<p>Recent literature has investigated the use of different clustering techniques for analysis of gene expression data. For example, self-organizing maps (SOMs) have been used to identify gene clusters of clear biological relevance in human hematopoietic differentiation and the yeast cell cycle (Tamayo et al., 1999). Hierarchical clustering has also been proposed for identifying clusters of genes that share common roles in cellular processes (Eisen et al., 1998; Michaels et al., 1998; Wen et al., 1998). Systematic evaluation of clustering results is as important as generating the clusters. However, this is a difficult task, which is often overlooked in gene expression studies. Several gene expression studies claim success of the clustering algorithm without showing a validation of complete clusterings, for example Ben-Dor and Yakhini (1999) and Törönen et al. (1999).</p><p>In this dissertation we propose an evaluation approach based on a relative entropy measure that uses additional knowledge about genes (gene annotations) besides the gene expression data. More specifically, we use gene annotations in the form of an enzyme classification hierarchy, to evaluate clusterings. This classification is based on the main chemical reactions that are catalysed by enzymes. Furthermore, we evaluate clusterings with pure statistical measures of cluster validity (compactness and isolation).</p><p>The experiments include applying two types of clustering methods (SOMs and hierarchical clustering) on a data set for which good annotation is available, so that the results can be partly validated from the viewpoint of biological relevance.</p><p>The evaluation of the clusters indicates that clusters obtained from hierarchical average linkage clustering have much higher relative entropy values and lower compactness and isolation compared to SOM clusters. Clusters with high relative entropy often contain enzymes that are involved in the same enzymatic activity. On the other hand, the compactness and isolation measures do not seem to be reliable for evaluation of clustering results.</p>

Cette thèse a pour sujet le développement de méthodes et de techniques permettant d'inférer l'état affectif d'une personne à partir d'informations visuelles. Plus précisement, nous nous intéressons à l'analyse d'expressions du visage, puisque le visage est la partie la mieux visible du corps, et que l'expression du visage est la manifestation la plus évidente de l'affect. Nous étudions différentes théories psychologiques concernant affect et émotions, et différentes facons de représenter et de classifier les émotions d'une part et la relation entre expression du visage et émotion sousjacente d'autre part. Nous présentons les dérivées Gaussiennes multi-échelle en tant que descripteur dímages pour l'estimation de la pose de la tête, pour la détection de sourire, puis aussi pour la mesure de l'affect. Nous utilisons l'analyse en composantes principales pour la réduction de la dimensionalité, et les machines à support de vecteur pour la classification et la regression. Nous appliquons cette même architecture, simple et efficace, aux différents problèmes que sont l'estimation de la pose de tête, la détection de sourire, et la mesure d'affect. Nous montrons que non seulement les dérivées Gaussiennes multi-échelle ont une performance supérieure aux populaires filtres de Gabor, mais qu'elles sont également moins coûteuses en calculs. Lors de nos expérimentations nous avons constaté que dans le cas d'un éclairage partiel du visage les dérivées Gaussiennes multi-échelle ne fournissent pas une description d'image suffisamment discriminante. Pour résoudre ce problème nous combinons des dérivées Gaussiennes avec des histogrammes locaux de type LBP (Local Binary Pattern). Avec cette combinaison nous obtenons des résultats à la hauteur de l'état de l'art pour la détection de sourire dans le base d'images GENKI qui comporte des images de personnes trouvées «dans la nature» sur internet, et avec la difficile «extended YaleB database». Pour la classification dans la reconnaissance de visage nous utilisons un apprentissage métrique avec comme mesure de similarité une distance de Minkowski. Nous obtenons le résultat que les normes L1 and L2 ne fournissent pas toujours la distance optimale; cet optimum est souvent obtenu avec une norme Lp où p n'est pas entier. Finalement, nous développons un système multi-modal pour la détection de dépressions nerveuses, avec en entrée des informations audio et vidéo. Pour la détection de mouvements intra-faciaux dans les données vidéo nous utilisons de descripteurs de type LBP-TOP (Local Binary Patterns -Three Orthogonal Planes), alors que nous utilisons des trajectoires denses pour les mouvements plus globaux, par exemple de la tête ou des épaules. Nous avons trouvé que les descripteurs LBP-TOP encodés avec des vecteurs de Fisher suffisent pour dépasser la performance de la méthode de reférence dans la compétition «Audio