22689 research outputs found
Sort by
A Needs Based Assessment to Determine how to Enhance On boarding Modules for Tenured Research Faculty
Get PDFThe purpose of the research study is to determine if the currently available onboarding modules for new faculty hires and the ongoing training for faculty is sufficient for understanding the internal process of Sponsored Research at Georgia Institute of Technology. This study is being conducted for a Capstone Project Final for Lea Brooks at Johns Hopkins University, Baltimore, MD. The information collected will be utilized to improve the resources available to faculty seeking sponsored funding at Georgia Institute of Technology
CHARACTERIZATION OF QRICH1 AS A NOVEL REGULATOR OF T CELL RECEPTOR SIGNALING
No full textPrecise regulation of T cell receptor (TCR) signaling is critical for defending against pathogens while preventing diseases such as autoimmunity, chronic inflammation, and lymphoma. The signaling hub and scaffold protein CARD11 controls T cell activation by linking TCR co-stimulation to activation of the NF-κB, JNK, mTOR, and Akt pathways. Dysregulation of CARD11 signaling is associated with diseases including primary immunodeficiency and lymphoma, and pharmacological modulation of CARD11 signaling components has emerged as a promising strategy for treating immunological disease and promoting anti-tumor immunity. Therefore, we sought to identify previously undescribed factors controlling the signaling output of CARD11 and discovered that QRICH1 is an intracellular checkpoint for CD8+ T cell activation that negatively regulates CARD11 signaling to NF-κB. QRICH1 is recruited to the activated CARD11 complex downstream of TCR co-stimulation, where it blocks subsequent signaling to the IKK complex and NF-κB, and this interaction is controlled by a previously undescribed inhibitory domain within the C-terminus of QRICH1. To study the physiological impact of QRICH1 deletion in the T cell lineage, we generated a novel QRICH1 conditional knockout mouse and found that QRICH1 controlled CD8+ T cell activation, proliferation, and effector function in response to ex vivo stimulation as well as in vivo infection with the bacterial pathogen Listeria monocytogenes. We also analyzed the QRICH1 domain requirements for inhibition of CARD11 signaling and found that most of the QRICH1 inhibitory activity was contained within the first 400 N-terminal amino acids, and deletion of the C-terminal inhibitory domain introduced additional redundancy in the domain requirements for inhibition of CARD11 signaling. Finally, we assessed the impact of QRICH1 on TCR signaling to NFAT and on TNF-⍺ receptor signaling to NF-κB and found that QRICH1 may slightly impact these signaling pathways. Our work defines a previously unappreciated regulator of T cell receptor signaling that modulates effector function
Probability Flow ODE in Infinite-Dimensional Space
No full textDiffusion models (DMs) have emerged as a cornerstone in the field of generative methods, with widespread applications in image creation, 3D data synthesis, video generation, and molecular design. Traditionally confined to finite Euclidean spaces, these models employ a forward process to introduce Gaussian noise and a reverse process that utilizes score-based functions for denoising and sample generation. However, the applicability of conventional diffusion models is limited when dealing with non-Euclidean geometric data such as point clouds, meshes, and graph-based structures.To overcome this limitation, recent advancements have expanded diffusion models into Hilbert spaces, significantly enhancing their ability to represent and manipulate infinite-dimensional measures. This extension is essential for capturing the complex nature of non-Euclidean data. This paper presents the derivation of the Probability Flow Ordinary Differential Equation (PFODE) within an infinite-dimensional framework. Our findings suggest that, regardless of generation speed and image quality, PFODE outperforms traditional score-based generative models in infinite-dimensional spaces. Moreover, we compare the mathematical frameworks of diffusion models in both finite and infinite dimensions, providing a mathematically accessible explanation for researchers interested in applying diffusion models to the generative modeling of non-Euclidean geometric data. This investigation not only promises to extend the performance and applicability of diffusion models but also aims to broaden their utility across various domains
Enabling Shape-Based Approaches for Autonomous Percutaneous Interventions with Sensorized Needles
No full textFlexible needle insertion procedures are common in minimally-invasive surgeries for disease diagnosis and treatment. Manual needle insertion has been proven to be challenging, requiring for frequent reinsertions of the needle to adjust needle positioning, causing excessive damage to the surrounding tissue, sensitive nearby anatomy and thus increased patient discomfort. To provide needle placement feedback to the physician, sensors are embedded into these needles for determining the real-time 3D shape of the needle during operation without requiring direct, intra-operative visualization of the needle. Through expansive research in fiber optics, a plethora of bio-compatible, MRI-compatible, optical shape-sensors have been developed to provide real-time shape feedback, namely fiber Bragg gratings (FBGs). Furthermore, many recent technologies to provide autonomous needle insertion control demonstrated promising results to accurately place the needle while reducing excessive tissue damage. As such, this work enables autonomous percutaneous interventions through intelligent shape-based approaches of sensorized needles through real-time shape sensing and shape prediction.
