Rather than designing focused experiments to test individual hypotheses, scientists now commonly acquire measurements using massively parallel techniques, for post hoc interrogation. The resulting data is both high-dimensional and structured, in that observed variables are grouped and ordered into related subspaces, reflecting both natural physical organization and factorial experimental designs. Such structure encodes critical constraints and clues to interpretation, but typical unsupervised learning methods assume exchangeability and fail to account adequately for the structure of data in a flexible and interpretable way. In this thesis, I develop computational methods for exploratory analysis of structured high-dimensional data, and apply them to study gene expression regulation in Parkinson’s (PD) and Huntington’s diseases (HD). BOMBASTIC (Block-Organized, Model-Based, Tree-Indexed Clustering) is a methodology to cluster and visualize data organized in pre-specified subspaces, by combining independent clusterings of blocks into hierarchies. BOMBASTIC provides a formal specification of the block-clustering problem and a modular implementation that facilitates integration, visualization, and comparison of diverse datasets and rapid exploration of alternative analyses. These tools, along with standard methods, were applied to study gene expression in mouse models of neurodegenerative diseases, in collaboration with Dr. Myriam Heiman and Dr. Robert Fenster. In PD, I analyzed cell-type-specific expression following levodopa treatment to study mechanisms underlying levodopa-induced dyskinesia (LID). I identified likely regulators of the transcriptional changes leading to LID and implicated signaling pathways amenable to pharmacological modulation (Heiman, Heilbut et al, 2014). In HD, I analyzed multiple mouse models (Kuhn, 2007), cell-type specific profiles of medium spiny neurons (Fenster, 2011), and an RNA-Seq dataset profiling multiple tissue types over time and across an mHTT allelic series (CHDI, 2015). I found evidence suggesting that altered activity of the PRC2 complex significantly contributes to the transcriptional dysregulation observed in striatal neurons in HD.