Visual Emotion Challenge (AVEC) 2014». Nous disposons donc d'une technique effective pour l'evaluation de l'état dépressif, technique qui peut aisement être étendue à d'autres formes d'émotions qui varient lentement, comme l'humeur (mood an Anglais). / In this thesis we focus on the development of methods and techniques to infer affect from visual information. We focus on facial expression analysis since the face is one of the least occluded parts of the body and facial expressions are one of the most visible manifestations of affect. We explore the different psychological theories on affect and emotion, different ways to represent and classify emotions and the relationship between facial expressions and underlying emotions. We present the use of multiscale Gaussian derivatives as an image descriptor for head pose estimation, smile detection before using it for affect sensing. Principal Component Analysis is used for dimensionality reduction while Support Vector Machines are used for classification and regression. We are able to employ the same, simple and effective architecture for head pose estimation, smile detection and affect sensing. We also demonstrate that not only do multiscale Gaussian derivatives perform better than the popular Gabor Filters but are also computationally less expensive to compute. While performing these experiments we discovered that multiscale Gaussian derivatives do not provide an appropriately discriminative image description when the face is only partly illuminated. We overcome this problem by combining Gaussian derivatives with Local Binary Pattern (LBP) histograms. This combination helps us achieve state-of-the-art results for smile detection on the benchmark GENKI database which contains images of people in the "wild" collected from the internet. We use the same description method for face recognition on the CMU-PIE database and the challenging extended YaleB database and our results compare well with the state-of-the-art. In the case of face recognition we use metric learning for classification, adopting the Minkowski distance as the similarity measure. We find that L1 and L2 norms are not always the optimum distance metrics and the optimum is often an Lp norm where p is not an integer. Lastly we develop a multi-modal system for depression estimation with audio and video information as input. We use Local Binary Patterns -Three Orthogonal Planes (LBP-TOP) features to capture intra-facial movements in the videos and dense trajectories for macro movements such as the movement of the head and shoulders. These video features along with Low Level Descriptor (LLD) audio features are encoded using Fisher Vectors and finally a Support Vector Machine is used for regression. We discover that the LBP-TOP features encoded with Fisher Vectors alone are enough to outperform the baseline method on the Audio Visual Emotion Challenge (AVEC) 2014 database. We thereby present an effective technique for depression estimation which can be easily extended for other slowly varying aspects of emotions such as mood.

The blood-brain barrier, formed by brain microvessel endothelial cells, is the restrictive barrier between the brain parenchyma and the circulating blood. It was previously demonstrated in our laboratory that knock down of fatty acid transport proteins FATP-1 and CD36 attenuated apical to basolateral monounsaturated fatty acid transport across human brain microvessel endothelial cells (HBMEC). Arachidonic acid (AA; 5,8,11,14 - cis-eicosatetraenoic acid) is a conditionally essential, polyunsaturated fatty acid [20:4(n-6)] and a major constituent of brain lipids. We examined transport of AA across confluent monolayers of HBMEC. Control cells or HBMEC with knock down of FATP-1 or CD36 were cultured on Transwell® plates and incubated apically with [3H]AA and incorporation of [3H]AA into the basolateral medium was determined temporally. [3H]AA was rapidly incorporated into the basolateral medium with time in control cells. Surprisingly, knock down of FATP-1 or CD36 did not alter [3H]AA movement into the basolateral medium. The increased permeability mediated by AA was likely caused by a metabolite of AA produced de novo and was confirmed by an increased movement of fluorescent dextran from apical to basolateral medium. HBMECs expressed PGE2 synthase, cyclooxygenase-1 and -2, PGE2 receptors, tight junction proteins and prostaglandin transporters. The AA-mediated increase in membrane permeability was not attenuated by cyclooxygenase inhibitor drugs (NSAIDs). Incubation of the HBMEC monolayers with exogenous PGE2 resulted in attenuation of the AA-mediated permeability increases. The results indicate that AA increases the permeability of the HBMEC monolayer likely via increased production of metabolites or by-products of the lipoxygenase or epoxygenase pathways. These observations may explain the rapid influx of AA into the brain previously observed upon plasma infusion with AA.

