Dissertations / Theses on the topic 'Domains of topological association'

Le génome eucaryote est organisé dans l’espace et le long de sa séquence. Du chromosome aux gènes, son organisation 3D est étroitement liée à son activité transcriptionnelle. A une échelle inférieure à la mégabase, le génome est organisé physiquement en domaines d’auto-interaction (TAD, pour topologically associating domains), que l’on pense influencer l’action de séquences régulatrices des gènes, appelées ‘amplificateurs’. Les TAD sont parfois vus comme des ‘unités de régulation’ permettant de coréguler plusieurs gènes en les exposant aux mêmes amplificateurs. L’expression de gènes d’un même TAD est en effet souvent corrélée entre tissus et types cellulaires. On sait d’autre part que l’expression de gènes fonctionnellement liés est aussi souvent corrélée. Cependant, le rôle que joue l’organisation 3D d’un locus génomique dans la corégulation de gènes fonctionnellement liés reste mal compris. Par des techniques d’imagerie de molécules uniques, j’ai observé, dans des cellules individuelles, la position et la transcription de trois gènes adjacents et fonctionnellement liés, partageant –ou non– les mêmes amplificateurs. J’ai utilisé comme système modèle la réponse à l’œstrogène dans des cellules MCF7. En combinant les techniques de FISH ARN et ADN, j’ai mesuré le couplage transcriptionnel de paires de gènes en cis, mettant en évidence une augmentation de la corrélation entre gènes d’un même TAD en réponse à l’œstrogène. Une perturbation de la frontière entre TAD indique également le rôle des contacts génomiques dans la corégulation de gènes. Ces travaux posent les bases pour comprendre comment les amplificateurs et les gènes communiquent et coordonnent leur activité en 3D
The eukaryotic genome is highly organized in both space and sequence. From entire chromosomes to individual genes the 3D organization of the genome is linked to transcription and many regulatory mechanisms likely coexist at different scales. At the sub-megabase scale, the genome is physically organized into self-interacting topologically associating domains (TADs) that are thought to constrain the range of action of gene regulatory elements called ‘enhancers’. Current data suggest that TADs serve as ‘regulatory units’ to coregulate multiple genes by exposing them to the same enhancers. Genes from the same TAD indeed often display correlated expression across different tissues and cell types. Interestingly, correlated expression is seen between functionally related genes. However, how 3D organization at an individual locus plays a mechanistic role in coregulating functionally related genes is unknown. Using single-molecule imaging, I observed in single cells the spatial positions and transcription of three adjacent functionally related genes regulated by the same/different enhancers. I used estrogen stimulation in MCF7 cells as a model system to study hormone-responsive genes and enhancers. Using combined RNA-DNA FISH, I measured the coupling between genes as the correlation in cis of their transcription. I found, that stimulation with estrogen increases the correlation in cis between genes belonging to the same TAD. Perturbation of the TAD boundary revealed the contribution of contact insulation to gene coregulation. Together, this work lays the ground towards an understanding of how enhancers and genes communicate and coordinate their activity within the 3D genome

Die Netzwerktheorie ist eine wirksame Methode, um die Struktur realer Systeme, z.B. des Klimasystems, zu beschreiben und zu klassifizieren. Der erste Teil dieser Arbeit nutzt diese Diskriminanzfähigkeit um die Ost- und Zentralpazifischen Phasen von El Niño und La Niña mittels eines Index basierend auf der Evaluation zeitlich entwickelnder Klimanetzwerke zu unterscheiden. Nach dem Studium der klimatischen Einflüsse dieser unterschiedenen Phasen verlegt die Arbeit ihren Schwerpunkt von der Klassifikation einzelner klimatischer Schichten auf den generelleren Fall interagierender Netzwerke. Hier repräsentieren die Teilnetzwerke entsprechende Variabilitäten in Ozean und Atmosphäre. Es zeigt sich, dass die Ozean-Atmosphären-Wechselwirkung einer hierarchischen Struktur folgt wobei makroskopische Netzwerkmaße einzelne Atmosphärenschichten bezüglich ihrer Wechselwirkung mit dem Ozean unterscheiden. Der zweite Teil dieser Arbeit untersucht den Einfluss der räumlichen Einbettung von Knoten auf topologische Netzwerkeigenschaften. Hierzu werden Nullmodelle eingeführt, welche zufällige Surrogate eines gegebenen Netzwerks erzeugen, sodass globale und lokale räumliche Eigenschaften erhalten bleiben. Diese Modelle erfassen die makroskopischen Eigenschaften der studierten Netzwerke besser als bisherige Standardmodelle zur Erzeugung von Zufallsnetzwerken. Abhängig von der Performanz der vorgeschlagenen Modelle können gegebene Netzwerke schlussendlich in verschiedene Klassen eingeteilt werden. Die Arbeit schließt mit einer Erweiterung der bisherigen Netzwerkklassifikatoren um eine zweidimensionale Metrik, welche Netzwerke auf Basis ihrer Komplexität unterscheidet. Es wird gezeigt, dass Netzwerke des gleichen Typs dazu neigen in individuellen Bereichen der resultierenden Komplexitäts-Entropie-Ebene zu liegen. Die eingeführte Methode ermöglicht auch die objektive Konstruktion von Klimanetzwerken indem Schwellwerte gewählt werden, die die statistische Komplexität maximieren.