This dissertation studies prediction and classification models for disease risk through semiparametric modeling and statistical learning. It consists of three parts. In the first part, we propose several survival models to analyze the Cooperative Huntington's Observational Research Trial (COHORT) study data accounting for the missing mutation status in relative participants (Kieburtz and Huntington Study Group, 1996a). Huntington's disease (HD) is a progressive neurodegenerative disorder caused by an expansion of cytosine-adenine-guanine (CAG) repeats at the IT15 gene. A CAG repeat number greater than or equal to 36 is defined as carrying the mutation and carriers will eventually show symptoms if not censored by other events. There is an inverse relationship between the age-at-onset of HD and the CAG repeat length; the greater the CAG expansion, the earlier the age-at-onset. Accurate estimation of age-at-onset based on CAG repeat length is important for genetic counseling and the design of clinical trials for HD. Participants in COHORT (denoted as probands) undergo a genetic test and their CAG repeat number is determined. Family members of the probands do not undergo the genetic test and their HD onset information is provided by probands. Several methods are proposed in the literature to model the age specific cumulative distribution function (CDF) of HD onset as a function of the CAG repeat length. However, none of the existing methods can be directly used to analyze COHORT proband and family data because family members' mutation status is not always known. In this work, we treat the presence or absence of an expanded CAG repeat in first-degree family members as missing data and use the expectation-maximization (EM) algorithm to carry out the maximum likelihood estimation of the COHORT proband and family data jointly. We perform simulation studies to examine finite sample performance of the proposed methods and apply these methods to estimate the CDF of HD age-at-onset from the COHORT proband and family combined data. Our results show a slightly lower estimated cumulative risk of HD with the combined data compared to using proband data alone. We then extend the approach to predict the cumulative risk of disease accommodating predictors with time-varying effects and outcomes subject to censoring. We model the time-specific effect through a nonparametric varying-coefficient function and handle censoring through self-consistency equations that redistribute the probability mass of censored outcomes to the right. The computational procedure is extremely convenient and can be implemented by standard software. We prove large sample properties of the proposed estimator and evaluate its finite sample performance through simulation studies. We apply the method to estimate the cumulative risk of developing HD from the mutation carriers in COHORT data and illustrate an inverse relationship between the cumulative risk of HD and the length of CAG repeats at the IT15 gene. In the second part of the dissertation, we develop methods to accurately predict whether pre-symptomatic individuals are at risk of a disease based on their various marker profiles, which offers an opportunity for early intervention well before definitive clinical diagnosis. For many diseases, existing clinical literature may suggest the risk of disease varies with some markers of biological and etiological importance, for example age. To identify effective prediction rules using nonparametric decision functions, standard statistical learning approaches treat markers with clear biological importance (e.g., age) and other markers without prior knowledge on disease etiology interchangeably as input variables. Therefore, these approaches may be inadequate in singling out and preserving the effects from the biologically important variables, especially in the presence of potential noise markers. Using age as an example of a salient marker to receive special care in the analysis, we propose a local smoothing large margin classifier implemented with support vector machine to construct effective age-dependent classification rules. The method adaptively adjusts age effect and separately tunes age and other markers to achieve optimal performance. We derive the asymptotic risk bound of the local smoothing support vector machine, and perform extensive simulation studies to compare with standard approaches. We apply the proposed method to two studies of premanifest HD subjects and controls to construct age-sensitive predictive scores for the risk of HD and risk of receiving HD diagnosis during the study period. In the third part of the dissertation, we develop a novel statistical learning method for longitudinal data. Predicting disease risk and progression is one of the main goals in many clinical studies. Cohort studies on the natural history and etiology of chronic diseases span years and data are collected at multiple visits. Although kernel-based statistical learning methods are proven to be powerful for a wide range of disease prediction problems, these methods are only well studied for independent data but not for longitudinal data. It is thus important to develop time-sensitive prediction rules that make use of the longitudinal nature of the data. We develop a statistical learning method for longitudinal data by introducing subject-specific long-term and short-term latent effects through designed kernels to account for within-subject correlation of longitudinal measurements. Since the presence of multiple sources of data is increasingly common, we embed our method in a multiple kernel learning framework and propose a regularized multiple kernel statistical learning with random effects to construct effective nonparametric prediction rules. Our method allows easy integration of various heterogeneous data sources and takes advantage of correlation among longitudinal measures to increase prediction power. We use different kernels for each data source taking advantage of distinctive feature of data modality, and then optimally combine data across modalities. We apply the developed methods to two large epidemiological studies, one on Huntington's disease and the other on Alzhemeier's Disease (Alzhemeier's Disease Neuroimaging Initiative, ADNI) where we explore a unique opportunity to combine imaging and genetic data to predict the conversion from mild cognitive impairment to dementia, and show a substantial gain in performance while accounting for the longitudinal feature of data.

A dissertation submitted to the Faculty of Engineering and the Built Environment, University of the Witwatersrand, Johannesburg, in fulfilment of the requirements for the degree of Master of Science in Engineering. Johannesburg, 2017 / This study presents the development of a potentially new biomarker for three different movement disorders: Huntington’s Disease (HD), Parkinson’s Disease (PD) and Essential Tremors (ET). A Leap Motion® gaming device was used to record the trajectories of subjects’ forefinger as they trace simple patterns in the air. The patterns used were stepfunction, triangle and circle. The recorded signals were analysed using transform functions and Fourier analysis. Both analysis types yielded features from which differences between the four categories studied: PD, HD, ET and control subjects, were sought and displayed in both graphical and numerical forms. The X-axis and Y-axis of the signals were separately analysed and yielded different results. For the step-function pattern, no distinct differences between the four categories were found from the transfer function analysis whereas the Y-axis of the signal could distinguish between the categories. For the triangle pattern, the X-axis features provided a discrimination between the categories while the Y-axis feature did not. For the circle pattern, neither X-axis nor Y-axis features were able to distinguish between the categories. A Fourier analysis showed a better discrimination ability for both X- and Y- axis. This study is a preliminary one and all results indicate that more subjects of all categories are needed to develop a bio-marker for the diseases studied and that a higher order transfer function analysis is required. However, the methodology outlined in this work, comprising of both the experimental system and the analysis showed a potential to produce a biomarker for movement disorders. / MT2018

