Thèses sur le sujet « Hox clusters »

Il est largement documenté que les patrons épigénétiques sont altérés dans les cancers. Pour autant, l’étendue et la nature précise de ces altérations, tout comme leur impact sur l’expression des gènes, restent encore peu appréciés. Mon projet de thèse est bâti sur ce constat, et s’inscrit en particulier dans la recherche des causes et conséquences des altérations épigénétiques dans les gliomes. Ces tumeurs du système nerveux central représentent en effet un excellent modèle, car elles présentent des défauts de méthylation de l’ADN permettant de discriminer deux populations de tumeurs avec des caractéristiques cliniques différentes. Notre stratégie, basée sur des analyses moléculaires exhaustives, s’est appuyée sur une cohorte de 70 échantillons tumoraux, classés sur la base de leur statut IDH, et de six lignées de cellules souches de glioblastomes (CSG).Ces travaux ont tout d’abord permis de relativiser la contribution de la méthylation de l’ADN dans les dérégulations transcriptionnelles observées dans les gliomes. Il apparait en effet que ce sont plutôt les altérations au niveau de la chromatine bivalente, et plus spécifiquement de la marque H3K27me3, qui sont la cause principale de ces dérégulations transcriptionnelles. Spécifiquement, nos données supportent un modèle selon lequel l’altération dans le contrôle de la marque H3K27me3, et plus spécifiquement dans les interactions entre le complexe PRC2 et la machinerie de transcription spécifique au cerveau, est la cause principale des altérations transcriptionnelles dans les gliomes.Cette étude révèle également que les gènes à homéodomaine, et en particulier les gènes HOX, constituent une catégorie à part dans les gliomes les plus agressifs (IDHwt). Leur signature moléculaire, associant gain d’expression et gain de méthylation de l’ADN, est en effet atypique. Nos données révèlent que cette altération est généralisée aux quatre clusters HOX, et que la réactivation de ces gènes est liée à la perte drastique et spécifique de la marque H3K27me3 sur ces régions. Cette étude conduit également à proposer un modèle original selon lequel l’hypométhylation globale de l’ADN est un élément déclencheur de l’expression ectopique détectée au niveau de nombreux gènes, et dont l’altération des gènes HOX aurait, via un effet domino, un rôle central. L’observation de l’altération de la marque H3K27me3 dans les gliomes, et en particulier aux clusters HOX, nous a également amené à nous interroger sur le rôle des ARN non codants dans ces mécanismes. En ce sens, un transcrit non codant encore peu caractérisé, nommé HOXA-AS2 (situé en antisens au niveau du cluster HOXA), a été identifié. Ce transcrit est significativement et spécifiquement surexprimé dans les gliomes IDHwt. Des approches de sous-expression dans des lignées bien caractérisées de CSG suggèrent un rôle central de HOXA-AS2 dans la biologie de ces cellules. Il contribuerait ainsi au caractère pathologique des CSG en inhibant les voies de l’inflammation et en favorisant la capacité des cellules à proliférer.Dans son ensemble, ce travail revisite le lien entre altérations épigénétiques et défauts d’expression dans les cancers et met en évidence qu’une altération dans le contrôle de la marque H3K27me3 est la principale cause des défauts d’expression des gènes
Epigenetic alterations are a well-known signature of cancer cells. However, the causes of these defects, as well as their consequence on gene expression, remain elusive. My thesis project specifically lies in this thematic, and focuses on the causes and consequences of epigenetic alterations in gliomas. These brain tumors can be divided into two subsets, based on IDH mutation status, that are characterized by different methylation profiles. Interestingly, the mutation of IDH is also associated with a better prognosis. Our strategy, based on exhaustive molecular analyses, relies on the study of 70 glioma samples, classified according to their IDH status, and of six glioblastoma stem cell (GSC) lines.We found that most transcriptional alterations in tumor samples were DNA methylation-independent. Instead, altered histone H3 trimethylation at lysine 27 (H3K27me3) was the predominant molecular defect at deregulated genes. Our results also suggest that the presence of a bivalent chromatin signature at CpG island promoters in stem cells predisposes not only to hypermethylation, as widely documented, but more generally to all types of transcriptional alterations in transformed cells. In addition, the gene expression strength in healthy brain cells influences the choice between DNA methylation- and H3K27me3-associated silencing in glioma. Highly expressed genes were more likely to be repressed by H3K27me3 than by DNA methylation. Our findings support a model in which altered H3K27me3 dynamics, more specifically defects in the interplay between Polycomb protein complexes and the brain-specific transcriptional machinery, is the main cause of transcriptional alteration in glioma cells. Also, our study revealed that homeodomain genes, and in particular HOX genes, are characterized by an atypical defect in aggressive gliomas (IDHwt), associating a gain of expression with an aberrant gain of methylation. We determined that this alteration affect all the four HOX clusters, and that the reactivation of these genes is likely a consequence of the aberrant loss of H3K27me3 that specifically affect these clusters. This