Complex network theory provides a powerful tool to quantify and classify the structure of many real-world complex systems, including the climate system. In its first part, this work demonstrates the discriminative power of complex network theory to objectively classify Eastern and Central Pacific phases of El Niño and La Niña by proposing an index based on evolving climate networks. After an investigation of the climatic impacts of these discriminated flavors, this work moves from the classification of sets of single-layer networks to the more general study of interacting networks. Here, subnetworks represent oceanic and atmospheric variability. It is revealed that the ocean-to-atmosphere interaction in the Northern hemisphere follows a hierarchical structure and macroscopic network characteristics discriminate well different parts of the atmosphere with respect to their interaction with the ocean. The second part of this work assesses the effect of the nodes’ spatial embedding on the networks’ topological characteristics. A hierarchy of null models is proposed which generate random surrogates from a given network such that global and local statistics associated with the spatial embedding are preserved. The proposed models capture macroscopic properties of the studied spatial networks much better than standard random network models. Depending on the models’ actual performance networks can ultimately be categorized into different classes. This thesis closes with extending the zoo of network classifiers by a two-fold metric to discriminate different classes of networks based on assessing their complexity. Within this framework networks of the same category tend to cluster in distinct areas of the complexity-entropy plane. The proposed framework further allows to objectively construct climate networks such that the statistical network complexity is maximized.

Members of the RASSF family (RASSF1-10) have been identified as candidate tumour suppressors that are frequently downregulated by promoter hypermethylation in cancers. These adaptor proteins carry a common Ras-association (RA) and SARAH domain (RASSF1-6) that can potentially bind Ras oncoproteins and mediate protein-protein interactions with other SARAH domain proteins (e.g. MST kinase). However, there is a notable lack of comparative characterisation of the RASSF family, as well as of molecular and structural information that facilitate their tumour suppressive functions. As part of our comparative analysis, we modelled the RA and SARAH domains of the RASSF members based on existing structures and predicted their potential interactions and the key residues involved. These in silico predictions were compared to in vitro studies and intracellular binding assays using Förster Resonance Energy Transfer (FRET). Several SARAH domain mutants were also investigated for their effects on RASSF interactions. Furthermore, we compared the interactions of the RASSF family with several key proteins involved in death and NFκB signalling. Our biochemical data show a diversity of interactions within the RASSF family RA domain, whereas interactions between RASSF and MST correlate with the presence of the SARAH domain, which is supported by the FRET experiments. Mutations of specific non-polar residues in the dimerisation interface of the SARAH domain also prove detrimental to the interaction between selected RASSF members and MST. Moreover, we observed stimulation-dependent interactions between specific RASSF members and MOAP1, TNF-R1, DAPK and TBK1. These results suggest that different members, despite shared general architecture, may have distinct binding properties, but ultimately could share overlapping functions. Current data also support an interaction model where RASSF serves as an adaptor for the assembly of multiple protein complexes and further functional interactions, involving MST kinases and other interacting partners, which could be regulated by Ras.

The baculovirus Autographa californica nucleopolyhedrovirus (AcNPV) has two viral forms, budded virus (BV) and occlusion derived virus (ODV). The envelopment of these two viral forms occurs at different locations: BV acquires envelopes at the plasma membrane while ODV acquires envelopes in the nucleus. The two viral forms carry out different functions in the viral life cycle. The purpose of this study is to investigate how viral envelope proteins sort/traffic to the nucleus. Of particular interest is BV/ODV E26 (E26). E26 is an envelope protein of both BV and ODV (Braunagel and Summers, 1994); therefore it must traffic to the plasma membrane and the nucleus during infection. Thus, E26 is a bi-directional trafficking protein, which interacts with membranes in both locations of the cell. As such it has been shown that there are several immunoreactive forms of E26 (Beniya, Braunagel, and Summers, 1998). The da26 gene produces at least 2 protein products of 26 and 28 kDa with different functions respectively, which correlate with localization, solubility, membrane association, and temporal requirements. The 28 kDa form is likely a soluble protein that interacts with transcriptional activators and DNA in the nucleus in the early stages of infection. A part of the 26 kDa population is a membrane bound form interacting with an integral membrane protein in the ER and likely functions as an INM sorting factor. The 26 kDa membrane bond form is also found in the inner nuclear membrane, intra-nuclear microvesicles, ODV envelopes, and ODV in the nucleus.

La relation entre le repliement du génome et les processus cellulaires majeurs, notamment la transcription, la réparation de l’ADN et la réplication est une question biologique centrale. De nouvelles méthodes de la génomique ont révélé un niveau inconnu de l’architecture tridimensionnelle de la chromatine. A l’échelle en dessous de la mégabase, certaines séquences génomiques se trouvent préférentiellement à proximité les unes des autres formant ainsi des domaines topologiques associés (TAD). Les gènes situés dans le même TAD ont des propriétés épigénétiques similaires, et leur expression au cours de la différentiation semble corrélée, ce qui suggère un lien fort entre la structure de la chromatine et la transcription. Les TADs sont à leur tour séparés par des régions avec peu de contacts, appelées frontières, qui sont généralement occupées par des protéines dites isolatrices. Les déterminants de cette organisation chromatinienne particulière et ses implications fonctionnelles sont largement méconnus. Selon une hypothèse récente, les TADs seraient formés par des contacts entre les séquences des frontières, stabilisés par la formation de boucles via les protéines isolatrices. Les travaux présentés ici ont pour but d’étudier le rôle des protéines isolatrices dans le mécanisme de formation des TADs chez la drosophile. La microscopie super-résolue a été implémentée et une série de développements ont été réalisés en microscopie à illumination structurée (SIM) et la microscopie à localisation de molécules uniques (SMLM), avec une attention