49082 research outputs found
Sort by
Investigating the history of cancer and its impact on multiple myeloma treatment methodologies and the emergence of monoclonal antibody maintenance therapy: daratumumab (DARA) a systematic review of the literature
No full text2024This thesis paper addresses the general history, epidemiology and pathogenesis of cancer, specifically defining a prevalent and incurable subtype of cancer, multiple myeloma (MM), and its sequelae within the human body while outlining the evolution of various categories of therapy used for the treatment of cancer and its impact on our contemporary understanding of this disease. In summary, the intent of this composition aims to feature the recent emergences in treatment methodologies, explicitly investigating the utilization of monoclonal antibodies against multiple myeloma and highlighting a promising new drug known as daratumumab (DARA) and the science behind its unique anti-cancer mechanisms, both solo and in tandem with other known cancer drug therapies, and its latest achievements in clinical trials
Investigation of the molecular mechanisms in UBE3A-dependent autism spectrum disorder: insights into sexual dimorphism and synaptic defects
No full text2024Autism spectrum disorders (ASDs) are characterized by social, communication, and behavioral challenges. UBE3A is one of the most common ASD genes, and transgenic mice with UBE3A overexpression exhibit typical autistic behaviors. ASDs display a remarkable sex difference with a 4:1 male to female prevalence ratio; however, the underlying mechanism remains largely unknown. Using the UBE3A-overexpressing mouse model, we studied sex differences at behavioral, genetic, and molecular levels. We found that male mice with extra copies of Ube3a exhibited greater impairments in social interaction, repetitive self-grooming behavior, memory, and pain sensitivity, whereas female mice with extra Ube3a displayed greater olfactory defects. Social communication was impaired in both sexes, with males making more calls and females preferring more complex syllables. At the molecular level, androgen receptor (AR) levels were reduced in both sexes due to enhanced degradation mediated by UBE3A. However, AR reduction significantly dysregulated AR target genes only in male, not female, transgenic mice. Importantly, restoring AR expression effectively rescued male-biased alterations in the expression of AR target genes, social preference, grooming behavior, and memory in male mice with extra copies of Ube3a, without affecting females. These findings suggest that AR plays an essential role in mediating the sexually dimorphic changes in UBE3A-dependent ASD. Regarding neuronal communication, because AMPA receptors (AMPARs) mediate most of the excitatory synaptic transmission in the brain, and synaptic dysregulation is considered one of the primary cellular substrates in ASD pathology, we investigated the involvement of AMPARs in UBE3A-dependent ASD. We found that the abundance of the AMPAR GluA1 subunit was decreased in UBE3A-overexpressing mice. Consistent with this, we detected a reduction in AMPAR-mediated neuronal activity with UBE3A overexpression. Interestingly, we discovered that GluA1 mRNA was trapped in the nucleus of UBE3A-overexpressing neurons, leading to suppressed GluA1 protein synthesis. Further, we identified that SARNP, an mRNA nuclear export protein, was downregulated in UBE3A-overexpressing neurons, which is responsible for GluA1 mRNA nuclear retention. Importantly, restoring SARNP levels not only rescued GluA1 mRNA distribution and protein expression, but also normalized neuronal activity and autistic behaviors in mice overexpressing UBE3A. These findings indicate that SARNP plays a crucial role in the cellular and behavioral phenotypes of UBE3A ASD by regulating the nuclear mRNA trafficking and protein translation of a key AMPAR subunit
Journal of African Christian Biography: v. 1, no. 6: Separate cover file (A4 format), print-ready
No full textThe full issue of Journal of African Christian Biography: v. 1, no. 6 is available at: https://hdl.handle.net/2144/3566
Characterizing the role of osteoclasts in the unique marrow fibrosis phenotype of osteoclast-rich severe autosomal recessive osteopetrosis
No full text2024Osteopetrosis is a rare bone disorder where patients have increased bone density due to decreased number or dysfunction of osteoclasts. Previous research has identified a unique marrow fibrosis phenotype in osteoclast-rich forms of this disease, and we hypothesize that the presence of marrow fibrosis is due to the presence of dysfunctional osteoclasts. Nfatc1 knockout and Clcn7 knockout mice were used to characterize and quantify the marrow fibrosis phenotype via histologic analysis. These models were crossed with Rank knockout mice to investigate the osteoclast dependence of this fibrotic marrow phenotype. Treatment with anti-RANK ligand will be used to further investigate the osteoclast dependence of marrow fibrosis.