The logistical complexities of performing multi-site longitudinal diffusion-weighted imaging (DWI) studies requires careful construction of analysis tools and procedures. Proposed clinical trials for therapies in neurodegenerative disease are known to re- quire several hundred subjects, thus mandating multiple site participation to obtain sufficient sample sizes. DWI is an important tool for monitoring diffusivity properties of white matter (WM) in disease progression. The multi-site nature of clinical trials requires new strategies in DWI processing and analysis to reliably measure longitudi- nal WM changes. This work describes the process of developing and validating robust analysis methodologies to process multi-site DWI data in a rare, neurodegenerative disease. Key processing components to accomplish a robust DWI processing system include: DICOM conversion, automated quality control, unbiased atlas construction, fiber tracking, and statistical analysis. Extensive validation studies were performed to characterize methodological results within and across the common confounds inherent in multi-site clinical trials. The conversion and automated quality control tools optimized for this work both enhanced the ability to reliably obtain repeat diffusion tensor image (DTI) scalar measurements in a reliability analysis of healthy controls scanned at multiple sites using multiple scanner vendors. A DTI scalar analysis performed on focused WM regions showed it was possible to detect significant mean differences of DTI scalars among separate groups of a neurodegenerative disease population. The DTI scalar analysis paved the way for an atlas-based cross-sectional fiber tracking analysis. In the cross-sectional fiber tracking analysis, multi-site data was brought into the same space, making major fiber tracts terminating in the focused WM regions of the scalar analysis from all participants comparable. Significant differences in diffusivity were found throughout each tract among separate groups of the neurodegenerative disease population. In addition, multiple neuropsychological cognitive variables that have a documented ability to track disease progression of the neurodegenerative disease, strongly correlated with many of the DTI scalars in each tract. The findings of the cross-sectional fiber tracking analysis were reinforced by similar findings produced by a longitudinal fiber tracking analysis. Collectively, these findings suggest that cogni- tive deficits seen in the neurodegenerative disease population could be explained by changes in diffusivity of the tracts explored in this work. In addition to the longi- tudinal fiber tracking analysis examining diffusivity, methods for a WM morphology analysis using parallel transport to apply longitudinal volume changes to a template was explored.

<p><strong>Objective: </strong>To evaluate the effect of a physiotherapeutic exercise programme for patients with Huntington´s Disease (HD) concerning motor function and disability, balance and fall related self-efficacy, and to investigate the correlation between the seven assessment tools.</p><p><strong>Participants:</strong> Twelve persons with genetically confirmed HD at an early or middle stage of the disease and with a mean age of 52 (16) years.</p><p><strong>Methods:</strong> The intervention comprised physiotherapy (PT) focused on training of transitions, balance and fall-related self efficacy, twice a week for six weeks. Each treatment session lasted for one hour, was individual and took place at an out-patient clinic. Baseline assessments including five clinical tests and two questionnaires were made 6 and 0 weeks prior to the intervention and 0 and 6 weeks after the intervention.</p><p><strong>Outcome measures:</strong> Motor function and disability were measured with the Unified Huntington's disease Rating Scale; the Total Motor Score and the Total Functional Assessment. Static and dynamic balance was measured with the One- leg stance- test, the Timed Up and GO Test, the Figure of Eight-test and the Berg Balance Scale.  Fall-related self-efficacy was measured with the Falls Efficacy Scale.</p><p><strong>Results:</strong> The physiotherapeutic exercise programme demonstrated a significant improvement in balance measured with the Berg Balance Scale (<em>p=.045). </em>The significant correlation coefficients between the different measurements of motor function, disability, balance and fall related self-efficacy ranged from 0.68 to 0.87.</p><p><strong>Conclusions: </strong>The contents of the out-patient clinic physiotherapeutic exercise programme, with a focus on balance and transitions, seemed to have clinical relevance. PT in different kinds of settings should be studied further to get a better knowledge about the effects of PT and physical activity at home, at an out- patient setting or at the hospital for patients with HD.</p><p><strong>Key Words: </strong>Huntington's disease; Physiotherapy; Motor function; Disability; Balance; Fall- related self efficacy.</p>