particulière sur le marquage fluorescent. Ces développements ont directement été appliqués à l’étude de l’organisation nucléaire de la protéine associée aux éléments frontières (BEAF), une des 11 protéines isolatrices identifiées à ce jour chez la drosophile. Le fort enrichissement aux frontières des TADs, ainsi que son activité dans la formation de boucles d’ADN font de BEAF un candidat intéressant pour tester l’hypothèse de regroupement de frontières comme mécanisme général de repliement de la chromatine. La technique SMLM multi-couleurs a systématiquement localisé BEAF à la périphérie des larges domaines portant la marque épigénétique H3K27me3. L’analyse quantitative des images SMLM a révélé que BEAF forme des centaines de foyers d’une taille de 45 nm, composés en moyenne de 5 molécules, ce qui est en désaccord avec la présence de boucles de chromatine à large échelle. Afin de tester le regroupement de gènes directement au niveau de l’ADN, des frontières ont été marquées par des oligonucléotides fluorescents. Le nombre de foyers détectés par SIM s’est à nouveau révélé incompatible avec le modèle de contacts entre les frontières tout le long du génome. Par ailleurs, les distances entre paires de frontières au niveau de deux régions génomiques ont montré <5% de contacts. Ensemble, ces résultats sont en désaccord avec l’établissement d’interactions entre barrières chez la drosophile. Enfin, ces travaux de thèse ont contribué au développement méthodologique de la microscopie super-résolue, ce qui a permis d’apporter des preuves expérimentales invalidant le modèle de regroupement des frontières comme mécanisme général du repliement chromatinien
The interplay between genome folding and key cellular functions such as transcription, DNA repair and replication is a fundamental question in chromatin biology. Recent genome-wide developments unveiled a new level of three-dimensional chromatin architecture. At the sub-megabase scale, some genome sequences are preferentially found in proximity with one another forming Topologically Associating Domains (TADs). Genes located within the same TAD display common epigenetic properties and tend to have coordinated dynamics of expression during differentiation, suggesting a strong link between chromatin structure and transcription. TADs are in turn separated by regions of low contact, termed TAD borders, which are generally occupied by factors called chromatin insulators. What are the determinants of this particular type of chromatin organization and what are the functional implications is still largely unknown. In an emerging hypothesis, TADs could be formed through contacts between TAD border sequences stabilized by the looping activity of insulator proteins. This thesis investigates the roles of insulator proteins in the TAD formation mechanism using Drosophila melanogaster as model system. Superresolution imaging was implemented and a series of developments were performed in Structured Illumination Microscopy (SIM) and Single-molecule Localization Microscopy (SMLM), with particular attention on fluorescent labeling for single-molecule detection. These developments were directly applied to study the nuclear organization of the Boundary Element Associated Factor (BEAF), one of the 11 insulator proteins discovered to date in Drosophila. The strong enrichment on TAD borders and its demonstrated looping activity make BEAF a potent candidate to test for the clustering of TAD borders as a general mechanism of chromatin folding. Multicolor SMLM systematically located BEAF foci at the periphery of large H3K27me3 chromatin domains. Quantitative analysis of SMLM images indicated BEAF forms hundreds of 45 nm foci, containing a mean of 5 molecules, which argues against a large-scale looping of BEAF-bound chromatin. To directly probe for gene clustering at the DNA level, TAD borders were labeled using fluorescent oligonucleotide probes. The number of foci detected by SIM was once more incompatible with a model of chromosome-wide contacting of multiple TAD borders. Furthermore, TAD border pairs distances were measured in two genomic regions, resulting in <5% of paired contacts among the measured barriers. Taken together, these results are inconsistent with constitutive interactions between consecutive or non-consecutive barriers in Drosophila.In conclusion, this study contributed to the methodological development of super-resolution microscopy which was applied to provide experimental evidence invalidating the TAD border clustering model as a general mechanism of chromatin folding

La chromatine sert de support à de multiples processus biologiques, cependant son organisation spatiale diffère fortement selon l'échelle considérée. L'expression des gènes est ainsi coordonnée par des éléments régulateurs dispersés dans le génome mais capables d'interagir entre eux. Chez les métazoaires, des expériences de capture de conformation de chromosome (3C) combinées au séquençage haut-débit (Hi-C) ont permis la découverte de domaines d'association topologique (TAD), à l'échelle de la mégabase. Puisque la résolution du Hi-C reste limitée, nous avons utilisé la 3C-qPCR pour explorer, dans des cellules souches embryonnaires murines, la dynamique chromatinienne à l'intérieur de ces domaines ainsi qu'à leurs bordures. Nous identifions ainsi une modulation des fréquences de contacts, sur quelques centaines de kilobases. Cette modulation est plus ou moins importante en fonction du contenu en gènes des domaines, mais elle semble néanmoins universelle. Des modèles dérivés de la physique des polymères permettent de décrire cette modulation sous la forme d'une hélice statistique, que la chromatine adopterait en moyenne et en l'absence d'interactions spécifiques, à l'intérieur des TAD. Cette hélice reflète certaines contraintes que la chromatine subit à l'échelle supranucléosomale. Elle est très affectée par les bordures, qui bloquent la modulation, mais elle l'est beaucoup moins par le contenu en histone de liaison H1. Par ailleurs, grâce à des résultats de Hi-C à haute résolution, nous montrons que la modulation observée chez les souris n'est pas retrouvée chez la drosophile, où les caractéristiques des TAD semblent avant tout liées au paysage épigénétique local. Pour ces deux organismes, la dynamique chromatinienne à l'intérieur des domaines est donc sous le contrôle de phénomènes différents