To enable autonomous prostate percutaneous interventions, this work provides intelligent shape-based approaches for real-time needle shape sensing and shape feedback. Needle shape sensing was achieved using a Lie-group theoretic model which demonstrated real-time, sub-millimeter accuracy, validated through extensive testing by robotic insertions into tissue. Incorporating stochasticity of needle insertion and uncertainty in FBG sensor measurements, a stochastic extension of the needle shape model was developed to provide needle shape confidence intervals. In an effort to optimize sensorized needle construction, robotic needle insertions into tissue were used to compare various FBG-based sensors for needle shape sensing.
For the first time, The proposed shape-sensing model enabled an autonomous robotic needle insertion closed-loop control method performed in in-vivo porcine experiments. A novel hybrid deep learning and model-based approach leveraging the presented Lie-group theoretic shape model was developed to predict the trajectory of the needle, achieving accurate, far horizon needle shape prediction
Air Reading: A Randomized Evaluation of a Virtual Tutoring Model
Get PDFThis report presents the findings of a randomized controlled trial evaluating the Air Reading program, a virtual tutoring model designed to improve reading skills among elementary students performing below grade level. The study, conducted in a rural district in Texas during the 2023-24 school year, involved 381 students from first to sixth grade. Students were randomly assigned to either the Air Reading treatment group or a control group receiving standard classroom instruction. Using the Northwest Evaluation Association's Measures of Academic Progress (NWEA MAP) as the primary outcome, the study found that students who received Air Reading tutoring demonstrated significant gains in reading achievement compared to their peers, with an effect size of +0.12, translating to approximately 1.5 additional months of learning. Subgroup analyses revealed no significant differences by race, ethnicity, or socioeconomic status, though first-grade students showed the largest gains. Dosage analysis indicated that students who received 40 or more tutoring sessions exhibited a larger effect size (+0.17), reflecting 2 months of additional learning. These findings suggest that virtual tutoring, when implemented consistently, can provide meaningful improvements in reading outcomes, especially when students receive sustained tutoring. The report discusses the implications for educational practice and policy, highlighting the importance of dosage and sustained engagement in virtual tutoring programs
Networking for Development: Interpersonal Relationships and Technical Cooperation in Cold War Taiwan and South Korea
No full textThis dissertation analyzes interpersonal relationships that facilitated development cooperation projects in Taiwan and South Korea during the 1950s. Drawing from government and non-government archival sources, it examines a diverse range of historical actors, such as private consultants, family members of expatriates and officials, and immigrant engineers, all of whom contributed to industrial infrastructure construction in East Asia. The involvement of these historical actors in development cooperation was not limited to a set of technical tasks. Rather, they built and utilized social skills to remedy the uncertainty and vagueness of developmental visions promoted by the state. By actively leveraging their personal identities, desires, and networks as part of professional interactions, these historical actors shaped the practice of decision-making for development cooperation.