The blood-brain barrier, formed by brain microvessel endothelial cells, is the restrictive barrier between the brain parenchyma and the circulating blood. It was previously demonstrated in our laboratory that knock down of fatty acid transport proteins FATP-1 and CD36 attenuated apical to basolateral monounsaturated fatty acid transport across human brain microvessel endothelial cells (HBMEC). Arachidonic acid (AA; 5,8,11,14 - cis-eicosatetraenoic acid) is a conditionally essential, polyunsaturated fatty acid [20:4(n-6)] and a major constituent of brain lipids. We examined transport of AA across confluent monolayers of HBMEC. Control cells or HBMEC with knock down of FATP-1 or CD36 were cultured on Transwell® plates and incubated apically with [3H]AA and incorporation of [3H]AA into the basolateral medium was determined temporally. [3H]AA was rapidly incorporated into the basolateral medium with time in control cells. Surprisingly, knock down of FATP-1 or CD36 did not alter [3H]AA movement into the basolateral medium. The increased permeability mediated by AA was likely caused by a metabolite of AA produced de novo and was confirmed by an increased movement of fluorescent dextran from apical to basolateral medium. HBMECs expressed PGE2 synthase, cyclooxygenase-1 and -2, PGE2 receptors, tight junction proteins and prostaglandin transporters. The AA-mediated increase in membrane permeability was not attenuated by cyclooxygenase inhibitor drugs (NSAIDs). Incubation of the HBMEC monolayers with exogenous PGE2 resulted in attenuation of the AA-mediated permeability increases. The results indicate that AA increases the permeability of the HBMEC monolayer likely via increased production of metabolites or by-products of the lipoxygenase or epoxygenase pathways. These observations may explain the rapid influx of AA into the brain previously observed upon plasma infusion with AA.

Cette thèse a pour sujet le développement de méthodes et de techniques permettant d'inférer l'état affectif d'une personne à partir d'informations visuelles. Plus précisement, nous nous intéressons à l'analyse d'expressions du visage, puisque le visage est la partie la mieux visible du corps, et que l'expression du visage est la manifestation la plus évidente de l'affect. Nous étudions différentes théories psychologiques concernant affect et émotions, et différentes facons de représenter et de classifier les émotions d'une part et la relation entre expression du visage et émotion sousjacente d'autre part. Nous présentons les dérivées Gaussiennes multi-échelle en tant que descripteur dímages pour l'estimation de la pose de la tête, pour la détection de sourire, puis aussi pour la mesure de l'affect. Nous utilisons l'analyse en composantes principales pour la réduction de la dimensionalité, et les machines à support de vecteur pour la classification et la regression. Nous appliquons cette même architecture, simple et efficace, aux différents problèmes que sont l'estimation de la pose de tête, la détection de sourire, et la mesure d'affect. Nous montrons que non seulement les dérivées Gaussiennes multi-échelle ont une performance supérieure aux populaires filtres de Gabor, mais qu'elles sont également moins coûteuses en calculs. Lors de nos expérimentations nous avons constaté que dans le cas d'un éclairage partiel du visage les dérivées Gaussiennes multi-échelle ne fournissent pas une description d'image suffisamment discriminante. Pour résoudre ce problème nous combinons des dérivées Gaussiennes avec des histogrammes locaux de type LBP (Local Binary Pattern). Avec cette combinaison nous obtenons des résultats à la hauteur de l'état de l'art pour la détection de sourire dans le base d'images GENKI qui comporte des images de personnes trouvées «dans la nature» sur internet, et avec la difficile «extended YaleB database». Pour la classification dans la reconnaissance de visage nous utilisons un