Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease causing degeneration of upper and lower motor neurons. Most ALS cases are sporadic; only 6% are associated with mutations in Cu, Zn superoxide dismutase (SOD1). It is believed, however, that sporadic and familiar forms of ALS share a common mechanism, where SOD1 plays an important role: SOD1 knockout mice do not develop ALS, whereas the overexpression of human SOD1 in mice produces ALS-like symptoms. Increasing evidence suggest that the SOD1 structure gains cytotoxic properties, but detailed description of the toxic species is missing. This thesis work is focused on understanding how structural and dynamic properties of SOD1 change along its folding free-energy landscape and indicates the structural hot-spots from where the cytotoxic species may originate. Thus, binding of the zinc controls folding, stability and turnover of SOD1: (i) miscoordination of Zn2+ by the Cu-ligands speeds up folding of the SOD1 core structure, however, it stabilizes SOD1 in a state where both active-site loops IV and VII are unfolded, (ii) coordination of Zn2+ in the Zn-site, induces the folding of loop VII and stabilizes the native and  functional fold of both active-site loops and (iii) the tremendous stability gain due to Zn-site metallation corresponds to a folded state’s lifetime of  &gt; 100 years, thus the cellular lifetime of SOD1 is likely controlled by Zn2+ release, which again is coupled to opening of active-site loops. Hence the active-site loops IV and VII stand out as critical and floppy parts of the SOD1 structure. Moreover, a number of ALS-associated mutations, benign to apo-SOD1 stability, are shown here to affect integrity of active-site loops in holo-SOD1, which, in turn, increases population of SOD1 species with these loops disorganized. Finally, the close relation between SOD1 and Zn2+ can also act in the reverse direction: a perturbed folding free-energy landscape of SOD1 can disturb Zn2+ homeostasis. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 3: Manuscript. Paper 4: Manuscript.</p>

The subject of this thesis is to develop a temporal neuro-fuzzy system for fore- casting the future behavior of a multivariate time series data. The system has two components combined by means of a system interface. First, a rule extraction method is designed which is named Fuzzy MAR (Multivari- ate Auto-regression). The method produces the temporal relationships between each of the variables and past values of all variables in the multivariate time series system in the form of fuzzy rules. These rules may constitute the rule-base in a fuzzy expert system. Second, a temporal neuro-fuzzy system which is named ANFIS unfolded in - time is designed in order to make the use of fuzzy rules, to provide an environment that keeps temporal relationships between the variables and to forecast the future behavior of data. The rule base of ANFIS unfolded in time contains temporal TSK(Takagi-Sugeno-Kang) fuzzy rules. In the training phase, Back-propagation learning algorithm is used. The system takes the multivariate data and the num- ber of lags needed which are the output of Fuzzy MAR in order to describe a variable and predicts the future behavior. Computer simulations are performed by using synthetic and real multivariate data and a benchmark problem (Gas Furnace Data) used in comparing neuro- fuzzy systems. The tests are performed in order to show how the system efficiently model and forecast the multivariate temporal data. Experimental results show that the proposed model achieves online learning and prediction on temporal data. The results are compared by other neuro-fuzzy systems, specifically ANFIS.

A novel broad spectrum neutron spectrometer has been created to extend the useful energy range of existing neutron Bonner Sphere Spectrometers (BSS). Both an active LiI(Eu) scintillator probe and passive gold foil detector were utilized to extend the energy response of the existing BSS. Above 20 MeV the energy structure for the standard polyethylene BSS are poor because the response functions are not unique. MCNPX was used to investigate several modifications to the BSS system which resulted in the Bonner Sphere Extension (BSE). This cost effective extension uses several concentric spheres of copper, lead, and tungsten heavy metal downscatter materials to extend the useful range of the current BSS from 20 MeV to above 1 GeV. Using both a 3" and 5" inner polyethylene spheres, aluminum shell sets were made with a 1" cavity and filled with the high Z materials for six total sets of spheres. a 12" and 8" polyethylene sphere were also milled to accept the heavy metal spheres. The system was validated at the Los Alamos Neutron Science Center (LANSCE) neutron beam. The system was calibrated at LANSCE for neutrons up to 800 MeV on target 4's 15 degree right flightpath (4FP15R) at 90 meters. Detailed models in MCNPX were made of the BSS, BSE and LANSCE facilities. Fine group responses were made and compared to the unfolded data from LANSCE. A vast improvement over the BSS system alone was seen with reasonable agreement with time of flight data measured at LANSCE and MCNPX calculated neutron spectra.