This dissertation argues that only through these interpersonal relationships could development visions be materialized into industrial infrastructure. The interaction between different actors, involving social and emotional labor, trust-building, and communications, sustained the chaotic processes of decision-making. Furthermore, interpersonal relationships provide a window through which hidden networks of development cooperation during the Cold War become visible. These relationships reveal that the existing hierarchy of gender, race, and language shaped decision-making processes. Furthermore, overlooked connections among private actors surface in these interactions, allowing us to examine hidden relationships between Taiwan and South Korea in their industrial infrastructure building.
The first chapter explores the governance of development cooperation in Taiwan and South Korea in the early 1950s and shows that even when their experience was not interchangeable, there were attempts to transfer Taiwan’s experience to South Korea. The following three chapters analyze how various groups of actors engaged in development cooperation, leveraging distinct sets of identities and social skills. The last two chapters delve into the planning of cement and fertilizer plants to show that social criteria, in fact, shaped the seemingly technical decisions in development cooperation projects. Together, this dissertation shows that the practice of development cooperation became an entanglement of the technical and the social through interpersonal relationships
A TALE OF TWO ARTICULATIONS: EXAMINING DIVERGENT OUTCOMES OF DOMESTIC WORKERS’ MOVEMENTS IN INDIA AND SOUTH AFRICA
No full textOver the last few decades, domestic workers have mobilized globally to demand labor rights and recognition as workers. India and South Africa are two such countries where domestic workers have waged struggles for recognition for several decades. As national contexts, the two countries share an essential similarity: two democracies in the global south with strong civil societies. Furthermore, as an occupation, paid domestic work is characterized by intersecting inequalities in gender, class, caste (India), and race (South Africa) in both these contexts. Despite these similarities, it is puzzling that domestic workers in South Africa have managed to gain substantive recognition while their Indian counterparts continue to fight for recognition. Motivated by this puzzle, this monograph examines: What explains the divergent outcomes of domestic workers’ movements in different countries? Drawing from more than two years of ethnographic fieldwork, qualitative interviews, and archival research in India and South Africa, I develop a framework that seeks to explain domestic workers movements’ divergent outcomes across national contexts by highlighting intersectionality, articulation, and hegemony.
I argue that domestic workers' movements in these two countries show divergent outcomes for two reasons: first, how the structural intersectionality inherent in the occupation found expression in articulations of the movements, and; second, how those articulations have interacted with the hegemonic projects. In the context of India, despite the intersections of caste, class, and gender in the occupation, the movement's emphasis on only gender and class (while omitting caste) projects domestic workers as poor and helpless women who seek the state's patronage. In contrast, by combining all the relevant axes of inequalities into its articulation, the domestic workers' movement in South Africa has amplified the scope of its influence. I argue that the domestic workers’ movement in India characterizes Selective Articulation while their South African counterpart represents Holistic Articulation. Selective articulation accommodates hegemony, while holistic articulation challenges hegemony. The different forms of articulation and how they interact with the hegemony in their respective contexts explain why the domestic workers’ movements in these two countries had differing outcomes
CAN FORMULA ONE EXIST IN A NET ZERO WORLD?