study allows to propose a model whereby global DNA hypomethylation triggers ectopic expression of numerous genes through a cascade of events, in which HOX gene alteration would have a central role.The observation that H3K27me3 is deregulated in gliomas, and particularly on HOX genes, also lead us to investigate for the role of non-coding RNA in these mechanisms. We have identified HOXA-AS2, a yet poorly characterized long non-coding RNA located at HOXA locus, that is specifically and significantly overexpressed in IDHwt gliomas. The inhibition of HOXA-AS2 in well-characterized CSG lines suggests that this transcript play a central role in the biology of these cells. Thus, it would contribute to the aggressiveness of CSG by inhibiting inflammatory pathways and promoting cell proliferation. Altogether, these works revisit the relationship between epigenetic alterations and aberrant transcription, and present the control of H3K27me3 as the main cause of transcriptionnel defects in cancer

High redshift galaxy protoclusters are the precursors of today’s massive clusters; the sites of formation of the most massive galaxies in the present-day Universe. By studying these immature structures we can directly analyse the formation of galaxies in the densest environments without relying on extrapolations from low redshift. Finding protoclusters is challenging due to the need for very wide and deep surveys. Radio-loud active galactic nuclei (RLAGN) have been shown to preferentially reside in overdense environments at z > 1. By using these bright radio sources as beacons, protoclusters may be efficiently selected, without the need for large, blind surveys. In this thesis I study the properties of galaxies in high redshift (z > 1.3) clusters and protoclusters selected around RLAGN. Using a sample of 37 clusters and protoclusters from the Clusters Around Radio-Loud AGN (CARLA) survey, I show that the protocluster galaxies have an approximately unevolving, red observed-frame i'-[3.6] colour across 1.3 < z < 3.2. This is at odds with the simple models which are commonly used to explain the cluster red sequence in the local Universe, which predict cluster galaxy colours to become more blue at higher redshifts. Taking the full cluster population into account, I show that the formation of stars within the majority of massive cluster galaxies occurs over at least 2 Gyr, and peaks at z ~ 2–3. This is consistent with the cosmic star formation history, with star formation ending in clusters at 1 < z < 2. I further show that massive galaxies at z > 2 must have assembled within 0.5 Gyr of them forming a significant fraction of their stars. This means that few massive galaxies in z > 2 protoclusters could have formed via dry mergers. Some of the CARLA structures exhibit signs of being mature, collapsed clusters. In a pilot project, I report on the discovery of a z = 1.58 cluster with a strong red sequence around the RLAGN 7C 1753+6311. I demonstrate that the cluster has an enhanced quiescent galaxy fraction that is three times that of the control field. I also show that this enhancement is mass dependent: 91 +/- 9% of the M* > 10^10.5 Msun cluster galaxies are quiescent, compared to only 36 +/- 2% of field galaxies, whereas the fraction of quiescent galaxies with lower masses is the same in the cluster and field environments. This is in contrast to low redshift studies which have shown that mass and environmental effects on quenching star formation are separable. In the literature there is some debate as to whether RLAGN preferentially reside in clusters of a certain stage of collapse. The presence of a dense core and a well-formed, quiescent red sequence suggest that 7C1753+6311 resides within a mature cluster. This means that distant RLAGN do not solely reside in young, uncollapsed protoclusters, rather they can be found in clusters in a wide range of evolutionary states. Finally I present results from surveys of Halpha emitters in the fields around three high redshift RLAGN. I find that there is more dust-obscured star formation in protocluster galaxies than in similarly-selected control field galaxies at z ~ 2.5 and there is tentative evidence of a higher fraction of starbursting galaxies in the denser environment. However, on average I do not find a difference between the star formation rate (SFR)-mass relations of the protocluster and field galaxies and so conclude that the SFR of these galaxies at z ~ 2.5 is governed predominantly by galaxy mass and not the host environment. The stellar mass distribution of the protocluster galaxies is also skewed towards higher masses and there is a significant lack of low mass (M < 10^10 Msun) galaxies within the protocluster core. These results have implications for future protocluster surveys. The lack of low mass galaxies affects the level of overdensity which is detected. If only high mass galaxies are considered, the density of the protocluster field may be over-estimated. This means that it is important when quantifying protoclusters to compare their mass functions, rather than simply number overdensities. I also find that some radio galaxies do not reside in the centre, or densest region of the surrounding structure, meaning the overdensity measured in an aperture centred on the RLAGN will be underestimated. This means that future studies of (proto)clusters around RLAGN should use larger fields of view in order to establish the existence of a (proto)cluster.

