Senior Project Abstracts

Abstracts are sorted alphabetically by category. Jump to the start of a particular category by clicking a category link below.

Behavioral & Social Sciences
Biology, Microbiology, & Genetics
Chemistry & Biochemistry
Civil & Energy Engineering
Computer Science & Applied Computational Methods
Environmental Sciences & Engineering
Electrical & Computer Engineering
Biomedical Engineering & Health Technologies
Mechanical & Materials Engineering
Medicine & Health Sciences
Physics, Astronomy & Math
Plant Sciences

Evaluation Of The Table Rock Petroglyph

BEHAVIORAL & SOCIAL SCIENCES

This research project focused on examining a Native American petroglyph (rock carving) resting on the western shore of Utah Lake, known colloquially as Table Rock. Upon examination, Table Rock seemed to possess connections to astronomical phenomena of import, such as the solstices and equinoxes. If these connections did exist, they would appear on the petroglyph in the form of an alignment, when markings on the rock line up or point towards the astronomical phenomena in the sky. Contextual evidence and an alignment on the petroglyph with Polaris, places the construction of Table around 1400 CE and suggests the Fremont tribe as the most likely constructors. A phone camera was used to photograph the possible alignment, and a meta-analysis of past research was carried out to better understand the importance of Table Rock to the tribe that carved it.

Public Health Intervention: A Novel Hybrid Framework For Early Suicide Risk Detection And Prevention Using Real Time Linguistic And Socioeconomic Indicators

BEHAVIORAL & SOCIAL SCIENCES

Suicide remains a leading cause of death in the United States, yet most prevention systems rely on delayed mortality reports and static socioeconomic indicators that detect risk only after harm has occurred. Psychological distress, however, is dynamic and often precedes official signals. This raises a critical question: can real-time behavioral data identify suicide risk earlier than traditional surveillance?

I developed a novel hybrid early-warning framework that integrates linguistic distress signals from approximately 12,500 geolocated suicide-related social media posts with four state-level socioeconomic indicators across all 50 U.S. states. Using natural language processing and machine learning, I generated contemporaneous state-level distress metrics and fused them with structural measures of income, education, unemployment, and inequality. Rather than treating behavioral and socioeconomic data independently, this framework aligns and integrates both domains into a unified risk modeling architecture.

The hybrid model revealed distinct geographic suicide risk patterns and achieved substantially stronger predictive performance than single-domain approaches, identifying dynamic risk signals not visible in census data alone. This integration enabled earlier, prevention-focused insights capable of informing targeted public health responses.

By shifting suicide surveillance from retrospective reporting to real-time risk forecasting, this scalable framework equips state health departments and policymakers with forward-looking intelligence for crisis outreach, strategic resource allocation, and early intervention planning. Instead of documenting tragedy after it occurs, the system transforms fragmented data streams into an actionable early-warning infrastructure designed to help prevent loss of life.

School Uniform Perceptions And Violations

BEHAVIORAL & SOCIAL SCIENCES

School uniforms are a significant topic of study due to the differing and often conflicting conclusions that are made of the benefits of students wearing them. While some research finds that uniforms may improve behavior and respect (Baumann & Krskova 2016; Sanchez et al. 2012), others find that there is little to no difference (Ansari et al, 2021; Sowell 2012). Moreover, the way in which uniforms control student dress (Babicka-Wirkus 2023; JDCHS Parent & Student Handbook 2023) may lead to issues of limiting the right to freedom of expression (Ralston 2009; Vopat 2025). All the Catholic Schools in Utah require uniforms to be worn (Utah Catholic Schools 2025). The social significance of clothing, the issue of freedom of expression, the intended behavioral impact, as well as parent and student perceptions of uniforms explore the ways in which uniforms are debated. There is disapproval among some students, despite the potential benefits, so many students do not adhere to the policies. Ultimately, the project is based on the question: How do student perceptions of uniforms influence uniform violations? A survey sent to students and an analysis of a database of uniform violations allowed for both quantitative and qualitative data to be collected through a mixed methods approach. These data were analyzed using descriptive statistics, cross-tabulation, and a chi square test for the quantitative data, and qualitative data were coded based on common themes. All together, the data collected will be used to provide an answer to the research question.

How Ai Speaks Changes How People Think: Framing Ai Advice To Keep Moral Reasoning Human

BEHAVIORAL & SOCIAL SCIENCES

AI advisors have become common even for consequential decisions in health (triage, emergency care, and who receives a scarce organ), finance (whether to approve a loan and on what terms), social welfare (who receives limited support for food, rent, or shelter), and legal settings (bail and sentencing). Moral dilemmas cause worry, regret, and fear of errors (Krosch et al., 2012). AI advice can provide psychological relief and invite "diffusion of responsibility" and over-reliance on AI advice. If we treat AI as an objective moral crutch, we risk moral de-skilling and weaker independent ethical judgment.