Mitochondrial respiratory chain deficiency and increased oxidative stress have been closely associated with major age-associated neurodegenerative diseases. I hypothesized that mitochondrial oxidative phosphorylation defects or elevated oxidative stress, which could arise in a stochastic manner during our normal aging process, might modulate the formation of protein aggregates or production of misfolded proteins, contributing to the initiation of these diseases. To test this hypothesis, we (i) have developed and characterized mouse and cellular models of Alzheimer's and Huntington's diseases expressing aggregate-prone pathogenic proteins, beta-amyloid and mutant huntingtin (Chapters 1 and 2), (ii) have developed mouse models that exhibit neuron-specific defects in mitochondrial oxidative phosphorylation (Chapters 2 and 3), and (iii) have evaluated the alterations in the amount of aggregate loads upon genetic and pharmacological manipulations of mitochondrial oxidative phosphorylation activities (Chapters 1 and 2). The evaluation of the impacts of mitochondrial defects on the amount of huntingtin aggregates has revealed that a defect in complex III promotes the accumulation of huntingtin aggregates via the impairment of proteasome activity (Chapter 1). On the other hand, ablation of complex IV activity in a subset of postmitotic neurons revealed that complex IV deficiency does not promote either oxidative stress or the deposition of amyloid plaques in a mouse model of Alzheimer's disease, questioning the mitochondrial origin of Alzheimer's disease (Chapter 2). However, as shown previously, the tight correlation between oxidative stress and accumulation of amyloid plaques was found. Chapter 3 involved the generation of an improved mouse model, in which mitochondrial defects can be induced in a subset of forebrain neurons (cortex, hippocampus, and striatum) in a doxycycline-dependent manner. This system relies on the regulated expression of a mitochondria-targeted restriction enzyme, PstI, which digests mitochondrial DNA and thereby impairs the activity of oxidative phosphorylation. In conclusion, our studies highlighted the disease-specific complex pathways that may modulate the accumulation of misfolded proteins during aging. Future studies employing the newly-developed mouse model may reveal a contribution of age-associated global defects of oxidative phosphorylation to oxidative stress and neurodegenerative diseases.

Objective: To evaluate the effect of a physiotherapeutic exercise programme for patients with Huntington´s Disease (HD) concerning motor function and disability, balance and fall related self-efficacy, and to investigate the correlation between the seven assessment tools. Participants: Twelve persons with genetically confirmed HD at an early or middle stage of the disease and with a mean age of 52 (16) years. Methods: The intervention comprised physiotherapy (PT) focused on training of transitions, balance and fall-related self efficacy, twice a week for six weeks. Each treatment session lasted for one hour, was individual and took place at an out-patient clinic. Baseline assessments including five clinical tests and two questionnaires were made 6 and 0 weeks prior to the intervention and 0 and 6 weeks after the intervention. Outcome measures: Motor function and disability were measured with the Unified Huntington's disease Rating Scale; the Total Motor Score and the Total Functional Assessment. Static and dynamic balance was measured with the One- leg stance- test, the Timed Up and GO Test, the Figure of Eight-test and the Berg Balance Scale.  Fall-related self-efficacy was measured with the Falls Efficacy Scale. Results: The physiotherapeutic exercise programme demonstrated a significant improvement in balance measured with the Berg Balance Scale (p=.045). The significant correlation coefficients between the different measurements of motor function, disability, balance and fall related self-efficacy ranged from 0.68 to 0.87. Conclusions: The contents of the out-patient clinic physiotherapeutic exercise programme, with a focus on balance and transitions, seemed to have clinical relevance. PT in different kinds of settings should be studied further to get a better knowledge about the effects of PT and physical activity at home, at an out- patient setting or at the hospital for patients with HD. Key Words: Huntington's disease; Physiotherapy; Motor function; Disability; Balance; Fall- related self efficacy.