In this thesis we have covered the importance of finding underlying conservation events to better understand the regulatory mechanisms of important development orchestrators like the Hox cluster. As an example of these non-evident conservation, we have shown two cases, as described below. The first case studied, after developing a software able to detect homologous long noncoding RNAs by means of microsynteny analyses, is the conservation of Hotairm1 in Chordata. For assessing the homology of this lncRNA, first we had to identify the lncRNA fraction within the B. lanceolatum transcriptome. With a reliable lincRNA dataset, we used our pipeline, LincOFinder, to identify orthologs between human and amphioxus through microsynteny. After the identification of Hotairm1 as one of the lincRNAs with conserved microsynteny, we used Xenopus as a proxy to analyse the homologies in the expression and the function. We had to proceed this way due to the difficulties associated with the inhibition of genes in B. lanceolatum, and the unavailability of expression patterns for Hotairm1 in the bibliography. After we successfully characterised Hotairm1 expression in amphioxus and Xenopus, we injected morpholino oligonucleotides to target and inhibit the splicing of Hotairm1 to promote an isoform imbalance. Through the phenotype obtained and the performing of qPCRs, we were able to deduct the mechanism of Hotairm1 and successfully relate this mechanism with the one described in human cells. With all the data obtained we were able to strongly suggest that the amphioxus Hotairm1 is homologous to the Xenopus and human Hotairm1, thus being conserved in most of the lineages within chordates. The second case studied was the conservation of the regulation of the Hox cluster mediated by Cdx. When analysing the B. floridae knockouts of Cdx and Pdx obtained using the TALEN technique, we found a severe phenotype of the developing larvae in Cdx-/- and a mild phenotype in Pdx-/-. The Cdx-/- phenotype consisted in the disruption of posterior gut development, as well as an underdevelopment of the postanal tail, coupled with a non-opening anus. When looking at changes in the expression of the Hox cluster in this Cdx-/- embryos, we found collinear misregulation of the expressed Hox genes, with the most anterior Hox cluster genes upregulated, and the most posterior ones downregulated. This is very similar to findings seen in triple morpholino knockdowns of the Cdx genes in Xenopus, indicating that in both, Xenopus and amphioxus, Cdx is regulating the Hox cluster through a homologous mechanism.

Klynge- og innovasjonsteorier har i de senere årene fått økt oppmerksomhet, noe som har bidratt til en oppblomstring av offentlig strategier for innovasjon og næringsutvikling. Klyngeinitiativ er blitt promotert som regionale utviklingsverktøy for offentlige myndigheter, med en forventning om at de skal påvirke konkurransekraften positivt gjennom økt produktivitet og innovasjonstakt. Næringsklynger defineres som “geografiske konsentrasjoner av tilknyttede selskaper, spesialiserte leverandører, serviceselskaper, selskaper fra relaterte industrier og omkringliggende institusjoner (som for eksempel universiteter, virkemiddelapparat og handelsorganisasjoner.” (Porter 2000:15) Denne studien argumenterer for at klynger initiert for å promotere regional utvikling, sjelden innfrir forventningene. Det er få utvetydige bevis for en sammenheng mellom initiert klyngesamarbeid og innovasjon. I slike miljøer er det stor sannsynlighet for at politiske egenskaper, i form av konkurrerende interesser og oppfatninger, fører til politisk oppførsel, slik som konflikter, strategiutvikling og maktspill. Fraværet av innovasjonsresultater i denne type samarbeid, kan potensielt forklares med den sterke tilstedeværelsen av politikk. Politikk påvirker feltet, koster tid og tar energi fra de involverte aktørene. Dette avleder fokus fra den opprinnelige intensjonen om samarbeid knyttet til næringsutvikling. Politikk må derfor tas med i betraktningen nå man snakker om, planlegger og gjennomfører klyngesamarbeid. Debatter om konkrete resultater av klyngestrategier og regionale klyngeinitiativ er nesten fraværende i den offentlige debatten. Den presenterte casestudien avslører et felt hvor politikk og rivalisering mellom ulike organisasjoner og konkurrerende klyngeinitiativer er sterkt tilstedeværende. Over tid, og som følge av åpenbar konkurranse, ser det ut til at enkelte klyngeinitiativ vokser, noen slår seg sammen, mens en tredje gruppe tvinges til å legge ned. I denne studien kalles dette fenomenet ”klyngekannibalisme”. Siden politikk dominerer både etablering og utvikling av klynger, ser klyngestrategiene ut til å ha blitt strategier for å organisere det politiske innovasjonsfeltet, snarere enn et verktøy for å øke innovasjonstakten