I tested whether framing AI advice, using theoretically proven persuasion methods could result in people deciding similarly to how they would have decided if there had not been an advisor, since morality is an innately human trait. In a large study with more than 1500 participants I tested whether tone (emotional vs rational), certainty (uncertain vs confident), and structure (one-sided vs two-sided) changed agreement with AI advice and shifted choices in dilemmas that contrasted deontological reasoning (rule-based) with utilitarian reasoning (outcome-based). The dependent variables were choice, responsibility, deliberation time while a follow-up test examined the impact of interactions with the AI advisor. Most participants followed the AI advice. Two-sided advice and uncertain advice (55% confidence) produced choices closest to the no-advice baseline. This research holds implications for AI-agent designs in health, legal, welfare, and lending settings, especially where emotional and rational framing can increase over-reliance while uncertain or two-sided messages can keep decisions thoughtful and human.

Short Term Memory Accuracy

BEHAVIORAL & SOCIAL SCIENCES

This project investigates how scent, background noise, and visual distraction affect short-term memory accuracy. It is hypothesized that scent will have the greatest effect on memory accuracy due to its strong connection to the brain and memory processes. To test this, participants will complete an online 16-card matching memory game under four conditions: a control condition with no added stimulus, a scent condition using an air freshener, a background noise condition, and a visual distraction condition in which random videos play on the screen during the task. Each condition will be repeated multiple times to improve reliability.
The data collected will be used to compare memory accuracy, number of attempts, and task performance across conditions in order to determine which environmental factor has the greatest impact on short-term memory. It is expected that different sensory distractions will affect attention and recall in different ways.
This experiment aims to provide insight into how environmental factors influence short-term memory in everyday settings such as classrooms. Information for this project was obtained from sources including Google, Cleveland Clinic, ScienceDirect, and Harvard Medical School publications. Acknowledgements are given to all participants involved in the study.

An Ensemble Learning Framework For Frailty Classification In Older Adults Using A Dual Screening Approach

BEHAVIORAL & SOCIAL SCIENCES

Background: Frailty among older adults is highly prevalent and strongly associated with hospitalization, disability, and mortality. However, definitive frailty identification typically requires a time-intensive Comprehensive Geriatric Assessment (CGA) conducted by trained clinicians. Since nutritional vulnerability and functional impairment are central components of frailty pathophysiology, integrating brief nutritional and functional screening tools may enable earlier identification of clinically frail individuals. Yet, optimal computational solutions integrating these instruments while maintaining feasibility and interpretability remain undefined.

Methods: This study analyzed a synthetic dataset mimicking a population-based cohort of older adults completing two pragmatic screening instruments: Seniors in the Community: Risk Evaluation for Eating and Nutrition, version II (SCREEN II) and Strength, Ambulation, Rising from a chair, stair Climbing, and history of Falling (SARC-F). Leveraging demographics, SCREEN II, and SARC-F items as predictors, classification algorithms for CGA-defined frailty were developed using (1) rule-based screening, (2) logistic regression, (3) eXtreme Gradient Boosting (XGBoost), and (4) an ensemble-learning model. Model interpretability was examined using SHapley Additive exPlanations (SHAP) to identify key predictive features. An interactive Shiny web application was developed to support scalable implementation.

Results: The ensemble-learning model integrating SCREEN II and SARC-F outperformed individual screening tools and single-model approaches in identifying CGA-defined frailty. SHAP analyses demonstrated complementary contributions of demographic, dietary, and functional features.

Conclusions: A dual-screening, ensemble-learning framework integrating nutritional and functional screenings effectively improves identification of CGA-defined frailty among older adults. This study developed an end-to-end computational pipeline supporting interpretable risk classification and provided a scalable solution for frailty evaluation in aging societies.

The Impact Of Mouse Whisker Stimulation On C Fos Expressions In Neuronal Activity

BEHAVIORAL & SOCIAL SCIENCES

Mice rely heavily on their whiskers to detect and judge objects and navigate and map their environments. These primary sensory organs send signals to the brain, and this whisker-brain connections can be used for studying a mouse model of neuronal activation. However, it is unknown whether stimulating whiskers on one side of the snout increases c-Fos expression in a population of neurons. This study examines the impact of whisker stimulation in mice by measuring c-Fos expression in the barrel cortex. C-Fos is an immediate early gene (IEG) that is rapidly activated in neurons following strong stimulation, making it an accurate way to quantify neuronal activation. We stimulate whiskers on one side of the snout and compare images of the stimulated with unstimulated hemispheres to determine differences in activity in the barrel cortex. For these images, I used Immunohistochemistry to identify VGLUT and c-Fos. VGLUT staining identified the barrel cortex, and regions of interest were drawn around layers 2/3 and 4. Additionally, DAPI staining was used to label cell nuclei. The QuPath cell-analysis program was used to quantify the number of c-Fos-positive neurons and DAPI-labeled nuclei. I then calculated the ratio of c-Fos-positive neurons to total DAPI-labeled nuclei as my data. My results showed consistent increases of c-Fos expression in the stimulated layer compared to the unstimulated layer. These results suggest that whisker stimulation selectively increases c-Fos in the stimulated hemisphere. Moving forward, we know that this specific model of inducing neuronal activity can be used to study the barrel cortex.