Marrow fibrosis is present and quantifiable in Clcn7 and Nfatc1 knockout mice, and we are able to restore hematopoietic marrow in Rank double knockout mouse models where dysfunctional osteoclasts are not present. Pilot studies appear to demonstrate restoration of hematopoeitc marrow elements in Clcn7 and Nfatc1 knockout mice following treatment with anti-RANK ligand.
This study demonstrates the marrow fibrosis is both quantifiable in Clcn7 and Nfatc1 knockout mice, and that hematopoetic marrow can be preserved through genetic deletion of osteoclasts. Hematopoietic marrow can also apparently be restored in Nfatc1 knockout mice through treatment with anti-RANK ligand. These findings provide an understanding of the role of dysfunctional osteoclasts in the marrow fibrotic phenotype of osteoclast-rich osteopetrosis, and lay the foundation for investigation of therapeutic targets
Evaluation of fibrosis and anti-fibrotic therapeutics in systemic sclerosis
No full text2025Systemic sclerosis (SSc) is a complex yet elusive connective tissue disease where vascular, immune, and fibrotic alterations occur. Nearly all organ systems are affected, leading to significant impacts on patients’ quality of life. A literature review was conducted to assess SSc pathogenesis and treatment options, focusing on diffuse cutaneous SSc. Vasculopathy and immune dysregulation culminate in the activation of fibroblasts into myofibroblasts. Fibrosis is the most lethal complication of SSc and is characterized by overaccumulation of extracellular matrix. Pathophysiology of skin fibrosis and lung fibrosis—comprising Idiopathic Pulmonary Fibrosis and Interstitial Lung Disease (ILD)—was examined. Fibrotic entities are generally conserved across phenotypes of fibrosis, yet some significant differences exist among the different patterns. Progressive ILD is the leading cause of death in SSc. There is an established association between decline in forced vital capacity and mortality. Though many current therapies target organ-related complications of SSc, treatments targeting fibrosis are limited. At present, there is no disease-modifying therapy for SSc. Therapies inhibiting fibrotic mechanisms were evaluated and were concluded to show promise for the treatment of SSc
Ground- and satellite-based observations of column nitrogen dioxide: instrument performance, column-to-surface relationships, and the role of meteorology in coastal urban environments
No full text2024Nitrogen dioxide (NO2) is a criteria air pollutant that is deleterious to human health and the environment, but characterizing its distribution is challenging. This challenge arises from its abundant and heterogeneous sources, short lifetime, and the limited spatial extent of surface monitoring networks. In lieu of comprehensive surface monitoring, space-based retrievals of NO2 abundance may address gaps in our understanding of its spatiotemporal variability. Space-based observations of NO2, however, have coarse-resolution sensors, requiring well-constrained inputs, and until recently have only collected one observation per day (at most), limiting their utility for characterizing diurnal variability or intra-urban heterogeneity. Throughout this dissertation, I constrain the precision of ground- and space-based remote sensing instruments dedicated to retrieving NO2 abundance, as well as explaining the spatiotemporal variability of NO2 to provide new insights relevant to urban air quality. Chapter 1 of this dissertation explains the motivation for this dissertation in more detail. In Chapter 2 of this dissertation, I quantify previously unexamined aspects of the diurnal precision of ground-based spectroscopic column NO2 observations using a high spatiotemporal resolution model of the 2013 DISCOVER-AQ campaign domain around the Houston, TX area. Pandora is a ground-based instrument commonly used to observe NO2 columns in the atmosphere. Networks of these instruments are distributed throughout the world, and their precision and accuracy make the instrument favorable for observing the spatiotemporal variability of NO2 and validating satellite instrument NO2 observations. Pandora-derived NO2 observations are often considered implicitly precise relative to satellite observations, thus motivating this evaluation. With this model I developed an instrument viewing “operator” to simulate the Pandora instrument’s operation. This operator creates synthetic direct-sun (DS) differential optical absorption spectroscopy (DOAS) columns which, when compared with modeled overhead columns, reveal that urban heterogeneity results in late-day (4-6 pm, LT) observations being less precise than previously estimated.