The size of data generated every year follows an exponential growth. The number of data points as well as the dimensions have increased dramatically the past 15 years. The gap between the demand from the industry in data processing and the solutions provided by the machine learning community is increasing. Despite the growth in memory and computational power, advanced statistical processing on the order of gigabytes is beyond any possibility. Most sophisticated Machine Learning algorithms require at least quadratic complexity. With the current computer model architecture, algorithms with higher complexity than linear O(N) or O(N logN) are not considered practical. Dimensionality reduction is a challenging problem in machine learning. Often data represented as multidimensional points happen to have high dimensionality. It turns out that the information they carry can be expressed with much less dimensions. Moreover the reduced dimensions of the data can have better interpretability than the original ones. There is a great variety of dimensionality reduction algorithms under the theory of Manifold Learning. Most of the methods such as Isomap, Local Linear Embedding, Local Tangent Space Alignment, Diffusion Maps etc. have been extensively studied under the framework of Kernel Principal Component Analysis (KPCA). In this dissertation we study two current state of the art dimensionality reduction methods, Maximum Variance Unfolding (MVU) and Non-Negative Matrix Factorization (NMF). These two dimensionality reduction methods do not fit under the umbrella of Kernel PCA. MVU is cast as a Semidefinite Program, a modern convex nonlinear optimization algorithm, that offers more flexibility and power compared to iv KPCA. Although MVU and NMF seem to be two disconnected problems, we show that there is a connection between them. Both are special cases of a general nonlinear factorization algorithm that we developed. Two aspects of the algorithms are of particular interest: computational complexity and interpretability. In other words computational complexity answers the question of how fast we can find the best solution of MVU/NMF for large data volumes. Since we are dealing with optimization programs, we need to find the global optimum. Global optimum is strongly connected with the convexity of the problem. Interpretability is strongly connected with local isometry1 that gives meaning in relationships between data points. Another aspect of interpretability is association of data with labeled information. The contributions of this thesis are the following: 1. MVU is modified so that it can scale more efficient. Results are shown on 1 million speech datasets. Limitations of the method are highlighted. 2. An algorithm for fast computations for the furthest neighbors is presented for the first time in the literature. 3. Construction of optimal kernels for Kernel Density Estimation with modern convex programming is presented. For the first time we show that the Leave One Cross Validation (LOOCV) function is quasi-concave. 4. For the first time NMF is formulated as a convex optimization problem 5. An algorithm for the problem of Completely Positive Matrix Factorization is presented. 6. A hybrid algorithm of MVU and NMF the isoNMF is presented combining advantages of both methods. 7. The Isometric Separation Maps (ISM) a variation of MVU that contains classification information is presented. 8. Large scale nonlinear dimensional analysis on the TIMIT speech database is performed. 9. A general nonlinear factorization algorithm is presented based on sequential convex programming. Despite the efforts to scale the proposed methods up to 1 million data points in reasonable time, the gap between the industrial demand and the current state of the art is still orders of magnitude wide.