Reliability Of Eeg To Identify Neural Signatures Of Cognitive Overload During Complex Task Performances

BEHAVIORAL & SOCIAL SCIENCES

Recent technological advancements have allowed the immersion of Brain-Controlled Flights, where an EEG can measure brain activity of an individual and translate those signals to a computer to allow for a plane to be flown hands off, solely based on what activity is happening in the brain. However, this system is unreliable as there has been limited research performed in this field, specially because of the lack of ability for a neuro-adaptive system to take full control over an aircraft and make fast paced decisions during an emergency crisis, where human judgement proves to be more efficient. This project measures if changes in brain wave activity as detected by an EEG can predict when cognitive overload has been reached and performance in quick decision making will start to decline. Participants were given a fast-paced Sustained Attention to Response Task (SART) to perform, as a way to replicate the level of cognitive awareness required by pilots while flying aircrafts, while connected to an EEG to track if the machine recognized repeated patterns in alpha, delta, theta, and beta waves activity during moments of critical decision making and split seconds before a wrong decision was made. The findings of this experiment can be used by researchers in the neuroergonomics field for further research, with the hope that brain controlled planes can become a more realistic, and safe mode of transportation in the near future.

The Power Of Pressure

BEHAVIORAL & SOCIAL SCIENCES

For my experiment, I tested the theory that people work better under pressure. To test this theory, I had five students complete a marble run twice with two different environments. One environment was high stress while the other was low stress. I then compared the two and found that students did in fact work better under pressure.

Scrutinizing Sound: Associations Between Phonemes And Game Provoked Emotions

BEHAVIORAL & SOCIAL SCIENCES

For years there has been research on the Bouba/kiki effect, or more broadly on sound symbolism, which refers to the concept of phonemes having inherent meaning and connotation outside of any specific language. Several previous studies have examined consonants and vowels and the apparently attached emotional responses. Using the results from these studies, a new pair of pseudowords could be designed, where one employs the combination of phonemes that should result in a maximally positive response, and the other a maximally negative response.
The pair of bisyllabic names I designed are each made up of CV syllables to ensure simplicity and minimize the risk of unnecessarily violating the phonotactics of any languages tested in the study. As an exercise in phonoaesthetics, these two pseudowords are theoretically opposites, and are significantly distinct from any widely-known words in the English language.
I am presenting these words in the context of a short video game, a uniquely immersive medium that could provide accurate results on emotional responses linked to the names of characters. Throughout the game, one non-player character will be portrayed positively and another non-player character negatively. At the end of the game, the player will be asked to assign both characters an appropriate name from a selection of pseudowords (aural stimuli), as well as rank the characteristics they felt the characters best embodied to avoid any confounding factors (e.g. one pseudoword connotes toward aggressiveness etc. instead of negativity). The results may corroborate the results found or warrant deeper study.

Humanized Vs. Dehumanized Opponents In First Person Shooter Games: How Do Different Designs Impact Adolescent Aggression?

BEHAVIORAL & SOCIAL SCIENCES

For multiple decades, scientists have studied how playing violent video games interacts with human behavior. Violent video games are games where the player intentionally inflicts harm upon other players or characters. Most often, these games are first-person shooters, where you play from a first-person perspective of the character inflicting harm. In the United States, concern over a connection between violent games and behavior has risen since the early 2000s, prompting more research in the field. Researching the broad relationship and searching for causation is out of this project’s scope in terms of complexity and resources. Instead, it will focus on how the character design of opponents impacts aggression towards those characters. In this study, aggression will be defined as the decision to react violently and the speed of that decision. Quickness to spare a character indicates the least aggression, and quickness to eliminate a character indicates the most aggression. Character designs differ by their humanity (“most human” vs. “least human”). This within-subject experiment asks participants to play a simple shoot/no shoot game to determine if there are differing reactions to different character designs.