The popularization of cluster and innovation theories has led to the development of public strategies for innovation enhancement. Cluster initiatives have been promoted as a regional development tool for public governments, proposed to lead to innovation and economic development. Clusters can be defined as “geographical concentrations of interconnected companies, specialized suppliers, service providers, firms in related industries, and associated institutions (e.g. universities, standards agencies, trade associations) in a particular field that compete but also cooperate.” (Porter, 2000:15) They are expected to affect competition by increasing productivity and drive innovation through business development. This study argues that cluster cooperation initiated to promote regional innovation seldom is successful in doing so. There is little unambiguous evidence for correlation between initiated clusters and innovation. In cluster environments, political features of actors in the form of diverging interests and perceptions often lead to political behavior such as conflicts, strategy development and power play. The lack of innovative output of initiated cluster efforts may be explained by the strong orientation towards political features and behavior. Politics influence the field, consume time and energy of involved stakeholders and deflect the original intention of cooperation related to business development. Consequently, politics must be taken into account when speaking of, planning and carrying out cluster cooperation. Debates on the outcomes of cluster strategies and cluster initiation have been almost absent in the public sphere. The case presented reveals a field where politics and rivalry both among inter-agencies and competing cluster initiatives are present. Eventually, some initiatives are enlarged, some join forces, while others die out. In this study, these phenomena are called cluster cannibalism. As politics come to dominate clusters, cluster strategies seem to have become strategies for organizing the policy field of innovation, rather than a way of actually achieving innovation.

A central question in evolutionary biology is how organisms evolve highly derived and novel morphologies. More specifically, what changes to conserved developmental genes lead to the evolution of divergent morphologies? Here, I investigate the genetic and genomic changes to the developmentally important Hox genes using comparative genomics, gene expression and gene editing approaches. Hox genes code for homeodomain transcription factors that are responsible for determining the body plan of an embryo along the anterior-posterior axis, and changes to these genes have paralleled the rise of morphological diversity in the vertebrate animals. I focus my studies in a group of fish that exhibit a striking departure from the typical fish body plan: the pipefish and seahorse family, Syngnathidae. The evolution of syngnathid fish involved major modifications to their vertebrate body plan, but the developmental genetic basis of those changes is largely unknown. I describe the genomic organization of Hox clusters in a species of syngnathid pipefish—the Gulf pipefish (Syngnathus scovelli). I present an initial investigation on phenotypic consequences to the loss of hox7 genes in teleost fish—a group of Hox genes that are missing in syngnathids—using of the CRISPR/Cas9 system to induce indels in all hox7 genes (hoxa7a, hoxb7a) in the threespine stickleback (Gasterosteus aculeatus). In the second half of my thesis, I investigate noncoding changes in the syngnathid Hox clusters. I use syngnathid representative species and compared their conserved noncoding sequences within the Hox clusters to other teleost fish, non-teleost fish, and non-fish vertebrates. I present a detailed study regarding the nature of the loss of one conserved non-coding element. Results from this research indicate that the divergent syngnathid body plan is not due to rampant change in throughout Hox clusters. Also, these data do not argue for the absence of any role of genetic changes in Hox clusters. Instead, the findings presented here support the intermediate hypothesis that certain key changes to the Hox genes, microRNAs, and regulatory elements have probably contributed to their body plan developmental evolution in this unique family of fish. This work includes published co-authored material.

Many clusters of galaxies contain large quantities of hot diffuse gas. I have studied the properties of waves in this gas using Lagrangian perturbations. The gas is far more thermally stable than is commonly thought. For bremsstrahlung cooling, all modes that remain oscillatory are damped. Galaxy motions, especially the oscillations of a central cD galaxy, are an important way of generating large amplitude waves in cluster gas. This is especially pertinent in view of the growing realization that cD galaxies are not at rest with respect to the cluster. I also present evidence that weak magnetic fields tangled on scales of ~ 10 kpc are common in cluster gas. Electrons responsible for the flow of heat in the gas must travel along the field lines, leading to a global reduction in heat flux. The superposition of many different field lines implies that the cluster gas is a multiphase medium. Such a picture has been suggested independently by the claimed observation of mass drop-out from cooling flows. I also show some results from a more advanced study using real magnetic fields rather than random walk models.

