136879 research outputs found
Sort by
Characterization of Ligand Binding Using Dissolution DNP Assisted NMR Spectroscopy
No full textBiomolecular interactions play essential roles in cellular processes including signaling, metabolism, and enzymatic synthesis of cellular components. Elucidating interactions between proteins and ligands using techniques such as nuclear magnetic resonance (NMR) spectroscopy provides fundamental insights into biological function, as well as guidance on the identification of new drug candidates. A significant NMR sensitivity improvement of several thousand-fold can be achieved by hyperpolarizing the ligand molecule using dissolution dynamic nuclear polarization (D-DNP). Spectra can be acquired in a reduced time, at or near physiological concentrations. Here, transverse (R2) NMR relaxometry is demonstrated to probe protein-ligand interactions. A 13C R2 relaxation dispersion measurement characterizes ligand binding epitopes through the observation of relaxation rates at different positions of the ligand, whereby the magnitude of the dispersion reflects the binding orientation of the ligand. The efficiency of the R2 measurement can be improved by an ultrafast approach to obtain the relaxation rates from all 13C spins in a single measurement. Numerous target proteins for pharmaceuticals are embedded in the cell membrane. Hyperpolarized 19F, due to its low NMR detection limit, is proposed for probing the interactions with membranes and the cell surface proteins. A model for the binding interaction combined with predictions of spin relaxation rates provides estimates of the binding affinity to membranes of different compositions in unilamellar vesicles. Applied to the measurement of ligand interactions with different cell types, the influence of the interactions between ligands and cell membrane proteins is identified
Advanced Autonomous Algorithms for Versatile Terrain Navigation and Multirobot Coverage Control
No full textThe primary objective of this research is to develop robust autonomous navigation and multirobot coverage control algorithms with broad applicability. To achieve this objective, this dissertation delves into three core topics: 1) terrain-aware path planning, 2) real-time stair detection, and 3) multi-robot coverage control. To address the challenges associated with autonomous navigation using vision-based terrain classification, a novel technique, the uncertainty rejection filter, is introduced. This filter, when combined with a neural networks-based terrain classification model, enhances the reliability of autonomous navigation by identifying uncertain regions and assigning appropriate traversal costs. Simulations and field tests demonstrate the effectiveness of this path-planning scheme. This dissertation also presents a real-time stair detection algorithm based on a decision boundary-aware model. Leveraging a support vector machine trained on RGB images, the algorithm outperforms existing models. Lastly, novel algorithms for multi-robot coverage control are proposed for both homogeneous and heterogeneous systems. A centralized approach incorporating agent dropout and reinsertion processes improves overall coverage, while a decentralized version achieves desired outcomes without a central computer. These algorithms exhibit improved coverage performance in diverse non-convex environments. Additionally, a user interface is introduced to enable users to define target areas for coverage by the proposed algorithms. Field experiments showcase the successful integration of the user interface with the coverage control algorithms
Enhancing Control Assessments and Risk Analysis in Batch Chemical Processes: A Focus on Safer Design Methodologies
No full textBatch and semi-batch processes involving critical chemical reactions pose significant risks to industrial safety and require rigorous control assessment and risk analysis. This study aims to survey current industry practices and methodologies for testing and assessing the controls in batch reactions. It identifies strengths and limitations in existing procedures and propose safer design methods to mitigate reactive hazards. In batch reactions, a diverse range of chemicals operates at dynamic parameters such as temperature and pressure. Assuming the worst case scenarios and improper evaluation of controls could lead to events with dire consequences. The study examines the typical tests conducted for batch reactions and identifies opportunities for optimization based on the outcomes. Layer of Protection Analysis is used to evaluate various preventive controls such as Inherently Safer Designs (ISDs), automated systems and relief sizing. Recognizing the pivotal role of critical controls such as quenching and relief sizing as the last line of safeguards in the batch processes, here we determine what are the different ways of performing quenching process to halt the reaction and the effective way of doing this operation. We also develop a model and calculate for effective quenching during worst-case scenarios, demonstrating how this proposed safer design method maximizes control functionality and minimizes risk. As relief is also considered in many cases where quenching is not practical, the relief sizing of reactive systems is quite complex and here we compare how drastic the change in relief size would be for a reactive system when compared to relief sizing performed in a conventional way. Another concern highlighted is relying solely on automated systems, which can lead to catastrophic consequences in case of process deviation. This research scrutinizes the conditions which are considered over conservatively by many of the process Industries, arriving at conclusion of adequacy in the Safety Systems and its impact on the availability of the Safety System on demand (PFD) using exSILentia tools. Overall, this study comprehensively explores industry practices in testing, standard controls and alternate safer designs along with conservative safety system settings
