22689 research outputs found
Sort by
DETECTING, CHARACTERIZING, AND TARGETING IMMUNE OLIGODENDROGLIA USING NOVEL MAJOR HISTOCOMPATIBILITY COMPLEX REPORTER MICE
No full textIneffective remyelination in multiple sclerosis (MS) is thought to contribute to neurodegeneration, underlying progressive disability accumulation. While many pro-myelinating drugs have been identified, few have been successful in clinical trials, and promoting remyelination remains a central goal of MS research. Our lab and others recently identified a population of immune oligodendroglia (iOL), characterized by the expression of major histocompatibility complex (MHC) proteins normally restricted to immune cells. These iOL may contribute to inflammation and further prevent remyelination. We sought to better understand these cells by developing novel reporter mouse lines for MHC class I and II. We found that MHC class I-expressing oligodendroglia are present in the naïve brain and spinal cord, while MHC class II-expressing oligodendroglia are specific to the inflamed central nervous system. Both MHC class I and II-expressing oligodendroglia are more abundant in disease contexts, specifically in areas of increased inflammation and immune cell infiltration. We next sought to determine if iOL induction varies with age, given that progressive MS is associated with aging. We found that at baseline, MHC class I-expressing oligodendroglia were present in the CNS at all ages, while MHC class II-expressing cells were only found in the aging spinal cord. After inducing inflammation using an interferon gamma (IFN) adeno-associated virus, we showed that MHC class I induction is consistent across age while MHC class II induction is once again elevated in aging oligodendroglia. This suggest that MHC class II-expressing iOLs may have a unique role in progressive MS. Finally, we leveraged the reporter mice to develop a high-throughput in vitro screening assay to evaluate MHC induction in oligodendroglia. We tested three existing MS therapies and show both dimethyl fumarate and interferon modulate MHC induction in the presence of IFN, highlighting the importance of testing drugs directly on oligodendroglia under disease-relevant conditions. We conclude that understanding the role of iOLs using the tools developed herein will be critical to meeting the unmet need for progressive MS therapies
THE COSMOS AND MAN: THE EMPIRE AND THE EMPEROR IN EARLY IMPERIAL THOUGHT
Full text linkRoman history is traditionally divided between two periods: the Republic, which refers to the time from the ejection of the monarchy until the ascension of Augustus; and the Empire, which refers to the time from the ascension of Augustus to power. In the period of the Principate, the figure of the emperor developed and became central to the governing of the empire. Imperial literature and propaganda of this period expressed the idea that Roman Empire and Augustus and his successors had been knit into the cosmos and had achieved a kind of permanence. This period was also witness to a flowering of moral and philosophical thought. In this dissertation, I examine the confluence of these two themes in the way three authors—Seneca, Plutarch, and Marcus Aurelius—wrote about the empire and the figure of the emperor. I argue that all these authors viewed the Roman Empire in metaphysical terms and explored this theme in their writing. These writers were deeply concerned with ethics and wrote about the emperor as a moral figure. However, the Roman Empire’s extent and seeming permanence formed an essential context for this moral thought. Political philosophy, in the context of an apparently permanent political structure, becomes coterminous with personal ethics. I argue that it is possible to discern in their writings about the emperor an expression of the emperor’s uneasy status in a kind of twofoldness to his character, expressing a reification of his public and private selves
VACCINE BENEFIT AND RISK ASSESSMENT FOR AN INFECTIOUS DISEASE WITH PANDEMIC POTENTIAL USING QUALITY-ADJUSTED LIFE YEAR
Full text linkDuring the COVID-19 pandemic, vaccines, including messenger ribonucleic acid (mRNA) vaccines, were rapidly developed, manufactured, and delivered. Even if very safe and effective vaccines are developed for the next infectious disease pandemic, populations may not be protected without a good acceptance of the new vaccine based on a solid scientific benefit-risk assessment. In this four-manuscript dissertation, a model development of vaccine risk and benefit evaluation for an infectious disease with pandemic potential was prepared followed by discussions regarding policy and practical implications of vaccine benefit risk assessment to prepare for the next infectious disease pandemic. The overarching research objective in this dissertation was to develop a model for vaccine risk and benefits for an infectious disease with pandemic potential to rapidly compare the risk and benefit of a new vaccine using a single health outcome scale.