Get PDFEvery year, over a hundred million people tune into Formula One (F1), arguably making it the most popular motorsport worldwide. While the world of Formula One continues to grow, tapping into new markets and expanding its fanbase, it is facing a significant sustainability challenge in the face of global warming. This Capstone project investigates how Formula One can tackle this challenge while maintaining the integrity of the sport. The results indicate that electricity consumption on and off the racetrack was the primary source of carbon dioxide emissions associated with the sport, representing 72% of total emissions. Additionally, the results indicate that the logistics associated with the annual Formula One World Championship were the second-most significant source of carbon emissions. Grouping races regionally could reduce these emissions by 62%, equivalent to an annual emission reduction of 15 million kg CO2, while maintaining the number of races and destinations of the 2023 racing calendar, representing a 14% decrease in total identified carbon dioxide emissions. Additionally, other proposed mitigation solutions that do not significantly impact the integrity of the sport include implementing Formula E technology to Formula One cars, introducing a regulation for race track owners and Formula One team factories to utilize 100% renewable energy, and finally, reducing the number of Formula One races in a single season. Even though mitigation pathways exist to significantly reduce the environmental impact of the sport while maintaining its integrity, the results suggest that Formula One cannot exist in a net-zero world using only currently available technologies
GENERATING A SINGLE CHASSIS INFLUENZA VIRUS LIBRARY ENCOMPASSING DIVERSE HEMAGGLUTININS WITH STRAIN-DISTINGUISHING MOLECULAR BARCODES FOR MULTIPLEXED PROFILING OF HUMORAL IMMUNITY
No full textInfluenza is a common respiratory infection which affects a myriad of people globally and causes thousands of deaths per year. Developing treatments and vaccines for influenza to protect susceptible populations is important, but complex and challenging. Vaccines can provide protection against influenza, but they have relatively low efficacy and may not provide long-term and broad protection. Immunity against influenza is affected by several factors, including environment, age, previous exposure, health, and a number of immunological parameters. This makes developing effective vaccines difficult, especially since influenza mutates rapidly, leading to the circulating strains to change quickly and potentially escape vaccine protection. As an initial step towards addressing these challenges, it is important to screen the vulnerable populations and understand their humoral immunity against influenza at a given time in a time efficient manner. Therefore, we developed a whole influenza virus library with single chassis barcoded-hemagglutinin (HA) viruses to profile anti-influenza neutralizing antibodies in a high-throughput manner. We discuss challenges and demonstrate the utility of generating this library with techniques to explore neutralization and binding of anti-influenza antibodies
MANIFESTATION OF AROMATICITY IN BORON BASED COMPOUNDS
No full textBorepins are seven-membered aromatic rings with 6 π-electrons and a central boron atom. These unusual compounds have long intrigued researchers because of their weak, non-benzenoid aromaticity. While a significant amount of research has been done on stabilizing these precarious compounds, little research has been done on applying the compounds for material applications. As such, this thesis can be divided into two main units: (1) the continual pursuit of finding novel ways of stabilizing and attenuating the aromaticity of these seven-membered rings, and (2) finding borepin-based material applications. Chapter I outlines the structural moieties that borepins possess (arene annulation, boron incorporation, and significant steric shielding) which inspired many of these arduous projects.
Chapters II and II are concerned with the effects of arene annulation on the aromaticity of the borepin ring. Chapter II describes the effects of pyridine annulation on the borepin ring and the use of pyridine—and its derivatives—as protecting groups for shielding borepins from oxidative/aqueous degradation. While pyridine annulation proved to be synthetically challenging, its use as a protecting group offers unique material properties. Chapter III presents the effects of bond order and biphenylene antiaromaticity on the local aromaticity of the borepin ring.
Digressing from the attenuation of aromaticity within the borepin core, the remaining chapters try to use structure-property relationships to predict material applications. Chapter IV spotlights the sterically hindered protecting group and attempts to utilize it for advantageous applications, notably aggregation induced emission. While the current synthetic strategy failed to result in the desired compound, alternate synthetic pathways are proposed. Chapter V focuses on the role of boron in organic dyes; notably, the most prolific dye (BODIPY) involves the formal destruction of one aromatic unit. In lieu of destroying aromaticity, this chapter focuses on preserving aromaticity via a borepin-like moiety. While overall fluorescence was relatively low, the architecture offers a novel way of understanding optoelectronic properties.
Finally, in Chapter VI, the boron incorporation is once again the focal point. Exploiting boron’s strong Lewis acidity, the field of frustrated Lewis pairs (FLP) was explored. While final experiments were not possible due to instrumentation limitations, the borepins synthesized in this chapter are promising avenues for FLP chemistry