Now that the genome has successfully been sequenced, the next aim will be to characterize and annotate features. With the development of high resolution tools, such as Chromosome Conformation Capture (3C) technology, the 3-dimensional architecture of large linear spans of DNA can be determined. We have established that the mouse HoxA gene cluster is structured in such a way as to contain four distinct looping regions with a central interacting rosette core in its resting pluripotent state. This central core feature might directly affect the spatiotemporal regulation of the Hox genes during normal development as well as differentiation via retinoic acid administration in the P19 embryonal carcinoma cell line. The DNA looping might be mediated by cis-regulatory elements, which could function in a cooperative manner to regulate HoxA gene expression. Regulatory elements might include the previously identified retinoic acid response elements (RAREs) known to mediate specific developmental cues. Interestingly, Hox genes expressed earlier and more anteriorly are within loops containing higher amounts of identified RAREs as well as other conserved non-coding sequences that also appear to play a potential role in orchestrating the collinear mechanism of action. Loops containing the more 5' located genes have fewer regulatory elements and therefore may respond later to developmental cues.
Maintenant que le génome été séquencé avec succès, le but suivant sera de caractériser et annoter les caractéristiques regulataire. Avec le développement d'instruments de haute résolution, comme la technologie du Capture de Conformation Chromatin (3C), l'architecture 3- dimensionnelle de grandes durées linéaires d'ADN peut être déterminée. Nous avons établi que le groupe de gène de HoxA dans la souris est structurée d'une telle façon pour contenir quatre régions boucles distinctes avec un centre de rosette réagissant réciproquement central dans son état de reposant pluripotent. On croit que cette caractéristique centrale de base directement affecte le règlement spatiotemporel des gènes Hox pendant le développement normal aussi bien que la différentiation via l'administration acide rétinoïque dans le P19 ligne de cellule de carcinome. Les éléments cis-de-contrôle responsables de définir ces boucles ont l'air de fonctionner dans une manière coopérative dans laquelle le pluripotent HoxA l'état est plein d'assurance de réagir à l'administration acide rétinoïque dans une manière colinéaire contrôlée d'expression de gène. Auparavant identifié les éléments de réponse acide rétinoïque (ERARs) sont organisés pour tenir compte de la réorganisation dynamique du groupe sur la réception de signaux spécifiques du développement. Les gènes de Hox ont exprimé tôt et sont plus antérieurement dans les boucles qui contiennent de plus hautes quantités de ERARs identifié aussi bien que d'autres ordres de non-codification conservés qui ont aussi l'air de jouer un rôle potentiel dans le fait d'orchestrer le mécanisme colinéaire d'action.

With the advancement of Internet technology and the changes in the mode of communications, it is found that much first-hand news have been discussed in Internet forums well before they are reported in traditional mass media. Also, this communication channel provides an effective channel for illegal activities such as dissemination of copyrighted movies, threatening messages and online gambling etc. The law enforcement agencies are looking for solutions to monitor these discussion forums for possible criminal activities and download suspected postings as evidence for investigation. The volume of postings is huge, for 10 popular forums in Hong Kong; we found that there are 300,000 new messages every day. In this thesis, we propose an automatic system that tackles this problem. Our proposed system downloads postings from selected discussion forums continuously and employs data mining techniques to identify hot topics and cluster authors into different groups using word based user profiles. Using these data, we try to locate some useful trends and detect crime from the data, the result is discussed afterward with include advantages and limitations of different approaches and at the end, there is a conclusion of the way to solve those problems and provide future direction of this research.