In the first aim, the global risk of myocarditis, pericarditis and myopericarditis attributable to COVID-19 vaccination was evaluated. Young males had the highest risk attributable to COVID-19 vaccination. The quantitatively evaluation of attributable risks of myocarditis stratified by age group, sex, vaccine dose, and vaccine type helped develop a model with quantitative estimation of burden associated with vaccine adverse reactions. In the second aim, health utility of outcomes related with infectious diseases with pandemic potentials and associated vaccine adverse reactions were evaluated.
Following two systematic reviews, this study developed a model for vaccine benefit risk assessment stratified by age, sex, and the presence of medical comorbidity by country or region level. A quantitative benefit risk assessment of vaccination can facilitate a straightforward comparison of vaccine benefits and risks. While the societal perspective is fundamental to evaluating vaccine benefits and risks as a population, it may not always support individual vaccine decision making. Benefit risk assessment of vaccination from an individual perspective may be helpful for some individuals. In the next pandemic, a rapid vaccine benefit risk assessment is required, followed by sequential analyses to update dynamic changes regarding epidemiology as well as vaccine effectiveness and safety data
STATISTICAL CONNECTOMICS: DEVELOPING METHODS TOWARDS UNDERSTANDING POPULATIONS OF NETWORKS
No full textAdvances in in vivo brain imaging techniques have empowered neuroscientists to generate extensive datasets characterizing brains. Detailed maps of neural connectivity – termed connectomes - offer two fundamental perspectives: structural connectomes chart the physical wiring of the brain, while functional connectomes map patterns of correlated activity across brain regions. Investigating this diverse population of connectomes is vital for deciphering the principles guiding brain development, the neurological changes associated with disease, how our brains are shaped by experience, and the evolutionary trajectory of the human brain. Despite the progress in measuring connectomes at the scale at large, techniques for extracting meaning from these complicated datasets have lagged behind.
This thesis develops novel methods to enhance our understanding of connectome populations, with a focus on human structural and functional networks. It begins by reviewing existing statistical models and algorithms for analyzing connectomes and networks as well as improvements to existing algorithms for comparing connectomes. Next, it presents novel methods for investigating the heritability of structural connectomes within a causal analysis framework, while leveraging statistical modeling for connectomes. Additionally, we introduce a new method to quantify temporal dependencies within functional connectomes using generalized correlation measures. To foster broader research impact, these tools are made accessible to the neuroscience community and beyond through a documented, tested, open-source Python package.
Collectively, this thesis advances the algorithmic analysis of connectome data, and this work brings us closer to a better understanding of the brain's intricate workings. As the field of connectomics expands, generating increasingly complex and large-scale datasets, these sophisticated analytical methods will be essential
Developing compressed linear pangenome indexes for rapid sequence classification
Full text linkA reference genome serves an important function for various genomic analyses; it acts as a template to be used to match sequencing reads to the genome and provides a coordinate system to help translate findings from one study to another.
However, being overly reliant on a single reference genome leads to an issue called "reference bias" where one's findings can be biased due to the genetic differences between the reference and donor genome.
In order to combat this bias, the community has worked on assembling a multitude of reference genomes spanning a wide array of genetic backgrounds in order to build a pangenome reference.
My thesis work will focus on the problem of trying to quickly map sequencing reads onto these large pangenome databases using a compressed linear pangenome index.
The first half of the thesis will present novel computational methods for quickly classifying whether a sequencing read appears to have originated from a pangenome reference.
We present an efficient string matching algorithm computing a quantity called pseudo-matching lengths and develop a hypothesis testing framework for classifying whether reads are present or not in the database.
We show how to integrate the concepts of minimizer digestion and run-length encoding to build an efficient and scalable full-text index for querying.
Utilizing these novel methods, we show that we can achieve comparable binary classification accuracy to state-of-the-art aligners while being substantially faster and more memory-efficient.
The second half of the thesis will focus on specifically identifying where in a pangenome reference a read appears to originate from and we explore three different solutions for this problem.
Firstly, we develop a novel data-structure that scales with pangenomes and allows users to identify a single genome that a substring match from a read occurs in thereby giving users information to classify which genome the full read is from.
Secondly, in contrast to the previous solution, we theorized and implemented a new document listing data-structure which provides the full scope of information by allowing users to identify all of the genomes that a substring occurs in.
Lastly, we showed a novel compression scheme that can reduce the size of the document listing data-structure by over two orders of magnitude.