In Chapter 3 of this dissertation (Adams et al., 2023) long-term collocated surface and column NO2 observations at Boston University were used to understand drivers of total column NO2 variability in a coastal urban setting. I found that variations in column and surface NO2 abundance were governed by different processes. The temporal variability of NO2 column density was highly dependent upon meteorology, while concentrations of NO2 at the surface were more dependent upon surface emissions patterns and boundary layer entrainment. I found that the apparent equal mixing height of NO2 plumes within the boundary layer were not sensitive to prevailing meteorology or boundary layer stability. Additionally, I found that the sea breeze fostered uniquely large temporal variations in column NO2. I demonstrated that sea breeze conditions challenge the ability of satellite-derived column NO2 observations to accurately characterize day-to-day variation.
In Chapter 4 of this dissertation, I use long-term measurements of Pandora-derived total column NO2 at Boston University, Blue Hill Observatory (Milton, MA) and Harvard University. This long-term record confirmed that variation in temporal gradients in column NO2 observed in chapter 3 correspond to spatial gradients. Differences in column NO2 between sites as a function of time of day allowed us to infer the scale and formation of spatial column NO2 gradients. Finally, I evaluated to what extent satellite-derived column NO2 retrievals are capable of interpreting emissions differences across time and space. Generally, the TROPOMI satellite instrument overpasses struggled to characterize changes in column NO2 gradients across the Boston and Harvard University measurement locations between 2020 and 2021 relative to Pandora. However, TROPOMI resolved differences in the distributions of NO2 across urban-suburban scales that were not as obvious in the Pandora measurements. My results suggest that this difference in strengths at various scales is a result of the Pandora’s sensitivity to near-field emissions perturbations, in contrast with TROPOMI’s satellite footprint method which averages across larger-scales.
Chapter 5 of this dissertation summarizes the conclusions from Chapters 2, 3, and 4 and provides suggestions for future investigators
Chemically coalescing liquid metal emulsions for 3D printed soft conductors
No full text2025Gallium-based liquid metal alloys (GaLMAs) have widespread applications ranging from soft electronics, energy devices, and catalysis. GaLMAs can be transformed into liquid metal emulsions (LMEs), a composite form with modified rheology, for simpler patterning, processing, and material integration in GaLMA-based device fabrication. One major drawback of using LMEs is reduced electrical conductivity, owing to the oxides that form on the surface of dispersed liquid metal droplets. LMEs thus need to be activated by coalescing liquid metal droplets into an electrically conductive network, which usually involve techniques that subject the LME to harsh conditions. In this thesis, we present a way to coalesce these droplets through a chemical reaction at mild temperatures (T ~ 80°C). This chemical activation is enabled by inclusion of halide compounds that chemically etch the oxide on dispersed microdroplets of eutectic gallium indium (eGaIn). We investigate the use of a covalent halide compound as an activator and elucidate its activation mechanism. Through nuclear magnetic resonance spectroscopy, we discover the ability of eGaIn to catalyze the dehalogenation of our covalent halide activator and confirm through X-ray photoelectron spectroscopy that chemical oxide etching is occurring. Consequently, we establish the mechanism for self-catalyzing chemically coalescing LMEs. We then optimize this emulsion as a functional ink for 3D printing by exploring activator concentrations that maximize post-heat electrical conductivity, compatibility with direct ink writing, and post-activation shape retention. As a result, we select a 3D printable formulation with an electrical conductivity of 1.5 × 10^3 S/cm for further characterization and 3D print parameter optimization. We also explore LME formulations containing halide salt activators, and find that chemically coalescing LMEs can reach a high electrical conductivity (2.4 × 10^4 S/cm) close to that of bulk eGaIn, but at the expense of shorter shelf-life and poorer shape retention. Rheology of the selected covalent halide-based emulsion reveals that the LME is shear thinning and shear yielding. Additionally, it exhibits a high plateau modulus (1.0 × 10^5 Pa) and high yield stress (~2 kPa), thus requiring high pressure and high print velocities, which is desirable for rapid fabrication of GaLMA-based