published_or_final_version
Computer Science
Master
Master of Philosophy

Nowadays, clusters of computers are used to solve computation intensive problems. These clusters take advantage of a large number of computing nodes to provide a high degree of parallelization. Interconnection networks are used to connect all these computing nodes. The interconnection network should be able to efficiently handle the traffic generated by this large number of nodes. Interconnection networks have different design parameters that define the behavior of the network. Two of them are the topology and the routing algorithm. The topology of a interconnection network defines how the different network elements are connected, while the routing algorithm determines the path that a packet must take from the source to the destination node. The most commonly used topologies typically follow a regular structure and can be classified into direct and indirect topologies, depending on how the different network elements are interconnected. On the other hand, routing algorithms can also be classified into two categories: deterministic and adaptive algorithms. To evaluate interconnection networks, metrics such as latency or network productivity are often used. Throughput refers to the traffic that the network is capable of accepting the network per time unit. On the other hand, latency is the time that a packet requires to reach its destination. This time can be divided into two parts. The first part is the time taken by the packet to reach its destination in the absence of network traffic. The second part is due to network congestion created by existing traffic. One of the effects of congestion is the so-called Head-of-Line blocking, where the packet at the head of a queue blocks, causing the remaining queued packets can not advance, although they could advance if they were at the head of the queue. Nowadays, there are other important factors to consider when interconnection networks are designed, such as cost and fault tolerance. On the one hand, a high performance is desirable, but without a disproportionate increase in cost. On the other hand, the fact of increasing the size of the network implies an increase in the network components, thus the probability of occurrence of a failure is higher. For this reason, having some fault tolerance mechanism is vital in current interconnection networks of large machines. Putting all in a nutshell, a good performance-cost ratio is required in the network, with a high level of fault-tolerance. This thesis focuses on two main objectives. The first objective is to combine the advantages of the direct and indirect topologies to create a new family of topologies with the best of both worlds. The main goal is the design of the new family of topologies capable of interconnecting a large number of nodes being able to get very good performance with a low cost hardware. The family of topologies proposed, that will be referred to as k-ary n-direct s-indirect, has a n dimensional structure where the k different nodes of a given dimension are interconnected by a small indirect topology of s stages. We will also focus on designing a deterministic and an adaptive routing algorithm for the family of topologies proposed. Finally we will focus on analyzing the fault tolerance in the proposed family of topologies. For this, the existing fault tolerance mechanism for similar topologies will be studied and a mechanism able to exploit the features of this new family will be designed. The second objective is to develop routing algorithms specially deigned to reduce the pernicious effect of Head-of-Line blocking, which may shoot up in systems with a high number of computing nodes. To avoid this effect, routing algorithms able of efficiently classifying the packets in the different available virtual channels are designed, thus preventing that the occurrence of a hot node (Hot-Spot) could saturate the network and affect the remaining network traffic.
Hoy en día, los clústers de computadores son usados para solucionar grandes problemas. Estos clústers aprovechan la gran cantidad de nodos de computación para ofrecer un alto grado de paralelización. Para conectar todos estos nodos de computación, se utilizan redes de interconexión de altas prestaciones capaces de manejar de forma eficiente el tráfico generado. Estas redes tienen diferentes parámetros de diseño que definen su comportamiento, de los cuales podríamos destacar dos: la topología y el algoritmo de encaminamiento. La topología de una red de interconexión define como se conectan sus componentes, mientras que el algoritmo de encaminamiento determina la ruta que un paquete debe tomar desde su origen hasta su destino. Las topologías más utilizadas suelen seguir una estructura regular y pueden ser clasificadas en directas e indirectas, dependiendo de cómo estén interconectados los diferentes elementos de la red. Por otro lado, los algoritmos de encaminamiento también pueden clasificarse en dos categorías: deterministas y adaptativos. Para evaluar estas redes se suelen utilizar medidas tales como la latencia o la productividad de la red. La productividad mide el tráfico que es capaz de aceptar la red por unidad de tiempo. La latencia mide el tiempo que utiliza un paquete para alcanzar su destino. Este tiempo se puede dividir en dos partes. La primera corresponde al tiempo utilizado por el paquete en alcanzar a su destino en ausencia de tráfico en la red. La segunda sería la debida a la congestión de la red creada por el tráfico