functional importance. Numerous studies demonstrated the involvement of such TR-containing proteins in human diseases. Furthermore, genetic instability of these regions can lead to emerging infection threats. Additionally, TR-containing structures have generated significant interest with respect to protein design as they can make excellent scaffolds for specific recognition of target molecules. Therefore, the discovery of these domains, understanding of their sequence–structure–function relationship promises to be a fertile direction for research.The growth of structural genomics initiatives, in combination with improvements in crystallographic and NMR techniques aimed at non-globular proteins, has resulted in an increase in structurally elucidated TR proteins. This has necessitated the development of classification schemes. Structural repeats were broadly divided into five classes mainly based on repeat length; Class I – crystalline aggregates; Class II – fibrous structures such as collagen; Class III – elongated structures where the repetitive units require each other for structural stability such as solenoid proteins; Class IV – closed repetitive structures, such as TIM-barrels and Class V – bead on a string structures such as tandems of Ig-fold domains. Despite this progress, the majority of bioinformatics approaches have focused on non-repetitive globular proteins.In recent years, efforts have been made to develop bioinformatics tools for the detection and analysis of repetitive elements in protein structures (3D TRs). Depending on the size and character of the repeats, some methods perform better than others, but currently no best approach exists to cover the whole range of repeats. This served as a motivation for the development of our method called the TAndem PrOtein detector (TAPO). TAPO exploits, periodicities of atomic coordinates and other types of structural representation, including strings generated by conformational alphabets, residue contact maps, and arrangements of vectors of secondary structure elements. Currently, seven feature based scores produced by TAPO are combined using a Support Vector Machine, producing a score to enable the differentiation between proteins with and without 3D TRs. TAPO shows an improved performance over other cutting edge methods, achieving 94% sensitivity and 97% specificity on the current benchmark. The development of TAPO provided new opportunities for large scale analysis of proteins with 3D TRs. In accordance with our analysis of PDB using TAPO, 19% of proteins contain 3D TRs. The large scale analysis of the 3D TR structures in PDB also allows us to discover several new types of TR structures that were absent in the existing classification. Some of them are described in the thesis manuscript. This suggests that TAPO can be used to regularly update the collection and classification of existing repetitive structures. In particular, a comprehensive analysis of 3D TRs related to Rossmann Fold (RF) was undertaken. Our special interest in RFs was based on the observation that many proteins with RFs represent borderline cases between repetitive and non-repetitive structures. In principle, α-helix-β-strand units of RFs should have a strong potential to stack one over the other, forming repetitive structures. To probe the question of how frequently RFs form long arrays of stacked repeats, we selected by using TAPO known RF-containing structures and classified them. Our analysis shows that typical RFs cannot be clearly defined as repetitive, rather they are part of globular structures with up to 3 αβ-repeats. We provide some explanations for this surprising observation.

La biogenèse des ribosomes eucaryotes est un processus complexe qui implique la production et l'assemblage de 4 ARNr et 80 protéines. La production des deux sous-unités du ribosome, 40S et 60S, débute dans le nucléole par la synthèse d'un long précurseur commun contenant les séquences des ARNr matures et se termine dans le cytoplasme où ont lieu les dernières étapes d'assemblage des protéines ribosomiques et de clivage des ARNr. La production de ribosomes nécessite la participation de plus de 200 co-facteurs, qui catalysent les clivages et modifications des ARNr, coordonnent leur repliement et leur association aux protéines ribosomiques, et assurent des étapes de contrôle-qualité. Ces protéines sont associées aux particules en cours de maturation et absentes des sous-unités matures. Cette voie de synthèse, globalement conservée chez les eucaryotes, a été principalement étudiée chez la levure. Cependant, des études récentes ont montré des différences importantes de ce processus entre levure et mammifères. Un des verrous importants pour comprendre la fonction des co-facteurs, est l'absence de données sur la structure des précurseurs des sous-unités ribosomiques. J'ai donc entrepris une étude structurale de l'assemblage cytoplasmique de la petite sous-unité ribosomique chez l'homme par cryo-microscopie électronique à transmission. Le but de ma thèse était de déterminer la structure 3D des précurseurs de la petite sous-unité ribosomique purifiés à différentes étape de leur maturation. Ce travail a été conduit en collaboration avec l'équipe du Pr. Ulrike Kutay (ETH Zurich) pour la purification des particules pré-40S à partir de cellules humaines. La première structure 3D de particule pré-40S intermédiaire purifiée en étiquetant le co-facteur LTV1 a été déterminée à 19Å de résolution. Dans un deuxième temps, la structure 3D de la particule pré-40S tardive purifiée à via RIO1(KD) a aussi été déterminée à 15Å de résolution. Ces données nous ont permis de proposer un modèle de localisation des co-facteurs sur les précurseurs de la petite sous-unité ribosomique et de montrer une nouvelle différence dans la formation de la petite sous-unité chez l'Homme comparé à la levure, du fait de la présence de la protéine RACK1 sur les particules pré-40S humaines. La comparaison des structures des précurseurs de la petite sous-unité obtenues a permis de mettre en lumière l'existence de remodelages structuraux de la particule pré-40S au cours de sa maturation. Ce travail met en lumière les premières structures 3D de particules pré-40S humaines et pose les fondements méthodologiques d'explorations futures de la dynamique structurale des particules pré-ribosomiques. / Ribosome biogenesis is a complex process that requires the production and the correct assembly of the 4 rRNAs with 80 ribosomal proteins. In Human, the production of the two subunits, 40S and 60S, is initiated by the transcription of a pre-ribosomal rRNA precursor to the mature 18S, 5.8S, and 28S rRNAs by the RNA polymerase I, which is chemically modified and trimmed by endo- and exoribonuclease, in order to form the mature rRNAs. The nascent pre rRNA associated with ribosomal proteins, small ribonucleoprotein particles (snoRNP) and so called co-factors leading to the assembly of an initial 90S particle. This particle is then split into pre-40S and pre-60S pre-ribosomal particles that fallow independent maturation to form the mature subunit into the cytoplasm. Production of eukaryotic ribosomes implies the transient intervention of more than 200 associated proteins and ribonucleoprotein particles, that are absent from the mature subunits. Synthesis of ribosome, globally conserved in eukaryotes, has been principally studied in yeast. However, recent studies reveal that this process is more complex in human compared in yeast. An important bottleneck in this domain is the lack of structural data concerning the formation of intermediate ribosomal subunits to understand the function of assembly factors. Determination of the structural remodeling of pre-ribosomal particles is crucial to understand the molecular mechanism of this complex process. So I have undertaken a structural study on the assembly of the small ribosomal subunit using cryo-electron microscopy and image analysis. The goal of my thesis is to determine the 3D structures of human pre-40S particles at different maturation stages to see the structural remodeling that occurs during the biogenesis of the small ribosomal subunit. We are collaborating with the group of Pr Ulrike Kutay at ETH Zurich, who purify human pre-40S particles. The 3D structures of human pre-40S particles purified at an intermediate and late maturation stages, has been determined with a resolution of 19 and 15Å respectively. Supplementary densities, compared to the mature subunit, indicate the presence of assembly factors and show the unexpected presence of the RACK1 protein in the precursor of the human small ribosomal subunit in the cytoplasm. The comparison of the 3D structures of human pre-40S particle allows showing the structural remodeling that occur during the maturation of the small ribosomal subunit. This work provides the first 3D structure of human pre-40S particles and laid the methodological foundations for future exploration of the structural dynamics of pre-ribosomal particles.