We utilize these new data-structures in the application of taxonomic classification and show that we achieve a higher classification accuracy over state-of-the-art tools
CHARACTERIZATION AND BIOLOGICAL EVALUATION OF ABSORBABLE POLYMERS
No full textThe use of absorbable polymers in medical devices has garnered increasing interest due to their applications in sutures, cardiovascular devices, orthopedic fixation devices, and surgical mesh. However, evaluating these polymers presents significant challenges as their material properties and the identity and quantity of degradation products continuously change over time.
This project investigates the impact of polymer degradation on biological responses throughout the device degradation process and seeks to correlate polymer material properties and degradation products with biological responses. The chosen polymer, poly(lactic-co-glycolic acid) (PLGA 50:50), was degraded at 37°C and 45°C to simulate biologically relevant and accelerated conditions, respectively. Section 1 introduces the concept of absorbable polymers and their growing role in the biomedical field. It discusses various degradation mechanisms, current advantages and disadvantages of absorbable medical devices, and existing assessment methods. Section 2 details the fabrication of PLGA 50:50 test articles and its characterization throughout the degradation process. Mass loss and pH changes of the supernatant were recorded at each timepoint. Changes in surface morphology were studied using scanning electron microscopy (SEM), while contact angle and protein absorption were also investigated. These characterizations support the premise that PLGA 50:50 undergoes bulk erosion and surface property changes during hydrolytic degradation. Section 3 examines the cytotoxicity of the two monomers that can be formed during the degradation processes, lactic acid and glycolic acid. Human Coronary Artery Endothelial Cells were exposed to increasing concentrations of lactic acid and glycolic acid. In vitro assays were used to determine cell viability and permeability relative to the untreated cell controls, assessing the cytotoxicity of these degradation products.
In conclusion, while lactic and glycolic acid monomers are naturally metabolized by the body, the acidic environment produced by these degradants can negatively affect cells. This suggests the need for adequate assessment of the impact of polymer degradation on biological responses in the intended clinical use environment for medical devices utilizing absorbable polymers
Unpacking and Healing Math Trauma to Improve the Efficacy of Novice Alternate Route Elementary Teachers
Full text linkThis qualitative study used semi-structured interviews to examine the motivation, lived experiences, and math efficacy of eight current and former alternate-route elementary teachers associated with Teach for America, a national teacher recruitment and leadership development organization. Through the lens of the networked ecological systems theory, the study explores constructs of motivation, sustainability, math efficacy, lived experience, and teacher development to better understand the root causes of math anxiety and academic trauma for this population. Thematic analysis was conducted to arrive at five themes: novice alternate-route elementary teachers are motivated by early exposure to social justice and service, as well as an affinity to the mission of the highly competitive teacher leadership development organization. Other themes discovered were that parents were important early socializers of mathematical identity and that current and former alternate-route elementary teachers had strong recommendations for professional development that leveraged content development and psychological math anxiety reduction strategies. Based on these findings and literature review on teacher wellness and professional development for the reduction of math anxiety, an intervention was designed in partnership with a licensed mental health professional to support novice alternate-route elementary teachers in healing past academic traumas and increasing their efficacy in math teaching for the youngest students
Learning the Sequence Determinants of Mammalian Transcriptional Gene Regulation Across Cell-Types
Full text linkGene expression is controlled by cis-regulatory DNA elements that contain binding sites for a class of proteins known as transcription factors. The set of transcription factors that bind one of these DNA elements determine its regulatory function. Mutations to these binding sites have been associated with or directly cause complex and common human disease. Prediction of the impact of regulatory variants requires knowledge of the sequence determinants of regulatory activity. Many machine learning algorithms have been developed to predict regulatory activity directly from DNA sequence. Although these models regularly achieve high predictive performance, they are difficult to biologically interpret due to their complexity.
This dissertation is focused on the performance and interpretation of these models. First, I compare current sequence-based models on their ability to learn TFBSs and investigate their strengths and weakness. Then, to extract biological and compact features from these models, I developed a novel algorithm, gkmPWM, which learns individual Transcription factor binding site (TFBS) information by modelling gapped kmer distributions. I mathematically derive the algorithm and present a Lagrangian optimization method that extracts all the TFBS learned in these models de novo. Lastly, I show that gkmPWM outperforms other methods that learn TFBS sequence features.
In the second part of my dissertation, I apply gkmPWM to a wide range of experiments to derive sequence rules for different classes of regulatory elements. I characterize the cell-type independent binding of promoters and insulators with a small set of transcription factors. Additionally, I identify the combinations of TFBSs that determine the cell-type specific activity of enhancers. I also use gkmPWM to learn the sequence preferences of different functional characterization assays and show that reporter assays have unique sequence preferences.