BATF3 and IRF4 Control iTreg Cell Fate Decisions
No full textFOXP3 is the lineage-defining transcription factor for Tregs, a cell type critical to immune tolerance, but the mechanisms that control FOXP3 expression in Tregs remain incompletely defined- particularly as it relates to signals downstream of TCR and CD28 signaling. In this dissertation, I studied the role of IRF4 and BATF3, two transcription factors upregulated upon T cell activation, to the conversion of conventional CD4+ T cells to FOXP3+ T cells (iTregs) in vitro. I found that BATF3 is a potent inhibitor of FOXP3 expression and iTreg differentiation but is dependent on interactions with IRF4 to mediate this inhibition. BATF3 allows IRF4 to bind an upstream regulatory region within the Foxp3 super enhancer. I further demonstrate that interactions of these transcription factors are necessary for glycolytic reprogramming of activated T cells that is antagonistic to FOXP3 expression and stability. However, I also show that BATF3 and IRF4 play critical roles to iTreg function, demonstrating unique roles for these transcription factors in FOXP3 induction vs. in differentiated iTregs. Thus, my findings highlight how BATF3/IRF4 interactions contribute to the complex interplay between TCR signaling, FOXP3 expression and stability, and iTreg function while providing important insights to cellular mechanisms that govern expression of the Foxp3 locus
Compositional Path Design for Graded Alloys Using Reinforcement Learning
No full textCompositionally Graded Alloys belongs to the category of Functionally Graded Materials (FGMs), distinguished by their varying spatial composition within the structure that results in material alloys of superior properties over traditional alloys. In the recent years, these compositionally graded alloys have gained significant recognition, primarily because of the advancements in the additive manufacturing techniques like Directed Energy Deposition (DED) which make the production of these alloys feasible. However, a linear gradient path of these alloys result in the inclusion of deleterious phases within the alloys micro-structure that can adversely affect the final alloy properties which may result in cracks & in the work done by Kirk et al [1] an innovative gradient path planning algorithm inspired by the state-of-the-art robotic route planning algorithms was adopted and a successful gradient path was designed in Fe-Ni-Cr material system which avoided the deleterious phases that was impacting alloy gradient when printed from 316L stainless steel to pure Cr. One significant drawback of this technique is that it limits the flexibility of material space exploration that could be done by the designer, any new composition exploration can only occur after re-configuring the path planner to compute the feasible gradient throughout the material domain and this limitation leaves designers with few alternatives. Q-Learning is a model-free Reinforcement Learning algorithm was used to solve this design space exploration problem in the ternary materials environment. An innovative method of encoding absolute position of the agent in barycentric coordinates along with the relative heading state to goal was formulated to model the system thus enabling design space exploration. The effectiveness of the proposed method was measured in a holdout dataset which produced a validation accuracy of 97%
Assessing Primate Eye Morphology
No full textEye morphology varies widely across primates with species exhibiting a range of eye colors and shapes, but we do not yet have a full understanding of what drives this diversity. We tested whether primate eye morphology is correlated with social factors, ecological factors, or both. We did so by examining the eye coloration and shape of 68 primate species (spanning great apes, lesser apes, prosimians, and African, Asian, and South American monkeys). We used modern methods of analyzing animal coloration that account for the specific visual system of each species. We found that primate eye coloration is correlated with both social and ecological factors. Primates with more discriminable gaze have larger social group sizes and canine size dimorphism. Primate eyes that have greater red chromaticity are associated with more open habitats. And, primates with more elongated eyes had larger body mass, lived in more open habitats, had larger group sizes, and lived closer to the equator. Our ancestral trait reconstruction suggested that the ancestral primate had moderately elongated eyes with brown irises and sclera. Our results suggest that both social and environmental factors have impacted the evolution of primate eye morphology
Self-Healing of Bulk Vitrimers as a Function of Thermomechanical Boundary Conditions
No full textThe subject of this study is vitrimers, a class of thermoset polymers with exchangeable covalent bonds, with a focus on their ability to improve material durability in the aerospace, civil infrastructure, and elastomer technology industries while requiring minimal intervention. The material used in this study was procured from ATSP Innovations, provided as a bulk block with the specified dimensions to meet our testing requirements. The selected material, a variant of ATSP characterized as CBAB, was chosen strategically owing to its unique availability in bulk form from the manufacturer at this juncture. This characteristic of CBAB was pivotal for our study as it is allowed for a consistent and uniform sample preparation, an essential factor for the integrity of our fracture mechanics analysis.