existente. Uno de los efectos de la congestión es el denominado Head-of-Line blocking, donde el paquete que encabeza una cola se queda bloqueado, por lo que el resto de paquetes de la cola no pueden avanzar, aunque pudieran hacerlo si ellos encabezaran dicha cola. Otros factores a tomar en cuenta son el coste y la tolerancia a fallos. Las prestaciones deben mantenerse conforme aumentamos el tamaño de la red, pero sin un aumento prohibitivo en el coste. Además, el hecho de aumentar el tamaño de la red implica un aumento en el número de elementos de dicha red, aumentando la probabilidad de la aparición de un fallo. Por ello, es vital contar con algún mecanismo de tolerancia a fallos en las redes para los grandes supercomputadores actuales. En otras palabras, es de esperar una buena relación coste-prestaciones con un alto nivel de tolerancia a fallos. Esta tesis tiene dos objetivos principales. El primer objetivo combina las ventajas de las topologías directas e indirectas para crear una nueva familia de topologías con lo mejor de ambas. En concreto, nos centramos en el diseño de una nueva familia de topologías capaz de interconectar una gran cantidad de nodos siendo capaz de obtener muy buenas prestaciones con un bajo coste hardware. La familia de topologías propuesta, que hemos llamado k-ary n-direct s-indirect, tiene una estructura n-dimensional, donde los diferentes k nodos de una dimensión se conectan entre sí mediante una pequeña topología indirecta con s etapas. También diseñaremos un algoritmo de encaminamiento determinista y otro adaptativo para la familia de topologías propuesta. Finalmente, nos centraremos en estudiar la tolerancia a fallos para la familia de topologías propuesta. Para ello se estudiarán los mecanismos de tolerancia a fallos existentes en topologías similares y se diseñará un mecanismo capaz de aprovechar al máximo las características de esta nueva familia. El segundo objetivo consiste en el desarrollo de algoritmos de encaminamiento capaces de evitar el pernicioso efecto Head-of-Line blocking, lo cual puede aumentar rápidamente en sistemas con un gran número de nodos de computación. Para evitar este efecto se diseñarán algoritmos de encaminamiento capaces de clasificar de forma eficiente los paquetes en los diferentes canales virtuales disponibles, evitando así que la aparición de un punto caliente (Hot-Spot) sat
Hui en dia, els clústers de computadors són utilitzats per solucionar grans problemes computacionals. Aquests clústers aprofiten la gran quantitat de nodes de computació per a oferir un alt grau de paral·lelització. Per a connectar tots aquests nodes de computació, s'utilitzen xarxes d'interconnexió d'altes prestacions capaços de manejar de manera eficient el trànsit generat. Aquestes xarxes tenen diferents paràmetres de disseny que defineixen el seu comportament, dels quals podríem destacar dues: la topologia i l'algoritme d'encaminament. La topologia d'una xarxa d'interconnexió ens defineix com es connecten els seus components, mentre que l'algoritme d'encaminament determina la ruta que un paquet ha de prendre des del seu node origen fins al seu node destí. Les topologies més utilitzades solen seguir una estructura regular i poden ser classificades en directes i indirectes, depenent de com estiguen interconnectats els diferents elements de la xarxa. D'altra banda, els algoritmes d'encaminament també poden classificar-se en dues categories: deterministes i adaptatius. Per avaluar estes xarxes es solen utilitzar mesures com ara la latència o la productivitat de la xarxa. La productivitat mesura el trànsit que és capaç d'acceptar la xarxa per unitat de temps. La latència mesura el temps que utilitza un paquet per arribar al seu destí. Aquest temps es pot dividir en dues parts. La primera correspon al temps emprat pel paquet a aconseguir al seu destí en absència de trànsit a la xarxa. La segona part seria la deguda a la congestió de la xarxa creada per el trànsit existent. Un dels efectes de la congestió és l'anomenat Head-of-line blocking, on el paquet que encapçala una cua es queda bloquejat, de manera que la resta de paquets de la cua no poden avançar, encara que poguessen fer-ho si ells encapçalessen la dita cua. Altres factors a tenir en compte són el cost i la tolerància a fallades. Per tant, les prestacions s'han de mantenir d'acord augmentem la mida de la xarxa, però sense un augment prohibitiu en el cost. A més, el fet d'augmentar la mida de la xarxa implica un augment en el número de elements d'aquesta xarxa, de manera que la probabilitat de l'aparició d'una fallada és més gran. Per això, és vital comptar amb algun mecanisme de tolerància a fallades en les xarxes d'interconnexió per als gran supercomputadors actuals. En altres paraules, és d'esperar bona relació cost-prestacions amb una alta tolerància a fallades. Aquesta tesi té dos objectius principals. El primer objectiu combina les avantatges de les topologies directes i indirectes per a crear una nova família de topologies amb el millor dels dos mons. En concret, ens centrem en el disseny de una nova família de topologies capaç d'interconnectar una gran quantitat de nodes sent capaç d'obtenir molt bones prestacions amb un baix cost hardware. La família de topologies proposada, que hem nomenat k-ary n-direct s-indirect, té una estructura n-dimensional, on els diferents k nodes d'una dimensió se connecten entre si mitjançant una petita topologia indirecta amb s etapes. També dissenyarem un algoritme d'encaminament determinista i un altre adaptatiu per a la família de topologies proposta. Finalment, ens centrarem en estudiar la tolerància a fallades per a la família de topologies proposada. Per a això s'estudiaran els mecanismes de tolerància a fallades existents en topologies similars i es dissenyarà un mecanisme capaç d'aprofitar al màxim les característiques d'aquesta nova família. El segon objectiu consisteix en la creació d'algoritmes d'encaminament capaços d'evitar el perniciós efecte Head-of-line blocking que pot créixer ràpidament amb un gran número de nodes de computació. Per a evitar aquest efecte es dissenyaran algoritmes d'encaminament capaços de classificar de forma eficient els paquets en els diferents canals virtuals disponibles, evitant així que l'aparició d'un punt calent (