apprentissage métrique avec comme mesure de similarité une distance de Minkowski. Nous obtenons le résultat que les normes L1 and L2 ne fournissent pas toujours la distance optimale; cet optimum est souvent obtenu avec une norme Lp où p n'est pas entier. Finalement, nous développons un système multi-modal pour la détection de dépressions nerveuses, avec en entrée des informations audio et vidéo. Pour la détection de mouvements intra-faciaux dans les données vidéo nous utilisons de descripteurs de type LBP-TOP (Local Binary Patterns -Three Orthogonal Planes), alors que nous utilisons des trajectoires denses pour les mouvements plus globaux, par exemple de la tête ou des épaules. Nous avons trouvé que les descripteurs LBP-TOP encodés avec des vecteurs de Fisher suffisent pour dépasser la performance de la méthode de reférence dans la compétition «Audio Visual Emotion Challenge (AVEC) 2014». Nous disposons donc d'une technique effective pour l'evaluation de l'état dépressif, technique qui peut aisement être étendue à d'autres formes d'émotions qui varient lentement, comme l'humeur (mood an Anglais). / In this thesis we focus on the development of methods and techniques to infer affect from visual information. We focus on facial expression analysis since the face is one of the least occluded parts of the body and facial expressions are one of the most visible manifestations of affect. We explore the different psychological theories on affect and emotion, different ways to represent and classify emotions and the relationship between facial expressions and underlying emotions. We present the use of multiscale Gaussian derivatives as an image descriptor for head pose estimation, smile detection before using it for affect sensing. Principal Component Analysis is used for dimensionality reduction while Support Vector Machines are used for classification and regression. We are able to employ the same, simple and effective architecture for head pose estimation, smile detection and affect sensing. We also demonstrate that not only do multiscale Gaussian derivatives perform better than the popular Gabor Filters but are also computationally less expensive to compute. While performing these experiments we discovered that multiscale Gaussian derivatives do not provide an appropriately discriminative image description when the face is only partly illuminated. We overcome this problem by combining Gaussian derivatives with Local Binary Pattern (LBP) histograms. This combination helps us achieve state-of-the-art results for smile detection on the benchmark GENKI database which contains images of people in the "wild" collected from the internet. We use the same description method for face recognition on the CMU-PIE database and the challenging extended YaleB database and our results compare well with the state-of-the-art. In the case of face recognition we use metric learning for classification, adopting the Minkowski distance as the similarity measure. We find that L1 and L2 norms are not always the optimum distance metrics and the optimum is often an Lp norm where p is not an integer. Lastly we develop a multi-modal system for depression estimation with audio and video information as input. We use Local Binary Patterns -Three Orthogonal Planes (LBP-TOP) features to capture intra-facial movements in the videos and dense trajectories for macro movements such as the movement of the head and shoulders. These video features along with Low Level Descriptor (LLD) audio features are encoded using Fisher Vectors and finally a Support Vector Machine is used for regression. We discover that the LBP-TOP features encoded with Fisher Vectors alone are enough to outperform the baseline method on the Audio Visual Emotion Challenge (AVEC) 2014 database. We thereby present an effective technique for depression estimation which can be easily extended for other slowly varying aspects of emotions such as mood.