Doctor of Philosophy (PhD) / In the last two decades, ambivalence has emerged as one of the primary concerns of attitude researchers. The acknowledgement that individuals can simultaneously evaluate an attitude object as both positive and negative has challenged a number of the status quos of the attitude literature. This thesis utilises an unfolding approach to investigate the implications of ambivalence for the conceptualisation of attitudes and their measurement. Firstly, the assumption that ambivalence is at odds with the bipolar understanding of attitudes was investigated. The results suggested that ambivalence is consistent with bipolarity, whereby ambivalent attitudes are located at the centre of the bipolar dimension. Secondly, attitude scales for the abortion, euthanasia and Aboriginal Australians issues were constructed to reflect this bipolar understanding. The fit of these statements to Coombs’ (1964) unidimensional unfolding theory provided evidence that ambivalence is also consistent with the quantitative conceptualisation of attitudes. Together these results provided further validation of the unfolding approach to attitude measurement. These models alleviate many of the problems faced by the ubiquitous method of summated ratings, including the assessment of ambivalent evaluations. Finally, these scales were used to investigate the latent structure of attitudes and its relationship with meta-psychological judgements of ambivalence. The conclusions drawn from these analyses were limited by a number of issues, highlighting the importance of rigorously considering measurement issues for all attitude parameters. Nonetheless, they presented preliminary validation of these scales’ ability to measure ambivalent evaluations and suggested a systematic relationship between proximity to the centre of the evaluative dimension and metapsychological ambivalence. Overall, it is concluded that a number of assumptions regarding the implications of ambivalence are misplaced and the proposed solutions are even more troublesome. Suggestions for future research are made, particularly with respect to differentiating between ambivalent, indifferent and uncertain evaluative states. Furthermore, attitude researchers are encouraged to discard their operationalist biases in order to engage with the measurement issues illuminated throughout this thesis. This is necessary to account for the complexity of the attitude construct, which ambivalence is testament to.

An examination of the logical relationships between attitude statements suggests that attitudes can be ordered according to favourability, and can also stand in relationships of implication to one another. The traditional representation of attitudes, as points on a single dimension, is inadequate for representing both these relations but representing attitudes as intervals on a single dimension can incorporate both favourability and implication. An interval can be parameterised using its two endpoints or alternatively by its midpoint and latitude. Using this latter representation, the midpoint can be understood as the �favourability� of the attitude, while the latitude can be understood as its �generality�. It is argued that the generality of an attitude statement is akin to its latitude of acceptance, since a greater semantic range increases the likelihood of agreement. When Coombs� Theory of Unidimensional Unfolding is reformulated using the interval representation, the key question is how to measure the distance between two intervals on the dimension. There are innumerable ways to answer this question, but the present study restricts attention to eighteen possible �distance� measures. These measures are based on nine basic distances between intervals on a dimension, as well as two families of models, the Minkowski r-metric and the Generalised Hyperbolic Cosine Model (GHCM). Not all of these measures are distances in the strict sense as some of them fail to satisfy all the metric axioms. To distinguish between these eighteen �distance� measures two empirical tests, the triangle inequality test, and the aligned stimuli test, were developed and tested using two sets of attitude statements. The subject matter of the sets of statements differed but the underlying structure was the same. It is argued that this structure can be known a priori using the logical relationships between the statement�s predicates, and empirical tests confirm the underlying structure and the unidimensionality of the statements used in this study. Consequently, predictions of preference could be ascertained from each model and either confirmed or falsified by subjects� judgements. The results indicated that the triangle inequality failed in both stimulus sets. This suggests that the judgement space is not metric, contradicting a common assumption of attitude measurement. This result also falsified eleven of the eighteen �distance� measures because they predicted the satisfaction of the triangle inequality. The aligned stimuli test used stimuli that were aligned at the endpoint nearest to the ideal interval. The results indicated that subjects preferred the narrower of the two stimuli, contrary to the predictions of six of the measures. Since these six measures all passed the triangle inequality test, only one measure, the GHCM (item), satisfied both tests. However, the GHCM (item) only passes the aligned stimuli tests with additional constraints on its operational function. If it incorporates a strictly log-convex function, such as cosh, the GHCM (item) makes predictions that are satisfied in both tests. This is also evidence that the latitude of acceptance is an item rather than a subject or combined parameter.