devices. To provide a parameter processing guide for our ink, we construct a print phase diagram describing extrusion pattern types across a normalized print velocity range from 0.45 to 1.35. We also show that our ink can span distances up to 3 mm in length, following a mathematical model for viscoelastic catenaries that predicts an elastic modulus in agreement with experiment. Finally, to demonstrate the utility of our shelf-stable chemically coalescing LME, we incorporate it as a conductive ink in the hybrid 3D printing of custom-designed battery-integrated light emitting diode arrays, demonstrating simpler fabrication of GaLMA-based applications. This technology pioneers a new class of LMEs, providing the material basis for designing future chemically coalescing LMEs and patterning soft metal catalyzed multifunctional materials.2026-02-07T00:00:00
Topics in sparse Bayesian machine learning
No full text2023This dissertation is devoted to addressing several challenging problems in machine learning via the Bayesian approach. One popular approach to Bayesian deep learning is to use Monte Carlo methods, such as Markov Chain Monte Carlo (MCMC), to approximate the posterior distribution. These methods generate a set of samples from the posterior, which can be used to quantify the uncertainty in the parameters and make probabilistic predictions. Bayesian methods in deep learning provide a framework for incorporating uncertainty into the learning process and can lead to more robust models with improved performance on unseen data. They have been applied to a wide range of problems, including image classification, reinforcement learning, and generative models, among others. This dissertation is organized as follows. First chapter is fast asynchronous sampler in sparse bayesian learning. In this chapter, We propose a very fast approximate Markov Chain Monte Carlo(MCMC) sampling framework that is applicable to a large class of sparse Bayesian inference problems, where the computational cost per iteration in several regression models is of order O(n(s + J)), where n is the sample size, s the underlying sparsity of the model, and J is the size of a randomly selected subset of regressors. This cost can be further reduced by data sub-sampling when stochastic gradient Langevin dynamics are employed. The algorithm is an extension of the asynchronous Gibbs sampler of Johnson et al. (2013), but can be viewed from a statistical perspective as a form of Bayesian iterated sure independent screening (Fan et al. (2009)). We show that in high-dimensional linear regression problems, the Markov chain generated by the proposed algorithm admits an invariant distribution that recovers correctly the main signal with high probability under some statistical assumptions. Furthermore we show that its mixing time is at most linear in the number of regressors. We illustrate the algorithm with several models. Second chapter is A one-step Laplace Approximation for high-dimensional variable selection. In this chapter, we introduce a rapid one-step Laplace approximation method, referred to as OLAP, which effectively tackles the computational burden of variable selection in high dimensions. Our findings demonstrate that this approximation offers a consistent variable selection procedure under reasonable assumptions. Additionally, we establish that the mixing time of the Gibbs sampler, employed for sampling from the posterior distribution of OLAP, scales linearly with the dimension p. Through comprehensive simulations, we validate the efficiency and accuracy of our proposed sampler, highlighting its potential to significantly enhance variable selection processes. Third chapter is Sparse(Cyclical) MCMC in Deep Neural Networks. In this chapter, we propose a general cyclical MCMC framework for a class of Bayesian inference problem, aiming to generate samples from one single mode in each cycle andhave mode swapping among different cycles to capture multimodality. We provide extensive results on the performance of prediction, multimodality of different cyclical MCMC methods on high-dimensional gaussian mixture models. We then introduce the sparse cyclical MCMC sampler in deep neural networks and present promising simulation results from the perspective of uncertainty estimation and calibration
Enhancing security of cryptographic protocols through better modularization