The chromatin hosts various biological processes. However, its organization differs considerably depending on the scale. For example, gene expression is coordinated by regulatory elements that are dispersed in the genome but that are able to interact within the tridimensional space of the nucleus. In the Metazoa, chromosome conformation capture (3C) assays combined with high-throughput sequencing (Hi-C) uncovered the existence of topologically associating domains (TADs), at the mégabase scale. Due to the limited resolution of Hi-C, we used the 3C-qPCR method to explore, in murine embryonic stem cells, the chromatin dynamics inside TADs as well as at their borders. We found that contact frequencies undergo a periodic modulation over large genomic distances (few hundred kilobases). This modulation is weaker in gene-deserts than in gene-containing domains but it seems nevertheless to be universal. Using models derived from polymer physics, we show that this modulation can be understood as a fundamental helix shape that chromatin tends to adopt statistically, when no strong locus-specific interaction takes place, within the TADs. This statistical helix reflects some constraints that the chromatin undergoes at the supranucleosomal scale. It is affected by TADs borders, which disrupt the modulation, but linker histone H1 depletion only leads to subtle changes in the helix characteristics. Furthermore, using high-resolution Hi-C data, we found that chromatin dynamics is unconstrained in Drosophila where it seems mainly linked to the local epigenetics landscape. Therefore, distinct genome organization principles govern chromatin dynamics within mouse and Drosophila topologically associating domains

Chromatin ist ein Makromolekül, dessen Genregulation innerhalb des räumlich eingeschränkten Zellkerns organisiert werden muss. Die Genomorganisation ist eng mit Genaktivierung und Genrepression verknüpft. In den vergangenen Jahren wurde gezeigt, dass die DNA hierarchisch organisiert ist. Die Faltung läuft in aufeinander folgenden Schritten ab, wobei jede Organisationsebene sowohl zur räumlichen Komprimierung, als auch zur Genregulation beiträgt. In dieser Dissertation wurden mit Hilfe von hochauflösender Mikroskopie verschiedene Ebenen der 3D Chromatinorganisation auf Einzelzell-Basis untersucht. Auf der kleinsten Organisationsebene wurde die Struktur zweier, nebeneinander liegender topologischer Domänen (TADs) am Sox9-Lokus erforscht. Mit Hilfe von Fluoreszenz in situ Hybridisierung (FISH) in 3D Zellen, sowie Cryoschnitten in embryonalen Stammzellen von Mäusen konnten Interaktionen zwischen den benachbarten TADs festgestellt werden. FISH in Zellen mit genomischen Duplikationen, zeigte das Entstehen von zwei unterschiedlichen, durch die Duplikation entstandenen, Konformationen. Unter Verwendung von FISH wurden long-range Kontakte, die zuvor mit GAM entdeckt wurden, untersucht und es zeigte sich, dass sie häufig zwischen TADs die regulatorischen Domänen enthalten auftreten. Zudem zeigte sich die Bildung von Clustern zwischen mehreren, weit auseinander liegenden, regulatorischen Elementen. Dies lässt unter Umständen auf das Entstehen von regulatorischen Zentren zwischen diesen Enhancer-reichen Regionen schließen. Weitere Untersuchungen zeigten Veränderung der sogenannten Super-Enhancer Cluster in unterschiedlichen Zelltypen. Des Weiteren sind Super-Enhancer TADs sehr dekondensiert und wurden häufig an Splicing-Speckle Regionen vorgefunden.
Chromatin needs to organize gene regulation whilst fitting into the confined space of the nucleus. Chromatin organization is therefore intertwined with gene activation and silencing. In recent years many advances in the field of chromatin architecture have been made showing that chromatin is organized hierarchically. Folding occurs in subsequent units, where each level of organization contributes to the spatial compaction of DNA and gene regulation. In this dissertation different levels of 3D chromatin organization were analysed using single-cell, high-resolution imaging. On the smallest scale, the 3D organization of two neighbouring Topologically Associating Domains (TADs) at the Sox9 locus was investigated. Performing Fluorescence in situ Hybridization (FISH) in 3D and cryosectioned mouse embryonic stem cells, extensive contacts between the two neighbouring TADs across the TAD boundary were detected. Applying FISH in a cell line bearing a genomic duplication within the Sox9 locus, the occurrence of two different conformations that result from the duplication was shown. Recent evidence from GAM showed the formation of long-range, multimer contacts between distal regulatory elements. Investigating the occurrence of long-range contacts between super-enhancer TADs in single cells by FISH, showed that they establish frequent interactions at close spatial distances. Furthermore the formation of clusters containing distal super-enhancer TADs could be demonstrated, indicating the possibility of higher-order regulatory hubs between these enhancer-rich regions. Further investigation showed that super-enhancer regions form different clusters in different cell types. Finally, it was shown that super-enhancers are highly decondensed and preferentially located at splicing speckles.

Abstract Radio communication is used in increasingly diversified device typologies. Telecommunications with a reduced detrimental impact on health and environment, and an improved cost-efficiency and working lifetime are expected by institutions, end-users, operators and manufacturers. Moreover, with more networks present or more articulated systems, dependability of the entirety is to be ensured. The related need of efficiency in various compartments – such as in the use energy or the radio spectrum – and of effectiveness in adapting to changing operating conditions can be achieved with cognitive features. This dissertation addresses network reconfiguration and dependability by cognitive measures from multiple perspectives – each covered by a respective part of this work – providing guidelines for cognitive networks design. A rationalising view on cognitive networks with related taxonomies and models includes a discussion on the dynamics and interactions in networks operating closely and simultaneously (here, concurrent networks). While cognitive domains are specified for cognitive functions, with a more generic scope control functions are assigned to topological domains. This allows a flexible exploitation of the system design by decoupling the specification of system functions from their mapping onto network devices that will host them. As interaction plays an important role in many topical scenarios, a model for networked engineered cognitive entities comprising four categories (observation, interworking, consolidation, operation) and two levels (a cognitive frontier and a metacognitive hub) is presented here. Its cognitive phases are considered with regard to the other architectural elements. Moving the focus down to the levers for exploitation of context awareness, are presented solutions for efficient use of resources and dependability in general, considering the network dynamics. For communication link and network adaptation, the effective capacity is captured by a compact-form expression also considering imperfections, while learning is exploited for reducing overhead, and collaboration for fairly maximising energy save