In the last part of my dissertation, I map the TFBSs learned by gkmPWM to their specific positions in regulatory elements at nucleotide resolution. I present a dynamic programming algorithm to efficiently map combinations of binding sites using information from gkm-SVM models. I show that these TFBS predictions align with experiments that target specific TFBSs. I mapped TFBSs to distal enhancers in a wide range of cell-types, which are publicly available
Symmetry in Graph Learning
No full textGraphs are ubiquitous representations to capture complex interactions and structural patterns. Graph neural networks have emerged as popular machine learning tools to learn and reason with graph data. Despite their promise and some success stories, graph neural networks have not managed to absolutely outperform classical methods such as spectral embeddings and graph kernels. Moreover, it is not completely clear when and why graph neural networks work or fail.
In this dissertation, we develop principled graph neural networks by exploiting symmetries. Symmetries appear naturally in graph learning, such as permutation symmetry due to the choice of node labeling, or graph automorphisms arising from self-similarity. Leveraging symmetry can offer elegant descriptions of physical objects and encourage abstract reasoning -- a key feature of intelligence. Therefore, we enforce symmetry in graph neural networks to ensure their performance and rapid generalization to novel situations. Furthermore, symmetry is a fundamental concept in mathematics, spanning across geometry, algebra, and probability. Thus, we utilize rich mathematical insights of symmetry to understand and improve graph neural networks.
Using the notion of permutation symmetry, we study the expressivity of graph neural networks, leading to stronger architectures by incorporating graph spectral invariants or faster algorithms by using random graph embeddings. Beyond permutation symmetry, we analyze the generalization properties of equivariant graph networks when choosing different kinds of natural symmetries induced from the graphs. Leveraging probabilistic symmetry, we evaluate graph neural networks based on novel random graph models arising from joint exchangeability. Empirically, we validate our theoretical insights in numerous graph learning applications across social science, chemistry, and computer vision. We conclude by discussing other notions of symmetries and future research directions that exploit symmetry within and beyond graph learning
MCRIP2: A REGULATOR OF BROWN ADIPOSE TISSUE OXIDATIVE CAPACITY AND THERMOGENESIS
Full text linkNon-shivering thermogenesis, the brown adipose tissue (BAT) specific process of heat generation, relies on mitochondrial function. Therefore, mitochondrial content and activity within BAT displays a high degree of plasticity, mediated by cold-dependent adrenergic signaling and the activation of various transcription factors. Work from our lab has shown that a subfamily of nuclear receptors, called Estrogen Related Receptors (ERRα/β/γ), interact with Peroxisome Proliferator Activated Receptor (PPAR) gamma Coactivator 1α (PGC-1α) to remodel BAT gene expression in response to cold. Specifically, ERR/PGC-1α complexes are necessary for basal expression of oxidative phosphorylation, TCA cycle, and mitochondria-associated genes, as well as their full induction in response to adrenergic signaling in BAT. In addition to these targets, ERRs and PGC-1α also regulate genes with currently unknown roles in BAT, that are not predicted to have mitochondrial function, and, therefore, may be co-regulators of the ERR/PGC-1α program. Identification and characterization of these putative regulators will provide a more comprehensive understanding of the pathways modulating BAT thermogenic capacity.
MAPK Regulated Corepressor Interacting Protein 2, (MCRIP2), is a previously unexplored ERR/PGC-1α target whose putative protein interactors suggest it could provide insight into a post-transcriptional regulatory mechanism in BAT. In my thesis, I show that MCRIP2 is required for the expression of many mitochondrial genes, and overall BAT oxidative and thermogenic capacities. Primary brown adipocytes and BAT lacking MCRIP2 have reduced expression of oxidative phosphorylation complexes, branched chain amino acid and fatty acid catabolism enzymes, as well as mitochondrial translation proteins. Further, the absence of adipose MCRIP2 diminishes thermogenesis in response to cold, and lipid oxidation in response to adrenergic activation of BAT. Mechanistic studies performed in brown adipocytes reveal that MCRIP2 regulates the transcription rate of some, but not all, of its target genes, indicating that it may act at the post-transcriptional level to control some targets. Altogether, this study demonstrates that MCRIP2 is a novel regulator important for maintaining the oxidative function necessary for BAT thermogenesis. Work from this study provides greater understanding of how BAT function is maintained, specifically downstream of ERR/PGC1-α transcriptional complexes and highlights the many levels at which BAT gene expression is regulated