No full text2023Cryptographic protocols are usually designed using a modular approach, breaking down the task into several smaller tasks that involve constructing several cryptographic primitives. The next step is to identify primitives which serve the purpose and yet are simple enough to be instantiated. Simplicity can be defined in different ways - instantiable from a wider set of assumptions (or information-theoretically secure if possible), relaxed round complexity, and weaker security guarantees. However, finding the right way to break down the protocol into simpler primitives is not always easy. If the primitives are too simple, building the protocol becomes challenging. Conversely, if the primitives are too strong, it can be difficult to build them. Therefore, identifying the correct notion of "simplicity'', choosing the right set of primitives, and building the protocol requires a delicate balance. In this thesis, we revisit three problems in the areas of secure computation (MPC) and non-interactive zero-knowledge (NIZK) and we propose new ways to modularize the problems by relying on simpler primitives: 1. In the setting of two-party computation (2PC) where one of the participating parties is corrupted by a computationally unbounded adversary, we build a round-optimal 2PC protocol using a relaxed version of the oblivious transfer (which is not round-optimal) primitive and a non-interactive commitment scheme. 2. We build a round-optimal MPC protocol that withstands adaptive corruptions using an oblivious transfer which only provides sender and receiver privacy guarantees and satisfies some additional sampling properties. 3. We construct triply adaptive NIZKs which are secure against an adversary that corrupts parties and chooses statements to be proven in an adaptive way. We apply the Fiat-Shamir transform (instantiated using correlation intractable hash functions) to commitment-based Sigma protocols and demonstrate security in a modular way via a new approach for capturing universally-composable non-interactive commitments
Population genetics, environmental tolerances, and natural variation in infection frequency of the parasitic anemone, Edwardsiella lineata
No full text2024Mnemiopsis leidyi, the sea walnut, is one of the most destructive invasive species in the world. The ctenophore has broad environmental tolerances, a high rate of zooplankton consumption, extensive regeneration capabilities, and extraordinary fecundity. While there is a substantial body of research on M. leidyi’s role as a predator of zooplankton and as prey to the ctenophore Beroe ovata, only a few studies have examined M. leidyi’s relationship with its parasite, the lined sea anemone, Edwardsiella lineata. Previous research has shown that E. lineata may be exerting top-down control on the ctenophore in areas where present along the Northwestern Atlantic; the parasite has not established a population in M. leidyi’s invasive range. While a handful of studies have examined the ecological relationship between host and parasite, all studies have taken place at the same site, Woods Hole, and some failed to report key parameters such as host size and parasite number. More detailed examinations of this relationship are necessary to understand the parasite’s potential for controlling host populations in both native and European waters and how this control may be affected under various abiotic conditions. In my dissertation I performed three studies exploring the ecological relationship between M. leidyi and E. lineata. First, I profiled the variation in infection frequency in populations in Massachusetts and Rhode Island and examined the relationship between host size, parasite load, and seasonality. My results show that in natural populations, infected ctenophores are larger than non-infected conspecifics. The pattern of infection was also highly variable across sites and time. Second, I tested the temperature and salinity tolerances of E. lineata larvae to determine if certain conditions, such as those found in the invasive range of M. leidyi are refuges for the host from parasitism. My data indicate that temperature and salinity conditions alone will not prevent E. lineata from establishing a population in Europe. However, developmental outcomes and survival are both reduced at lower salinities and higher temperatures. Finally, using a RAD-seq approach I profiled population connectivity between juvenile parasite and adult polyp populations of E. lineata in Massachusetts and Rhode Island. E. lineata populations were found to be panmictic like host M. leidyi, though some structuring appeared between parasite and polyp populations. While asexual reproduction was seen in polyp mats, none was found within hosts.These data help contribute to the wider body of research on the M. leidyi-E. lineata system and also provide data on sea anemone genetic structure, which is currently understudied. I suggest that future studies examine cues for asexual reproduction in both the parasite and host and profile infection frequency and genetic structure in populations further south.2027-02-10T00:00:00