Tiivistelmä Käyttäjät, operaattorit ja laitevalmistajat toivovat tulevilta tietoliikennejärjestelmiltä sekä aiempaa pienempiä haitallisia vaikutuksia terveyteen ja ympäristöön että parannettua kustannustehokkuutta ja toiminta-aikaa. Lisäksi olisi varmistettava useiden verkkojen ja niiden muodostamien monimutkaisten järjestelmien kokonaisuuden luotettava toiminta. Tarvittava tehokkuus energian ja radioresurssien käytössä, samoin kuin kyky sopeutua muuttuviin käyttötilanteisiin, voidaan saavuttaa kognitiivisilla radioteknikoilla. Tämä väitöskirja käsittelee kognitiivisten menetelmien tuomaa radioverkkojen mukauttamista ja luotettavuutta eri näkökulmista. Samalla esitetään kognitiivisten verkkojen suunnittelun periaatteita ja lähtökohtia. Väitöskirja sisältää katsauksen kognitiivisiin radioverkkoihin niihin liittyvine luokitteluineen ja malleineen, sekä tarkastelee samanaikaisesti ja läheisesti toimivien verkkojen (rinnakkaisten verkkojen) dynamiikkaa ja vuorovaikutuksia. Työssä määritetään kognitiiviset lohkot kognitiivisine toimintoineen, kun taas topologiset tasot hallintatoimintoineen määritetään yleisemmin. Tämä mahdollistaa järjestelmäsuunnittelun joustavan hyödyntämisen erottamalla järjestelmän toimintojen määrittelyn toteuttavista verkkolaitteista. Koska vuorovaikutus on merkittävä tekijä useissa sovellusskenaarioissa, verkottuneille keinotekoisille kognitiivisille yksiköille ehdotetaan tässä neljä luokkaa (havainnointi, yhteistoiminta, vakauttaminen, toiminta) sekä kaksi vyöhykettä (kognitiivinen raja-alue ja metakognitiivinen keskus) sisältävää mallia. Mallin kognitiiviset vaiheet käsitellään suhteessa muihin arkkitehtuurin elementteihin. Järjestelmän kontekstitietoisuuden hyväksikäyttöön liittyen esitetään ratkaisuja resurssien tehokkaaseen käyttöön ja yleisemmin luotettavuuteen ottaen huomioon verkkojen dynamiikkaa. Yhteyksien ja verkkojen mukauttamisesta esitetään analyyttinen ratkaisu saavutettavan tehollisen kapasiteetin määrittämiseksi, huomioiden mahdolliset epäideaalisuudet. Kognitiivista oppimista hyödynnetään hallintalikenteen vähentämiseksi ja yhteistyötä energiansäästön maksimoimiseksi verkon alueella tasapuolisesti

Cette thèse présente: 1) le développement d'une nouvelle approche pour trouver des associations directes entre des paires d'éléments liés indirectement à travers diverses caractéristiques communes, 2) l'utilisation de cette approche pour associer directement des fonctions biologiques aux domaines protéiques (ECDomainMiner et GODomainMiner) et pour découvrir des interactions domaine-domaine, et enfin 3) l'extension de cette approche pour annoter de manière complète à partir des domaines les structures et les séquences des protéines. Au total, 20 728 et 20 318 associations EC-Pfam et GO-Pfam non redondantes ont été découvertes, avec des F-mesures de plus de 0,95 par rapport à un ensemble de référence Gold Standard extrait d'une source d'associations connues (InterPro). Par rapport à environ 1500 associations déterminées manuellement dans InterPro, ECDomainMiner et GODomainMiner produisent une augmentation de 13 fois le nombre d'associations EC-Pfam et GO-Pfam disponibles. Ces associations domaine-fonction sont ensuite utilisées pour annoter des milliers de structures de protéines et des millions de séquences de protéines pour lesquelles leur composition de domaine est connue mais qui manquent actuellement d'annotations fonctionnelles. En utilisant des associations de domaines ayant acquis des annotations fonctionnelles inférées, et en tenant compte des informations de taxonomie, des milliers de règles d'annotation ont été générées automatiquement. Ensuite, ces règles ont été utilisées pour annoter des séquences de protéines dans la base de données TrEMBL
This thesis presents: 1) the development of a novel approach to find direct associations between pairs of elements linked indirectly through various common features, 2) the use of this approach to directly associate biological functions to protein domains (ECDomainMiner and GODomainMiner), and to discover domain-domain interactions, and finally 3) the extension of this approach to comprehensively annotate protein structures and sequences. ECDomainMiner and GODomainMiner are two applications to discover new associations between EC Numbers and GO terms to protein domains, respectively. They find a total of 20,728 and 20,318 non-redundant EC-Pfam and GO-Pfam associations, respectively, with F-measures of more than 0.95 with respect to a “Gold Standard” test set extracted from InterPro. Compared to around 1500 manually curated associations in InterPro, ECDomainMiner and GODomainMiner infer a 13-fold increase in the number of available EC-Pfam and GO-Pfam associations. These function-domain associations are then used to annotate thousands of protein structures and millions of protein sequences for which their domain composition is known but that currently lack experimental functional annotations. Using inferred function-domain associations and considering taxonomy information, thousands of annotation rules have automatically been generated. Then, these rules have been utilized to annotate millions of protein sequences in the TrEMBL database

Yes
Mammalian genomes contain several dozens of large (>0.5 Mbp) lineage-specific gene loci harbouring functionally related genes. However, spatial chromatin folding, organization of the enhancer-promoter networks and their relevance to Topologically Associating Domains (TADs) in these loci remain poorly understood. TADs are principle units of the genome folding and represents the DNA regions within which DNA interacts more frequently and less frequently across the TAD boundary. Here, we used Chromatin Conformation Capture Carbon Copy (5C) technology to characterize spatial chromatin interaction network in the 3.1 Mb Epidermal Differentiation Complex (EDC) locus harbouring 61 functionally related genes that show lineage-specific activation during terminal keratinocyte differentiation in the epidermis. 5C data validated by 3D-FISH demonstrate that the EDC locus is organized into several TADs showing distinct lineage-specific chromatin interaction networks based on their transcription activity and the gene-rich or gene-poor status. Correlation of the 5C results with genome-wide studies for enhancer-specific histone modifications (H3K4me1 and H3K27ac) revealed that the majority of spatial chromatin interactions that involves the gene-rich TADs at the EDC locus in keratinocytes include both intra- and inter-TAD interaction networks, connecting gene promoters and enhancers. Compared to thymocytes in which the EDC locus is mostly transcriptionally inactive, these interactions were found to be keratinocyte-specific. In keratinocytes, the promoter-enhancer anchoring regions in the gene-rich transcriptionally active TADs are enriched for the binding of chromatin architectural proteins CTCF, Rad21 and chromatin remodeler Brg1. In contrast to gene-rich TADs, gene-poor TADs show preferential spatial contacts with each other, do not contain active enhancers and show decreased binding of CTCF, Rad21 and Brg1 in keratinocytes. Thus, spatial interactions between gene promoters and enhancers at the multi-TAD EDC locus in skin epithelial cells are cell type-specific and involve extensive contacts within TADs as well as between different gene-rich TADs, forming the framework for lineage-specific transcription.