Doctor of Philosophy (PhD) / In the last two decades, ambivalence has emerged as one of the primary concerns of attitude researchers. The acknowledgement that individuals can simultaneously evaluate an attitude object as both positive and negative has challenged a number of the status quos of the attitude literature. This thesis utilises an unfolding approach to investigate the implications of ambivalence for the conceptualisation of attitudes and their measurement. Firstly, the assumption that ambivalence is at odds with the bipolar understanding of attitudes was investigated. The results suggested that ambivalence is consistent with bipolarity, whereby ambivalent attitudes are located at the centre of the bipolar dimension. Secondly, attitude scales for the abortion, euthanasia and Aboriginal Australians issues were constructed to reflect this bipolar understanding. The fit of these statements to Coombs’ (1964) unidimensional unfolding theory provided evidence that ambivalence is also consistent with the quantitative conceptualisation of attitudes. Together these results provided further validation of the unfolding approach to attitude measurement. These models alleviate many of the problems faced by the ubiquitous method of summated ratings, including the assessment of ambivalent evaluations. Finally, these scales were used to investigate the latent structure of attitudes and its relationship with meta-psychological judgements of ambivalence. The conclusions drawn from these analyses were limited by a number of issues, highlighting the importance of rigorously considering measurement issues for all attitude parameters. Nonetheless, they presented preliminary validation of these scales’ ability to measure ambivalent evaluations and suggested a systematic relationship between proximity to the centre of the evaluative dimension and metapsychological ambivalence. Overall, it is concluded that a number of assumptions regarding the implications of ambivalence are misplaced and the proposed solutions are even more troublesome. Suggestions for future research are made, particularly with respect to differentiating between ambivalent, indifferent and uncertain evaluative states. Furthermore, attitude researchers are encouraged to discard their operationalist biases in order to engage with the measurement issues illuminated throughout this thesis. This is necessary to account for the complexity of the attitude construct, which ambivalence is testament to.

Thesis (Ph. D.)--University of Missouri-Columbia, 2005. / The entire dissertation/thesis text is included in the research.pdf file; the official abstract appears in the short.pdf file (which also appears in the research.pdf); a non-technical general description, or public abstract, appears in the public.pdf file. Title from title screen of research.pdf file viewed on (July 19, 2006) Vita. Includes bibliographical references.