Recent literature has investigated the use of different clustering techniques for analysis of gene expression data. For example, self-organizing maps (SOMs) have been used to identify gene clusters of clear biological relevance in human hematopoietic differentiation and the yeast cell cycle (Tamayo et al., 1999). Hierarchical clustering has also been proposed for identifying clusters of genes that share common roles in cellular processes (Eisen et al., 1998; Michaels et al., 1998; Wen et al., 1998). Systematic evaluation of clustering results is as important as generating the clusters. However, this is a difficult task, which is often overlooked in gene expression studies. Several gene expression studies claim success of the clustering algorithm without showing a validation of complete clusterings, for example Ben-Dor and Yakhini (1999) and Törönen et al. (1999). In this dissertation we propose an evaluation approach based on a relative entropy measure that uses additional knowledge about genes (gene annotations) besides the gene expression data. More specifically, we use gene annotations in the form of an enzyme classification hierarchy, to evaluate clusterings. This classification is based on the main chemical reactions that are catalysed by enzymes. Furthermore, we evaluate clusterings with pure statistical measures of cluster validity (compactness and isolation). The experiments include applying two types of clustering methods (SOMs and hierarchical clustering) on a data set for which good annotation is available, so that the results can be partly validated from the viewpoint of biological relevance. The evaluation of the clusters indicates that clusters obtained from hierarchical average linkage clustering have much higher relative entropy values and lower compactness and isolation compared to SOM clusters. Clusters with high relative entropy often contain enzymes that are involved in the same enzymatic activity. On the other hand, the compactness and isolation measures do not seem to be reliable for evaluation of clustering results.

This study represents the first comprehensive analysis of gene regulation underlying the distinct categories of resistance afforded to wheat (Triticum aestivum, L.) by different Dn genes. Russian wheat aphid (Diuraphis noxia, Mordv.) feeding on susceptible wheat cultivars causes leaf rolling, chlorosis and the eventual death of the plant. Plants expressing Dn genes are resistant to D. noxia infestation, but different Dn genes afford phenotypically distinct modes of resistance: the Dn1 gene confers an antibiotic effect to lower aphid fecundity; Dn2 confers tolerance to high aphid pressure; and Dn5 confers antixenosis, and aphids do not prefer such plants as hosts. Little is known about the components involved in establishing a successful defence response against D. noxia attack and how these differ between the distinct resistance categories. It is assumed that the Dn genes function as classic R genes in plant defence, being receptors for elicitors in aphid saliva. Upon recognition, defence response signalling is initiated, but the exact mechanics of subsequent cellular events in aphid resistance have only recently come under investigation. Evidence from cDNA microarray and subtractive hybridization experiments indicated the involvement of kinase signalling cascades and photosynthetic proteins in the response against D. noxia. However, expression analysis describing how these processes differ between plants carrying different Dn genes and how these differences account for antibiosis, antixenosis or tolerance had not been conducted. We consequently investigated the downstream components involved in or affected by the generation of these resistance mechanisms by comparing the responses in transcript regulation of Tugela near-isogenic lines with different Dn genes to D. noxia infestation. cDNA-AFLP analysis was selected as an appropriate functional genomics tool, since it is semi-quantitative, does not require prior sequence information and allows for the discovery of novel genes. cDNA-AFLP analysis yielded 121 differentially regulated transcript-derived fragments (TDFs) grouped into eight expression clusters. We cloned and sequenced 49 representative TDFs, which were further classified into five broad functional categories based on inferred similarity to database sequences. Transcripts involved in such diverse processes as stress, signal transduction, photosynthesis, metabolism and gene regulation were found to be differentially regulated during D. noxia feeding. Many TDFs demonstrated homology to proteins with unknown function and several novel transcripts with no similarity to previously published sequences were also discovered. Detailed expression analysis using quantitative RT-PCR and RNA hybridization provided evidence that the time and intensity of induction of specific pathways is critical for the development of a particular mode of resistance. This includes: the generation of kinase signalling cascades and the induction of several ancillary processes such as ubiquitination, leading to a sustained oxidative burst and the hypersensitive response during antibiosis; tolerance as a passive resistance mechanism countering aphid-induced symptoms through the repair or de novo synthesis of photosystem