This study was supported by the grants 5R01AR064580 and 1RO1AR071727 to VAB, TKS and AAS, as well as by the grants from MRC (MR/ M010015/1) and BBSRC (BB/K010050/1) to VAB.

Spider silk is a biomaterial of extraordinary toughness paired with elasticity. The assembly of silk proteins, so-called spidroins (from “spider” and “fibroin”), generates the silk threads we typically see in our garden or the corners of our houses. Although spider webs from different species vary considerably in geometry and size, many sections of spidroin sequences are conserved. Highly conserved regions, found in all spidroins, relate to the terminal domains of the protein, i.e., the N-terminal (NTD) and C-terminal domains (CTD). Both have an essential function in the silk fibre association and polymerisation. The NTD is a 14 kDa five-helix bundle, which self-associates via a pH-driven mechanism. This process is critical for starting the polymerisation of the fibre. However, detailed insights into how conserved this mechanism is in different species and the quantitative thermodynamic comparison between homologous NTDs was missing. For this reason, four homologous NTDs of the major ampullate gland (MaSp) from spider species Euprosthenops australis, Nephila clavipes, Latrodectus hesperus, and Latrodectus geometricus were investigated. I analysed and quantified equilibrium thermodynamics, kinetics of folding, and self-association. Methods involved dynamic light scattering (MALS), stopped-flow fluorescence and circular dichroism spectroscopy in combination with thermal and chemical denaturation experiments. The results showed conserved, cooperative two-state folding on a sub-millisecond time scale. All homologous NTDs showed a similarly fast association in the order of 10^9 M^−1 s^−1, while the resulting equilibrium dissociation constants were in the low nanomolar range. Electrostatic forces were found to be of great importance for protein association. Monomeric protein stability increased with salt concentration while enhancing its folding speed. However, due to Debye-Hückel effects, we found intermolecular electrostatics to be shielded, which reduced the NTDs association capacity significantly at high ionic strength. Altogether, the energetics and kinetics of the NTD dimerisation was conserved for all analysed homologs. Comparable to the NTD, the spider silks CTD is also a α-helix bundle, which covalently links two spidroins. The orientation of the domains predetermines the future fibre geometry. Here again, the detailed quantitative characterisation of the folding and dimerisation was missing. Therefore, the CTD from the E. australis was analysed in-depth. The protein folded via a three-state mechanism and was placed in the family of knotted proteins. By analysing the amino acid composition of the NTD of the MaSp1 of the Euprosthenops australis, we found an unusually high content of methionine residues (Met). To elucidate why this protein exhibits so many Met residues, I mutated all core Mets simultaneously to leucine (Leu). Results revealed a dramatically stabilised NTD, which now folded 50 times faster. After solving the tertiary structure of the mutant by NMR (nuclear magnetic resonance) spectroscopy, the structure of the monomeric mutant was found to be identical with the wild-type protein. However, when probing the dimerisation of the NTD, I could show that the association capacity was substantially impaired for the mutant. Our findings lead to the conclusion that Met provides the NTD with enhanced conformational dynamics and thus mobilises the protein, which results in tightly associated dimers. In additional experiments, I first re-introduced new Met residues into the Met-depleted protein at sequence positions containing native Leu. Hence, the mutated NTD protein was provided with the same number of Leu, which were previously removed by mutation. However, the protein did not regain wild-type characteristics. The functionality was not restored, but its stability was decreased as expected. To probe our hypothesis gained from the MaSp NTD, I transferred the experiment to another protein, namely the Hsp90 chaperone. Therefore, I incorporated methionine residues in the protein, which resulted in a slight improvement of its function. Finally, trial experiments were performed aiming at the synthesis of shortened spidroin constructs containing less repetitive middle-segments than the wild-type protein. The objective was to study the findings of the terminal domains in the context of an intact spidroin. The synthesis of these engineered spidroins was challenging. Nevertheless, preliminary results encourage the assumption that the characteristics observed in the isolated domains hold true in the context of a full-length spidroin
Spinnenseide ist ein Biomaterial mit außergewöhnlicher Widerstandsfähigkeit welche gepaart ist mit Elastizität. Das Zusammenfügen von Seidenproteinen aus so ge-nannten Spidroinen (ein Kunstwort aus „Spinne“ und „Fibroin“) erzeugt die Seiden-fäden, die wir typischerweise in unseren Gärten oder in den Ecken unserer Häuser finden. Obwohl Spinnennetze von verschiedenen Spinnenarten in Geometrie und Größe stark variieren, sind große Teile der Spidroin-Sequenzen konserviert. Stark konservierte Bereiche, die in allen Spidroinen vorkommen, sind die endständigen Bereiche des Proteins, die N-terminale (NTD) und C-terminale Domäne (CTD) ge-nannt werden. Beide haben wichtige Funktionen in der Assoziation der Proteine im Spinnkanal und deren Polymerisation zur Ausbildung des Seidenfadens. Die NTD ist ein kleines 14 kDa Protein, bestehend aus einem Bündel aus fünf Helices, dessen Selbstorganisation pH-abhängig ist. Dieser Prozess leitet die Poly-merisation der Faser ein. Allerdings fehlten bis heute Informationen darüber, ob dieser Mechanismus bei homologen Domänen aus verschiedenen Spinnenarten konser¬viert ist, da kaum