48440 research outputs found
Sort by
Characterization of A. annua Extracts: Investigating the Effects of Extract Composition on Breast Cancer Cells
No full textArtemisia annua (A. annua) is a herb native to the temperate regions of Asia where it was used as a traditional treatment for a variety of ailments for thousands of years. Recent studies have identified artemisinin, a compound derived from A. annua best known for its antiplasmodial properties, as a promising candidate for cancer treatment. This study aimed to investigate the chemical composition of various A. annua solutions and their effects on the viability of T47D breast cancer cells. The experimental treatments were prepared using several traditional methods, including two variations of wrung or soaked juices, two pounded extracts, and a tea infusion. From the collected data, it was determined that the pounded extracts had the greatest effect in reducing viable T47D breast cancer cell counts, followed by the 12-hour wrung juice, the tea, and the 2-hour wrung juice, relative to the water control. Thin layer chromatography (TLC) was used to analyze the chemical profiles of each preparation, revealing the presence of artemisinin in the wrung solutions and the tea. Artemisinin was not detected in either of the pounded solutions, suggesting either a concentration below the detection threshold or potential experimental inconsistencies that led to inconclusive results. Future studies should standardize extract concentrations across treatment groups to better isolate the effects of the preparation method, and test these preparations across multiple cancer cell lines to determine if the observed effects are consistent across various cancer types. Refinement of the TLC protocol is also recommended to reduce procedural variability, along with the integration of digital imaging software to enable quantitative spot analysis
Effect of cannabinoids (CBD) on worm behavior and physiology
No full textAlzheimer’s disease (AD) is associated with increased oxidative stress, known to be triggered by β-amyloid (Aβ) plaque accumulation. Increased reactive oxygen species (ROS) contribute to neuronal damage and reduced function. Cannabidiol (CBD), a non-psychoactive compound, has been shown to modulate oxidative stress by scavenging ROS and influencing antioxidant pathways. In Caenorhabditis elegans models of AD, CBD affects both redox balance and behavior. Preliminary observations by previous MQPs showed that stress induced worms with CBD treatment exhibited decreased motility compared with those without CBD. This suggests a complex interaction between CBD exposure and oxidative stress. To investigate this further, ROS levels will be quantified using a dihydroethidium (DHE) fluorescence assay. Expression of sod-1 and sod-2, genes encoding antioxidant enzymes, will be measured via RT-qPCR to assess molecular changes. Together, this data aims to clarify how CBD influences oxidative pathways and neuronal function in an AD context, offering insights into its potential therapeutic or adverse effects
The molecular basis of neuronal transition metal homeostasis
No full textZIP14, encoded by the SLC39A14 gene, is one member of the ZIP family of proteins and functions to increase cytosolic levels of zinc (Zn²⁺), iron (Fe²⁺), and manganese (Mn²⁺) in the intestines, liver, brain, and heart. These transition metals are essential for a variety of cellular processes including enzymatic activity, oxidative stress regulation, and metabolism. This report investigates ZIP14’s role in neuronal metal homeostasis. Disruption of ZIP14 function due to SLC39A14 mutations is associated with two rare disorders: hypermanganesemia with dystonia type 2 (HMDT2), caused by Mn²⁺ accumulation in the brain, and hyperostosis cranialis interna (HCI), a sclerosing craniofacial bone disorder. The experimental procedures performed included plasmid transformation, SH-SY5Y neuroblastoma cell culture, transfection with hZIP14, zinc uptake assays, BCA protein quantification, and atomic absorption spectrometry. Zinc uptake was analyzed as a function of time and varying concentrations to assess ZIP14’s transport function. Results support ZIP14’s role in regulating intracellular zinc levels and thus supports its single nucleotide polymorphisms (SNP) causing dysfunctions. Understanding ZIP14’s molecular transport mechanisms provide insight into its involvement in metal-related neurodegeneration and presents a future target for therapeutic intervention in metabolic and neurological disorders
Topic Detection in a Set of Documents using Large Language Models