Polyglutamine-expanded fragments, derived from the human huntingtin protein, are aggregation-prone and toxic in yeast cells, bearing endogenous QN-rich proteins in the aggregated (prion) form. Attachment of the proline-rich region targets polyglutamine aggregates to the large perinuclear deposit (aggresome). Aggresome targeting ameliorates polyglutamine cytotoxicity in the presence of the prion form of Rnq1 protein, however, aggresome-forming construct remains toxic in the presence of the prion form of translation termination (release) factor Sup35 (eRF3). Disomy by chromosome II partly ameliorates polyglutamine toxicity in the strains containing Sup35 prion. The chromosome II gene, coding for another release factor, and interaction partner of Sup35, named Sup45 (eRF1), is responsible for amelioration of toxicity. Plasmid-mediated overproduction of Sup45, or expression of the Sup35 derivative that lacks the QN-rich domain and is unable to be incorporated into prion aggregates, also ameliorate polyglutamine toxicity. Protein analysis indicates that polyglutamines alter aggregation patterns of the Sup35 prion and promote aggregation of Sup45, while excess Sup45 counteracts these effects. In the absence of Sup35 prion, disomy by chromosome II is still able to decrease polyglutamine toxicity. However, SUP45 is no longer the gene responsible for such an effect. Taken together with the finding that the presence of both the Rnq1 prion and the Sup35 prion has an additive effect on polyQ toxicity, one gene or few genes on chromosome II are able to ameliorate polyQ toxicity through a SUP45-independent pathway. The identification of such a gene is currently ongoing. Monosomy by chromosome VIII in diploid heterozygous by AQT (Anti-polyQ Toxicity mutants that are disomic by chromosome II) counteracted the effect of AQT. Similarly, deletion of the arg4 gene in chromosome VIII in AQT haploid was able to eliminate the AQT effect. Moreover, analysis of genes involved in the arginine and polyamine synthesis indicated that loss of genes in later stages of arginine biosynthesis causes increase of polyglutamine toxicity. Deletion of genes arg1, arg4, arg8 (arginine pathway) and spe1 (polyamine pathway) all suppressed the Sup35 prion phenotype expression in the nonsense suppression system. Further analysis regarding the mechanisms behind those effects is needed. Our data uncover the mechanisms by which genetic and epigenetic factors may influence polyglutamine toxicity, and demonstrate that one and the same type of polyglutamine deposits could be cytoprotective or cytotoxic, depending on the prion composition of a eukaryotic cell.

Expanded polyglutamine proteins are known to be the causative agents of a number of human neurodegenerative diseases but the molecular basis of their cytoxicity is still poorly understood. Polyglutamine tracts may impede the activity of the proteasome, and evidence from single cell imaging suggests that the sequestration of polyglutamine proteins into inclusion bodies can reduce the proteasomal burden and promote cell survival, at least in the short term. The presence of misfolded protein also leads to activation of stress kinases such as p38MAPK, which can be cytotoxic. The relationships of these systems are not well understood. We have used fluorescent reporter systems imaged in living cells, and stochastic computer modeling to explore the relationships of expanded polyglutamine proteins, p38MAPK activation, generation of reactive oxygen species (ROS), proteasome inhibition, and inclusion body formation. In cells expressing a polyglutamine protein, inclusion body formation was preceded by proteasome inhibition but cytotoxicity was greatly reduced by administration of a p38MAPK inhibitor. Computer simulations suggested that without the generation of ROS, the proteasome inhibition and activation of p38MAPK would have significantly reduced toxicity. Our data suggest a vicious cycle of stress kinase activation and proteasome inhibition that is ultimately lethal to cells. There was close agreement between experimental data and the predictions of a stochastic computer model, supporting a central role for proteasome inhibition and p38MAPK activation in inclusion body formation and ROS-mediated cell death.

Expanded polyglutamine proteins are known to be the causative agents of a number of human neurodegenerative diseases but the molecular basis of their cytoxicity is still poorly understood. Polyglutamine tracts may impede the activity of the proteasome, and evidence from single cell imaging suggests that the sequestration of polyglutamine proteins into inclusion bodies can reduce the proteasomal burden and promote cell survival, at least in the short term. The presence of misfolded protein also leads to activation of stress kinases such as p38MAPK, which can be cytotoxic. The relationships of these systems are not well understood. We have used fluorescent reporter systems imaged in living cells, and stochastic computer modeling to explore the relationships of expanded polyglutamine proteins, p38MAPK activation, generation of reactive oxygen species (ROS), proteasome inhibition, and inclusion body formation. In cells expressing a polyglutamine protein, inclusion body formation was preceded by proteasome inhibition but cytotoxicity was greatly reduced by administration of a p38MAPK inhibitor. Computer simulations suggested that without the generation of ROS, the proteasome inhibition and activation of p38MAPK would have significantly reduced toxicity. Our data suggest a vicious cycle of stress kinase activation and proteasome inhibition that is ultimately lethal to cells. There was close agreement between experimental data and the predictions of a stochastic computer model, supporting a central role for proteasome inhibition and p38MAPK activation in inclusion body formation and ROS-mediated cell death.