La dégradation intracellulaire des protéines est un processus fondamental qui a lieu dans tous les organismes, depuis les bactéries jusqu’aux êtres humains. La dégradation continuelle des protéines est nécessaire pour réguler les concentrations intracellulaires d’enzymes qui contrôlent toutes les réactions métaboliques, ainsi que le contenu général de toutes les autres protéines, en réponse aux modifications physiologiques. Un état d’équilibre dynamique se crée ainsi où la concentration intracellulaire d’une protéine peut être modulée aussi par des modifications de la vitesse de dégradation et de la vitesse de synthèse. Le travail présenté dans cette thèse porte sur le complexe d’activation du protéasome des cellules d’archaea (PAN). PAN est un complexe hexamerique énergie-dépendant découvert chez les archaeas et impliqué dans le dépliement des protéines substrat pour faciliter leur dégradation par le protéasome 20S. Toutes les études structurales du complexe PAN assemblé n’ont pas réussi à révéler la structure de la protéine à cause des difficultés rencontrées au niveau de la préparation et la stabilité des échantillons. La première partie du travail présenté a permis de déterminer une structure par Cryo-microscopie électronique et un modèle pseudo-atomique du complexe PAN hexamerique de Pyrococcus horikoshii. De plus, l’étude des différents états conformationnels du complexe induits par liaison du nucléotide ont permis de gagner plusieurs informations sur le mécanisme des unfoldases AAA+ et de proposer un mode d’action du complexe PAN. La seconde partie de l’étude a conduit à élucider la question sur la dynamique et les changements conformationnels des systèmes AAA+ unfoldases en général et du complexe PAN en particulier pendant le dépliement des protéines substrats. La méthode de variation de contraste en diffusion de neutrons aux petits angles (SANS) couplée avec de la spectroscopie de fluorescence appliqué à l’étude en temps réel du processus de dépliement de substrat par le système PAN de Methanococcus jannaschii a permis de révéler, pour la première fois, des mouvements de contraction de PAN pendant le dépliement du substrat induits par l’hydrolyse de l’ATP et suivis par une relaxation de la molécule à la fin du processus. Le mécanisme de dépliement par PAN semble être un mouvement de pompage péristaltique qui entraine un dépliement directionnel du substrat. L’ensemble de ce travail contribue à mieux comprendre la structure et le mécanisme d’action de la machine PAN dans les cellules et d’avoir une idée plus claire sur la dynamique fonctionnelle des complexes AAA+ à l’origine de leurs fonctions biologiques. / Intracellular protein degradation is a fundamental process occurring in all organisms, from bacteria to human. The continual degradation of proteins is necessary for regulating the intracellular levels of enzymes that control all metabolic reactions, as well as the general content of all other proteins, in response to physiological changes. A state of a dynamic equilibrium is created where the intracellular concentration of a protein can be modulated by changes in the synthesis rate as well as the degradation rate. The work presented in this thesis deals with the proteasome activating complex from archaeal cells (PAN). PAN is an energy dependent hexameric complex discovered in archaea and involved in the unfolding of protein substrates to facilitate their degradation by the 20S proteasome. All the previous structural studies on the assembled PAN complex have failed in revealing the structure of the whole complex because of the difficulties encountered during sample preparation and stabilization. In the first part of this work we determined a Cryo-electron microscopy structure and a pseudo-atomic model of the hexameric PAN complex from Pyrococcus horikoshii. In addition, the study of the different conformational states of the PAN complex induced by nucleotide binding helped in gaining several information about the AAA+ unfoldases mechanism and to propose a mode of action of the PAN complex. The second part of the study led to elucidate the question about the dynamic and the conformational changes of the AAA+ unfoldases in general and the PAN complex in particular. The method of contrast variation in Small Angle Neutron Scattering (SANS) coupled with online fluorescence spectroscopy applied to study, in real-time, the substrate unfolding process by the PAN complex from Methanococcus jannaschii allowed to reveal, for the first time, a contraction movements of PAN during substrate unfolding induced by ATP hydrolysis and followed by a relaxation of the molecule at the end of the process. The unfolding mechanism processed by PAN appears to be a peristaltic pumping motion that leads to a directional unfolding of the substrate. The whole work presented in this thesis contributes in understanding the structure and the mechanism of action of the PAN molecular machine inside the cells and to have a clearer idea about the functional dynamics of the AAA+ complexes at the origin of their biological functions.

abstract: The properties of materials depend heavily on the spatial distribution and connectivity of their constituent parts. This applies equally to materials such as diamond and glasses as it does to biomolecules that are the product of billions of years of evolution. In science, insight is often gained through simple models with characteristics that are the result of the few features that have purposely been retained. Common to all research within in this thesis is the use of network-based models to describe the properties of materials. This work begins with the description of a technique for decoupling boundary effects from intrinsic properties of nanomaterials that maps the atomic distribution of nanomaterials of diverse shape and size but common atomic geometry onto a universal curve. This is followed by an investigation of correlated density fluctuations in the large length scale limit in amorphous materials through the analysis of large continuous random network models. The difficulty of estimating this limit from finite models is overcome by the development of a technique that uses the variance in the number of atoms in finite subregions to perform the extrapolation to large length scales. The technique is applied to models of amorphous silicon and vitreous silica and compared with results from recent experiments. The latter part this work applies network-based models to biological systems. The first application models force-induced protein unfolding as crack propagation on a constraint network consisting of interactions such as hydrogen bonds that cross-link and stabilize a folded polypeptide chain. Unfolding pathways generated by the model are compared with molecular dynamics simulation and experiment for a diverse set of proteins, demonstrating that the model is able to capture not only native state behavior but also partially unfolded intermediates far from the native state. This study concludes with the extension of the latter model in the development of an efficient algorithm for predicting protein structure through the flexible fitting of atomic models to low-resolution cryo-electron microscopy data. By optimizing the fit to synthetic data through directed sampling and context-dependent constraint removal, predictions are made with accuracies within the expected variability of the native state. / Dissertation/Thesis / Ph.D. Physics 2011