proteins; and the possible involvement of ethylene-mediated wounding pathways in generating volatile organic compounds during antixenosis. This is the first report on the involvement of KCO1, a vacuolar K+ channel, in assisting cytosolic Ca2+-influx and preventing leaf rolling, as well as on the role of iron homeostasis as a gene regulatory mechanism for sustaining the oxidative burst during the antibiotic defence response. This study opens up several areas of investigation heretofore unexplored in cereal-aphid interaction research. Of particular interest is the induction of genes involved in photosynthetic compensation during Dn2 tolerance responses, since these constitute a novel, passive resistance mechanism exclusive to aphid defence as opposed to the active resistance triggered in the presence of the Dn1 gene in the form of a general hypersensitive response. / Dissertation (MSc)--University of Pretoria, 2008. / Genetics / unrestricted

Impact of NKR-P1 polymorphism on Ly49 receptors expression in hybrid mouse strains (C57BL/6 x Balb/c, F10-12) Abstract Natural killer (NK) cells constitute the subpopulation of large granular lymphocytes which mediate spontaneous immune response against infected, transformed or allogeneic cells and thus represent an important component of the innate immunity. NK cells express a wide repertoir of surface receptors which can be either activating or inhibitory and which mediate NK cell recognition and regulation of cytolytic activity. NKR-P1 and Ly49 receptor families belong to the most important murine NK receptors. Both NKR-P1 and Ly49 families are members of C-type lectin-like superfamily of receptors encoded by natural killer gene complex (NKC) on chromosome 6 and include both activating and inhibitory members. The aim of this diploma thesis was to elucidate the impact of Nkr-p1c gene divergence on Ly49 receptors expression and to find out whether the Ly49 and Nkr-p1 gene clusters (which are localized on opposite ends of NKC) are inherited independently or whether the NKC domain is inherited as a complex. The second research interest was to illustrate the influence of the above mentioned divergence on cytotoxic activity of NK cells and tumor growth. In this study, inbred mouse strains C57BL/6 and Balb/c...

Philosophiae Doctor - PhD / Expressed sequence tag database is a rich and fast growing source of data for gene expression analysis and drug discovery. Clustering of raw EST data is a necessary step for further analysis and one of the most challenging problems of modem computational biology. There are a few systems, designed for this purpose and a few more are currently under development. These systems are reviewed in the "Literature and software review". Different strategies of supervised and unsupervised clustering are discussed, as well as sequence comparison techniques, such as based on alignment or oligonucleotide compositions. Analysis of potential bottlenecks and estimation of computation complexity of EST clustering is done in Chapter 2. This chapter also states the goals for the research and justifies the need for new algorithm that has to be fast, but still sensitive to relatively short (40 bp) regions of local similarity. A new sequence comparison algorithm is developed and described in Chapter 3. This algorithm has a linear computation complexity and sufficient sensitivity to detect short regions of local similarity between nucleotide sequences. The algorithm utilizes an asymmetric approach, when one of the compared sequences is presented in a form of oligonucleotide table, while the second sequence is in standard, linear form. A short window is moved along the linear sequence and all overlapping oligonucleotides of a constant length in the frame are compared for the oligonucleotide table. The result of comparison of two sequences is a single figure, which can be compared to a threshold. For each measure of sequence similarity a probability of false positive and false negative can be estimated. The algorithm was set up and implemented to recognize matching ESTs with overlapping regions of 40bp with 95% identity, which is better than resolution ability of contemporary EST clustering tools This algorithm was used as a sequence comparison engine for two EST clustering programs, described in Chapter 4. These programs implement two different strategies: stringent and loose clustering. Both are tested on small, but realistic benchmark data sets and show the results, similar to one of the best existing clustering programs, 02_cluster, but with a significant advantage in speed and sensitivity to small overlapping regions of ESTs. On three different CPUs the new algorithm run at least two times faster, leaving less singletons and producing bigger clusters. With parallel optimization this algorithm is capable of clustering millions of ESTs on relatively inexpensive computers. The loose clustering variant is a highly portable application, relying on third-party software for cluster assembly. It was built to the same specifications as 02_ cluster and can be immediately included into the STACKPack package for EST clustering. The stringent clustering program produces already assembled clusters and can apprehend alternatively processed variants during the clustering process.

No description available.