No full textAs the amount of data collected for eDiscovery increases, traditional human-led document review has become increasingly unsustainable. This study explores the application of Large Language Models (LLMs) in enhancing the document review process of eDiscovery. Many LLMs, including Claude 3.7 Sonnet, GPT 4o, Gemini 2.0 Flash, Microsoft Copilot, Llama 3.2, DeepSeek-R1, Qwen2.5, Mistral, Phi4, and more, were tested for classification on both a legal and a non-legal dataset. Following this, various methods were tested and compared to improve accuracy, in addition to using both zero-shot and few-shot prompting. Testing revealed that processing techniques significantly impacted results such as individual processing performing twice as well as batch processing, while individual repeated binary classification (IRBC) tests consistently produced worse results than default pipelines. Additionally, hypergraph neural network (HGNN) implementation was used to fix enhance low-confidence predictions and increased the overall accuracy of the models by a range of 5-10 percent. The best performing model with an overall accuracy of 94 percent was Claude 3.7 Sonnet when HGNN was used. o3-mini had the second-best performance with 78 percent accuracy, followed by Qwen2.5 14B with 76.6% accuracy
Paying It Back: The Impact of Data Visualization on Loan Repayment Behavior
No full textAs student loan debt continues to rise, borrowers’ understanding of long-term impact of interest rates and repayment plans becomes increasingly important. This project applies behavioral economic theories to data visualizations to bridge a gap between the fields. By leveraging insights into human decisionmaking and cognitive biases, we aim to enhance understanding of student loan repayment and support more informed financial choices through data visualization. We designed a study comparing the student loan repayment decisions made by those who received a visual aid and those who did not. The results of this experiment contributed to finding more effective ways in which data visualization can be used to help individuals understand the distant and abstract process of student loan repayment. All supplemental materials can be found at (https://osf.io/b3tkf/?view only=c8b823f6dac64522b4acaf3f665724e0.) & For many borrowers with student loans, the process of repaying those loans can be confusing. The Office of Federal Student Aid (FSA) provides public, online resources to help guide borrowers through the repayment process. However, no one has investigated how effective these guides are in helping borrowers to understand and apply the advice they provide. In this MQP, I used genre criticism with a focus on digital rhetoric to identify the current rhetorical strategies that these resources use. I used my analysis to write a research manuscript examining and evaluating the genre of online FSA loan repayment resources. This report describes the process of researching related concepts in rhetorical analysis, conducting a case study on three FSA resources, and writing the subsequent manuscript
Urban Cooling Infrastructure Design Informed by Social Network Data
No full textRising temperatures and increased frequency of extreme heat events create a need for accessible, energy-efficient cooling infrastructure. This project seeks to improve cooling infrastructure in urban environments as informed by social network data. Social media networks serve as a valuable source of textual and spatial-temporal data that can highlight the connection between population preferences and geographic locations. In this project, we combine topic modeling on X and Reddit textual data with spatial analysis on Foursquare and Google Popular Times location data to identify patterns of public gathering, specifically during hot weather. Location data was cross-referenced with text and compiled into maps that identify gaps in Austin's current public cooling infrastructure to inform the selection of a site for a new public cooling center. We perform an exploratory data analysis of general locations within textual data to provide insights into major locations during high-temperature days, namely parks and cafés. This information was used to develop the program of our building. The building design’s performance criteria were then defined by the IBC and IECC codes as they apply to climate zone 2A. Nine case studies around Austin were analyzed to determine common passive cooling strategies, such as natural ventilation, shading, and thermal mass. These passive systems were incorporated into the building design alongside its active systems–the efficiency of which were analyzed through Building Information Modeling. Case studies and topic modeling showed that the most frequented public spaces prioritized cultural richness, flexibility, and connectivity, suggesting that social identity and comfort might play a role in drawing residents to cooling centers. Our final design incorporates popular opinion from social networks into site selection, building program, passive cooling methods, and overall design. By integrating environmental performance analysis with behavioral data and cultural contact, the proposed design framework supports a holistic, community-driven approach to urban heat resilience