quantitative biophysikalische Daten vorhanden sind. Aus diesem Grund wurden vier homologe NTDs der Spinnenarten Euprosthenops australis, Nephila clavipes, Latrodectus hesperus und Latrodectus geometricus vergleichend untersucht und deren Gleichgewichts-Thermodynamik, die Kinetik der Faltung und die Selbstassoziation quantitativ analysiert. Dazu wurden Methoden wie dynamische Mehrwinkel-Lichtstreuung (MALS), Stopped-Flow Fluoreszenz-spektroskopie und Zirkulardichroismus in Kombination mit thermischen und chemischen Denaturierungs¬experimenten angewandt. Die Ergebnisse lieferten die Erkenntnis einer kooperativen Zwei-Zustands-Faltung, die auf einer Zeitskala von weniger als einer Millisekunde stattfand. Alle homologen NTDs zeigten eine schnelle Assoziationsratenkonstante in der Größenordnung von 10^9 M^-1 s^-1, während die Gleichgewichts-Dissoziationskonstante für alle Homologe im nied¬rigen nano-molaren Bereich lag. Die Proteinassoziation wurde durch elektrostatische Kräfte gesteuert, wobei hohe Salzkonzentrationen die Stabilität des monomeren Proteins und dessen Faltungsgeschwindigkeit erhöhten. Die Assoziation zweier Domänen wurde jedoch durch Abschirmung intermolekularer elektrostatischer Kräfte, dem Debye-Hückel-Gesetz zufolge, reduziert. Die Energetik und Kinetik der NTD-Dimerisierung aller untersuchten Homologen erwies sich konserviert. Ebenso wie die NTD, ist auch die CTD der Spinnenseide ein α-helikales Bündel, welche jedoch zwei Spidroine kovalent miteinander verbindet. Die Orientierung der verknüpften Domäne bestimmt bereits die zukünftige Faserstruktur. Ähnlich wie bei der NTD, waren Faltung und Dimerisierung der CTD bisher nicht im Detail be-schrieben. Durch eine detaillierte Analyse der CTD der E. australis konnte gezeigt werden, dass das Protein sich in einem dreistufigen Mechanismus faltet und außerdem der Familie der geknoteten Proteine angehört. Bei genauerer Betrachtung der Aminosäurezusammensetzung der E. australis NTD konnte ein ungewöhnlich hoher Anteil der Aminosäure Methionin (Met) festge¬stellt werden. Um diesen überraschenden Sachverhalt zu verstehen, habe ich alle im Kern liegenden Met zu Leucin (Leu) mutiert. Die Ergebnisse zeigten eine extrem stabilisierte NTD, welche sich nun 50-fach schneller faltete. Die Protein¬struktur der Mutante wurde in Lösung mittels NMR Spektroskopie ermittelt. Das Ergebnis lieferte deckungsgleiche Strukturen von Mutante und Wildtyp im monomeren Zustand. Allerdings zeigten NTD Dimerisierungs-Versuche, dass die Assoziations-fähigkeit der Mutante erheblich beeinträchtigt war. Untersuchungen der nativen Dynamik mittels NMR und Fluoreszenzkorrelationsspektroskopie zeigten, dass Met diese entscheidend verstärkt, was zu einem eng assoziierten Dimer führte. Im Versuch die Dynamik wieder künstlich herzustellen, habe ich neue Met in die Mutante eingeführt, auf Sequenzpositionen welche natürlicherweise Leu aufweisen. Somit wurde die ursprüngliche Anzahl an Met in der NTD wiederher¬gestellt, jedoch an anderen Positionen. Obwohl das Protein wie erwartet an Stabilität verlor, konnte dessen Funktionalität nicht wiederhergestellt werden. Um unsere Erkenntnisse auf andere Proteine zu übertragen, wurden Met Reste künstlich in ein Hsp90 Protein eingeführt. Es konnte eine leicht verbesserte Funktionalität des Proteins beobachtet werden. Schließlich wurde versucht, die für die CTD und NTD gewonnen Erkenntnisse auf intakte, jedoch verkürzte Spidroine zu übertragen. Dazu wurden Spidroine mit weniger repetitiven Mittelsegmenten mittels rekombinanten Methoden hergestellt. Die Synthese dieser Spidroine erwies sich als Herausforderung. Allerdings zeigten die vorläufigen Ergebnisse, dass eine Verallgemeinerung der Erkenntnisse der isolierten Domänen auf das Volllängen-Spidroin möglich ist

Pattern recognition and machine learning problems are often conceived to work on metric vector spaces, where patterns are described by multi-dimensional feature vectors. However, many real-world complex systems are more conveniently modelled by complex data structures such as graphs, which are able to capture topological and semantic information amongst entities. This Thesis helps in bridging the gap between pattern recognition and graphs, with major emphasis on the hypergraphs domain. Six different strategies for solving graph-based pattern recognition problems are proposed, spanning several paradigms including kernel methods, embedding spaces and feature generation. The first two techniques map a graph towards a vector space by means of the spectral density of the Laplacian matrix and by means of topological invariants called the Betti numbers, respectively. Two additional techniques, according to the Granular Computing paradigm, map a graph towards a vector space by means of symbolic histograms. In a first case, simplices extracted from the simplicial complexes evaluated over the underlying graph are considered as candidate pivotal substructures for synthesising the symbolic histograms; in a second case, each path along a graph can be assigned a score that consider its specificity and sensitivity with respect to one of the problem-related classes and its inclusion in the candidate pivotal substructures is strictly related to its score. The final two techniques fall under the kernel methods umbrella: the first defines novel hypergraph kernels on the top of the simplicial complexes, the latter embraces a multiple kernel paradigm to exploit multiple graph representations simultaneously. These techniques are tested on real-world problems related to three biological case studies, namely the solubility prediction and enzymatic properties discrimination in protein networks and the analysis of metabolic networks. Further, the most cutting-edge techniques are also tested on well-known benchmark datasets for graph classification and compared against current approaches in graph-based pattern recognition.