Peñaranda Cebrián, R. (2017). Improvement of interconnection networks for clusters: direct-indirect hybrid topology and HoL-blocking reduction routing [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/79550
TESIS

En raison de l’essor rapide des technologies de communication sans fil et de la demande croissante de services dans les réseaux ad hoc véhiculaires (VANETs), le besoin d’algorithmes de clustering efficaces est nécessaire pour résoudre le problème de mise à l’échelle et le besoin croissant d'applications dans les VANETs. Cependant, à cause de la nature dynamique des VANETs, les algorithmes de clustering doivent assurer la stabilité du cluster. Pour cela, l’objectif de la thèse est de proposer un cadre générique servant à définir des algorithmes de clustering afin d'améliorer la stabilité du cluster et d'augmenter l'efficacité de la transmission de données. Dans cette thèse, nous introduisons un algorithme de clustering (MoDyC) qui se base sur la métrique de mobilité pour former des clusters. Ensuite, nous proposons un cadre générique servant à définir des algorithmes de clustering appelé UFC. Nous évaluons l’impact des métriques de clustering et des méthodes de fusion de clusters sur le critère de stabilité. Nous proposons une nouvelle méthode de fusion de cluster appelée LCM. Cette méthode basée sur la métrique Leadership permet d’augmenter la stabilité du cluster. Nous analysons les performances des algorithmes de clustering proposés. Dans le but de modéliser le processus de clustering, nous proposons un modèle basé sur les chaînes de Markov. A partir de ce modèle, nous analysons la stabilité du cluster. Cette thèse se termine par l’étude d’un scénario de diffusion de messages d'urgence en utilisant un schéma de diffusion de données en cluster avec cache de contenu
Due to the rapid development of wireless communication technologies and the growing demand of services in Vehicular Ad-hoc Networks (VANETs), efficient clustering algorithms are mandatory to solve the network scalability problem and to support more applications in VANETs. However, because of the dynamic network topology of VANETs, clustering algorithms are required to guarantee the cluster's stability. Therefore, this thesis aims to propose a framework for clustering algorithms for VANETs, to improve cluster's stability and to increase the efficiency of data transmission. In this thesis, we firstly investigate a new mobility-based scheme for dynamic clustering (MoDyC) in VANETs using the mobility-based clustering metric to construct clusters. Then, we propose our framework for clustering algorithm named UFC. Based on this framework, we evaluate the impacts of different clustering metrics and cluster merging schemes on cluster's stability. Meanwhile, a leadership-based cluster merging scheme (LCM) is presented to increase the cluster stability. To analyze the cluster stability, we propose a Markov chain model to model the clustering process and to investigate its performance. Finally, this thesis studies a scenario, aiming to disseminate vehicular emergency messages through cluster-based data dissemination scheme

Class I Homeobox genes (Hox) are an evolutionary conserved family of genes that encode master regulators in developmental processes. Hox gene expression has been shown to be high in normal haematopoietic stem/progenitor cells (HSPCs) and decreases during normal differentiation, however elevated Hox expression remains in subtypes of leukaemia. Polycomb repressor complexes (PRCs) and histone modifiers, e.g. Mixed Lineage Leukaemia (MLL), are key regulators of Hox expression. MLL rearrangements, frequent in acute leukaemia, are associated with high HOXA expression. However, the necessity of the Hoxa cluster in MLL-Ieukaemia maintenance is not fully elucidated. Ectopic overexpression of MLL-AF9 in HSPCs in conditional compound transgenic mouse backgrounds MxCre+/Hoxafloxlflox (MAFF) or Hoxafloxlflox (Aflox) models resulted in increased colony formation and growth in liquid culture. Transformed colonies, serially replated in methylcellulose and transplanted into sub-lethally irradiated recipient mice, resulted in primary leukaemia. Direct treatment of MLL-AF9 leukaemias generated in the MAFF background (MAFF-MA9) with interferon-alpha (IFN)resulted in further deletion of the Hoxa cluster and significant reduction in colony formation compared to controls. Although non-leukaemic MAFF HSPCs retained colony forming ability after complete Hoxa cluster deletion (Hoxa-/-), no Hoxa-/- colonies were recovered from the IFN treated MAFF-MA9 cultures. Cre-recombinase induced deletion of the Hoxa cluster from Aflox-MA9 leukaemia cells was confirmed by gDNA-PCR and sequencing. Transplantation of Cre-treated Aflox-MA9 cells resulted in significant increased survival (p