To evaluate the bulk fracture, compact tension (CT) tests were done on samples cut out of a panel. The samples were then mended by putting the two halves of the sample together and applying heat and mechanical pressures to study the kinetics of bond formation and their selfrepair mechanisms. Pre- and post-healing fracture toughness measurements were methodically performed, demonstrating a significant restoration of mechanical strength, confirming the effectiveness of the repair procedure. Furthermore, the study investigated the interplay between time and temperature influencing vitrimers ability to self-heal, offering light on the best circumstances for maximal healing efficiency. The time and temperature dependence of the healing mechanism was explained quantitatively by means of an Arrhenius relationship, and the activation energy of the bond reformation was assessed. Furthermore, by revealing the intricate link between stress levels and the healing process, this study sheds light on the fundamental mechanisms governing vitrimer self-repair, paving the path for improved material design and application. A comparison was made between the healing performance of bulk samples and ATSPcarbon fiber reinforced composites, obtained from other studies, to shed light on the effect of confinement (in the composite) on the healing performance of the vitrimers. Crucially, this study emphasizes the critical roles that temperature, time, and stress levels play in shaping the dynamics of self-healing, providing invaluable insights for both advancing our understanding of vitrimers and facilitating their wider adoption in the field of self-repairing materials
High-Dimensional Analysis of the Linear-Quadratic Regulator Problem Using First Order Methods on GPU
No full textThere has been a growing desire to bridge the gap between the fields of machine learning and optimal control theory. While optimal control typically operates on known dynamical systems, machine learning uses large data sets based on sampled data. The differences in input data used have made it difficult to adapt optimal control concepts to machine learning applications. For example, a major challenge in utilizing a linear-quadratic regulator (LQR) is how computationally demanding it can be for high-dimensional systems. Solving the algebraic Riccati equation (ARE)
directly is time-consuming and is typically O(n��) complexity. Posing the problem as a Linear Matrix Inequality (LMI) is even worse, this is typically solved in O(n���) time.
This thesis will examine the discrete-time LQR in the context of first order methods. These gradient-based methods provide an advantage in that they can be parallelized and run on GPUs. Multiple gradient-based algorithms will be proposed, and their performance will be compared with the traditional solutions for optimal LQR gains. The convergence rates of these gradients to the global optimum will be discussed and compared as the dimensionality of the problem increases. The ability to solve the LQR problem faster for high-dimensional systems may be useful for future large-scale optimal control problems in the aerospace field. Additionally, framing the LQR in terms of gradient-dominated policies may allow the LQR to be used in broader fields such as reinforcement learning
Nature of Science in Science Education: Establishing Common Ground
No full textA deep and robust understanding of the nature of science (NOS) has been increasingly recognized as a core component of any form of meaningful scientific literacy. However, little progress has been made with NOS teaching and learning over the past century. This has become an increasingly problematic situation as citizens have struggled with numerous high-profile socioscientific issues over the past several decades (e.g., climate change). Unfortunately, scholars��� discordant approaches to NOS in science education have created an acrimonious environment that has likely impeded efforts to promote attention to NOS teaching and learning. This problematic state of affairs is exacerbated by a lack of clarity regarding the specific ways that NOS understanding affects SSI decision-making, and a dearth of quantitative instruments that can provide valid and reliable inferences about respondents��� understanding of NOS constructs. This dissertation research contributes to the improvement of NOS teaching, learning, and research by: (1) Identifying the major themes that exist within the NOS approach literature, along with the areas of agreement and disagreement related to each theme; (2) Determining the NOS constructs related to climate change beliefs and attitudes; and (3) Creating a novel, high-resolution, quantitative NOS instrument that targets a specific NOS construct (i.e., methodological pluralism). The thematic analysis of NOS approach literature yielded seven central themes that contained numerous areas of agreement and disagreement that can help to provide order to future scholarly discussions. The examination of NOS connections to climate change beliefs and attitudes within the literature identified 15 relevant NOS constructs that provide an important focal point for science educators, communicators, and researchers. Instrument development efforts resulted in a tool that can provide high-resolution, valid, and reliable inferences about students��� views regarding methodological pluralism. While further refinement of the instrument is needed, this work represents a significant advance in quantitative NOS assessment. Collectively, the research related to this dissertation has addressed several important gaps in the field of NOS education, and establishes an important future line of research
Cricket Song Classification Using Transformers
No full textAs interest in studying animal sounds for biodiversity monitoring grows, the need for automatic methods to classify species based on their unique songs becomes crucial. This thesis presents an innovative approach to identifying various cricket species and genera by analyzing their audio recordings through advanced pretrained transformer models, specifically utilizing the AST (Audio-Spectrogram Transformer). The dataset includes 592 audio files for Gryllus species and 441 audio files for different genera and is meticulously curated to overcome challenges like uneven class distribution and variations in audio file durations. Customized label mapping and strategies such as undersampling and oversampling are applied to adapt the pretrained model to the specific classification task and balance the dataset respectively. The experimental setup includes three to four distinct datasets, each focusing on different subsets of cricket species and genera each. Training involves varying learning rates, with evaluation metrics encompassing validation and training accuracy, precision, recall, and F1-score. Further analysis is conducted to visualize the distribution of the data points. For this, first, Principal Component Analysis (PCA) is used to reduce the dimensionality of the features. Next, t-SNE visualization is used to provide insights into the spatial relationships between different species in the data space. This work differentiates itself from previous approaches that used CNNs, and it explores the capabilities of transformers in classifying cricket species and genus and aims to understand how these models perform in comparison. The transformer model achieved high accuracy rates of 95.31% for classifying Gryllus species and 94.27% for genus classification. This study has potential applications in education, conservation, agricultural pest management, and other ecological studies