Reframing Stress: A Mixed-Methods Study of the stART Program’s Impact on Youth Mental Health
No full textOur team has collaborated with the National Gallery Singapore and the National University of Singapore (NUS) to evaluate and enhance the stART program. The stART program is an art-based initiative aimed at improving the emotional literacy of adolescents in Singapore. Through three studies, we identified and evaluated a number of factors which we believe are paramount to expanding the stART program beyond its current workshop setting. An emphasis was placed upon the cost to the participant, artistic focus, technological implementation, and social application. These recommendations aim to improve the well-being of the Singaporean youth by providing more feasible and sustainable solutions for similar initiatives
The Design and Prototyping of a Low-Cost & Efficient Ocean Cleanup Robot
No full textOver the past year, approximately 16.5 million tons of plastic waste entered the world's oceans (Oceana). This waste poisons ecosystems and wildlife, and degrades into microplastics, entering the human food supply. Current strategies to address this issue are costly, require extensive human support, and focus mainly on preventing additional accumulation rather than removing existing debris. This project proposes the development of a cost-effective, scalable robotic system for efficient trash collection along shorelines. Our goal is to create an autonomous robot that matches the performance of existing solutions, with reduced cost and a limited need for human intervention. We designed and built a prototype robot, with a catamaran-inspired design. The robot has a durable trash collection system, two brushless motors, and a variety of sensors for object detection and navigation. While operating autonomously, the robot identifies and collects surface waste using a custom computer vision system optimized for aquatic environments
Unlocking heterogeneous photocatalysis in scalable μLED reactors
No full textOver the past decade, photocatalysis has emerged as an attractive method for synthesizing certain compounds and degrading others. Not only is photocatalysis “greener,” but it also allows for higher specificity than conventional catalysis methods and unlocks new pathways in chemical synthesis. However, photocatalytic reactor development has lagged behind in research, and there is still a void in the market for a scalable photocatalytic reactor that can be used in an industrial setting: most current reactor designs are limited in their applications due to light penetration issues and scalability concerns. Over the past few years, the Teixeira Lab has been developing a photocatalytic μLED packed bed reactor which shows immense potential for scalability and could potentially be the answer for industrial-scale photocatalytic chemistry. In this report, reaction kinetics and photonic efficiencies of this reactor were measured using a pharmaceutically relevant single-phase synthesis reaction to benchmark the reactor design. As part of these trials, the packed bed reactor was emulated by batch fixed-bed reactors, allowing for high modularity of the system for testing purposes. The potential of the system to facilitate heterogeneous chemistry was then explored by integrating a photocatalyst into the packed bed. The conversion of a compound by this photocatalyst was measured, and conclusions regarding the potential of the system to incorporate heterogeneous reactions in the future were drawn. Recommendations for future improvements to the design and potential experiments were made, suggesting that the reactor design exhibits high potential for becoming the preeminent industrial-scale flow photocatalytic reactor design
A Filtroporation Transfection Device to Modify Mammalian Cells
No full textTransfection is the process of introducing genetic material, such as DNA, RNA, or proteins, into mammalian cells, often for use in cell therapy or genetic engineering. One of the most well-known applications is CAR T-cell therapy, which uses lentiviral vectors to deliver therapeutic genes to a patient’s immune cells. However, this approach is extremely expensive, can trigger immune responses, and lacks repeatability. To address these limitations, this project focused on developing a non-viral transfection method using a custom-designed filtroporation system controlled by an Arduino board with custom-coded software, a stepper motor, and a linear actuator. Proof of concept was demonstrated using dextran sulfate and fluorescent imaging to confirm molecular delivery. Successful DNA transfection was achieved using red and green fluorescent protein-expressing plasmids, validating the device’s ability to introduce genetic material into mammalian cells without viral vectors