The Implicit Association Test (IAT) is a computer-based psychological test that measures implicit attitudes, stereotypes and beliefs. In an effort to better understand the applicability and limitations of the IAT researchers have investigated the effects of manipulating a variety of procedural variables that comprise the IAT, not least the IAT categories and the exemplars that are instances of those categories. This study investigated the effects of manipulating the IAT's target categories that define the attitudinal domain that the IAT measures. Experiments were devised to determine the IAT's sensitivity to minor and major semantic manipulations to its target categories while keeping exemplars and attribute categories constant. It was found that the IAT was sensitive to major semantic differences in its target categories, but was apparently insensitive to minor semantic category differences, implying that it is unable to discriminate between subtle distinctions in attitude. It was hypothesised that this latter finding could have been partly due to a temporary cognitive re-definition of the categories in accordance with the salient characteristics of the exemplars.
Thesis (M.Soc.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2008.

Understanding the relationships between chromatin structure and gene regulation is a fundamental problem in genetics. However, for a long time there has been little progress in the field of genome architecture except for studying low scale chromatin organization. This situation changed during last two decades due to the advancement in the development of NGS technology, which gave rise to Chromosome Conformation Capture (3C) methods. The availability of 3C techniques enabled genome organization studies on an unprecedented scale. In particular, the 3C-derived Hi-C protocol allowed researchers to interrogate millions of chromatin interactions between pairs of regions genome-wide at a very high resolution. One of the main applications of Hi-C is the differential analysis, which aim to identify the structural differences of chromatin influencing regulatory processes across various cell types, treatments or species. In this thesis we focus on the issue of comparing Hi-C contact matrices. First, we study the problem of assessing the similarity between chromosome segmentations arising from identification of Topologically Associating Domains (TAD) - an inherent feature of mammalian Hi-C maps, which were shown to shape regulatory landscape of the genome. We present a novel distance measure called BP-score, tailored for comparison of TAD partitionings and prove that our measure satisfy metric properties. Evaluation of the BP-score on real and simulated datasets demonstrates that it performs competitive against existing approaches. Additionally, we introduce local measures of domain rearrangement and show their correlation with functional measurements. Second, we develop a normalization-free method for discovery of Hi-C differential interactions called DiADeM. Our method introduces an intuitive definition of differential interaction, which takes into account the cross-dataset contact profile similarity. Finally, we assess DiADeMs ability to detect differential interactions using simulated contact maps and show it performs well against other available methods for Hi-C differential analysis. In summary, the tools developed by us may help researches in discovering unknown structural alterations driving regulatory mechanisms.
Problem zrozumienia relacji pomiędzy strukturą chromatyny, a regulacją genów ma kluczowe znaczenie w genetyce. Niestety przez wiele lat możliwe były wyłącznie badania architektury genomu w niskiej rozdzielczości lub małej skali. Sytuacja zmieniła się w ciągu ostatnich dwóch dekad głównie ze względu na postęp w rozwoju technologii NGS, która dała początek metodom 3C (ang. Chromosome Conformation Capture). Dostępność technik 3C umożliwiła badania organizacji genomu na niespotykaną dotąd skalę. W szczególności wysoko-rozdzielczy wariant metody 3C - protokół Hi-C pozwala uzyskać dane dotyczące milionów interakcji pomiędzy parami regionów chromatyny w całym genomie. Jednym z głównych zastosowań protokołu Hi-C jest analiza różnicowa, która ma na celu zidentyfikowanie różnic w strukturze chromatyny wpływających na procesy regulacji genów w różnych typach komórek, warunkach eksperymentalnych lub gatunkach. W tej pracy koncentrujemy się na problemie porównywania macierzy kontaktów Hi-C. Po pierwsze, badamy problem oceny podobieństwa między segmentacjami chromosomów wynikającymi z identyfikacji domen topologicznych (TAD) - nieodłącznej cechy map Hi-C organizmów ssaków, które, jak wykazano, kształtują krajobraz regulacyjny genomu. Prezentujemy nową miarę odległości o nazwie BP-score, dostosowaną do porównania segmentacji TAD oraz dowodzimy, że nasza miara jest metryką. Przykładowe analizy porównawcze przeprowadzone na danych symulowanych i rzeczywistych pokazują, że odległość BP jest konkurencyjna w stosunku do innych metryk wykorzystywanych dotychczas podczas badania podobieństwa segmentacji. Dodatkowo wprowadzamy lokalne miary rearanżacji domen topologicznych i pokazujemy, że pomiary rearanżacji uzyskane przy użyciu wprowadzonych przez nas miar korelują z pomiarami ekspresji genów lub metylacji. Po drugie, opracowujemy metodę do wykrywania różnicowych oddziaływań HiC o nazwie DiADeM działającą na danych nieznormalizowanych. Nasza metoda wprowadza intuicyjną definicję interakcji różnicowych, która uwzględnia podobieństwo profili kontaktów pomiędzy zestawami danych. Na koniec oceniamy zdolność naszej metody do wykrywania interakcji różnicowych przy użyciu symulowanych map kontaktów i pokazujemy, że osiąga konkurencyjne wyniki w porównaniu z innymi dostępnymi metodami służącymi do analizy różnicowej Hi-C. Podsumowując, opracowane przez nas narzędzia mogą pomóc badaczom w odkrywaniu nieznanych zmian strukturalnych wpływających na mechanizmy regulacji genów.