Ph It Mediated M Kate2 Fluorescent Tagging Of The Unc 25 Gene In Caenorhabditis Elegans

BIOLOGY, MICROBIOLOGY, & GENETICS

Endogenous gene tagging enables the study of gene expression and protein localization while preserving native regulatory control. In Caenorhabditis elegans, many tagging methods rely on random integration, which can disrupt gene function or produce inconsistent expression. This project investigated whether PhiC31-mediated Integration of Transgenes (PhIT) could precisely fluorescently tag the unc-25 gene, which encodes glutamic acid decarboxylase (GAD), an essential enzyme for GABA synthesis in motor neurons.

A three-strain genetic crossing strategy inserted a red fluorescent reporter (mKate2) into the endogenous unc-25 locus using PhiC31 integrase. PhiC31 recombined attB sequences at unc-25 with attP sequences attached to mKate2 for targeted insertion. Histamine chloride selection eliminated worms carrying non-integrated DNA. Fluorescence microscopy and behavioral analysis assessed expression and gene function.

Resulting worms showed stable red fluorescence along the ventral nerve cord, consistent with known unc-25 expression. These worms displayed wild-type movement, showing that PhIT tagging did not disrupt gene function. Absence of non-integrated DNA further supported successful insertion.

These results demonstrate that PhIT enables efficient, precise endogenous tagging while preserving native expression. GAD and GABA are present in various species, including humans. GAD and GABA synthesis problems can cause neurological disorders such as epilepsy, anxiety disorders, and motor dysfunction. In addition, PhIT complements powerful gene-editing tools such as CRISPR-Cas9 by offering a reliable and site-specific integration strategy that can reduce experimental variability and increase efficiency. This fluorescent unc-25 strain provides a valuable platform for studying neural conditions and demonstrates the efficiency of the novel and cost-effective PhIT method for genetic research.

Evaluating The Necessity Of Bone Morphogenetic Protein Binding Endothelial Cell Precursor Derived Regulator (Bmper) In Modulating Beige Adipocyte Differentiation.

BIOLOGY, MICROBIOLOGY, & GENETICS

Obesity is a major public health burden, as 40% of adults in the United States are classified as obese. This results from chronic excess energy intake, leading to triglyceride storage in white adipose tissue (WAT). However, adipose tissue (AT) is metabolically heterogeneous, with distinct depots promoting either energy storage or expenditure. AT is broadly classified into WAT, which stores excess energy, and brown adipose tissue (BAT), which oxidizes fatty acids to generate heat through non-shivering thermogenesis. In addition, beige adipocytes resemble white adipocytes at baseline but acquire thermogenic properties in response to stimuli such as β-adrenergic agonists. Because beige adipocytes are capable of dissipating energy, understanding the molecular mechanisms may reveal new therapeutic strategies to combat obesity.

This project investigates BMPER protein, identified through single-cell and RNA sequencing as a potential regulator of beige adipocyte differentiation. BMPER modulates BMP signaling, which is critical for adipocyte fate determination; however, its role in beige differentiation remains unclear. To determine BMPER’s necessity for beige differentiation, the SVF from control mice and mice with BMPER deleted in Pdgfra⁺ adipose progenitor cells were induced to differentiate. Thermogenic markers (Ucp1, Cidea, Dio2, PPAR 𝛾, Leptin, Cpt1b) were quantified via qPCR to assess transcriptional programming. BMPER-deficient cells exhibited significantly reduced thermogenic expressions compared to control, with reductions in Ucp1 and Dio2, indicating impaired thermogenic programming. Control cells maintained robust activation of thermogenic programming. These findings established BMPER as a necessary regulator of beige adipocyte differentiation and suggest that modulation of BMPER-mediated signaling enhances thermogenic capacity in obesity.

Crop Specific Microbial Biofertilizer For Indian Ricegrass: A Sustainable Strategy To Supplement Alfalfa Cultivation

BIOLOGY, MICROBIOLOGY, & GENETICS

Over the past half-century, an increase in the use of fertilizer has led to environmental pollution as well as a long-term decline in agricultural yield. Biofertilizers are a potential alternative, but greenhouse-based experiments testing new combinations have tended to prioritize broad application rather than crop specificity. This study aims to explore a crop-specific microbial consortium to address the environmental stress associated with the production of alfalfa in the Intermountain West; specifically that when used as a widely grown feedstock, alfalfa is a water intensive crop. The consortium would combat this issue by improving the nutrient viability of an alternative a native and drought tolerant crop, Indian ricegrass, which has similar agricultural potentials to alfalfa while also being more suited to the arid climate of the Intermountain West. The goal of this work is to create a crop-specific microbial-based biofertilizer for Indian Ricegrass using a combination of algal, bacterial, and fungal sources. Individually each biofertilizer is expected to improve growth and biochemical parameters, such as dry weight, shoot length, and phenol levels, in comparison to the negative and positive controls. The combined consortium is anticipated to yield the greatest increase in growth and biochemical parameters, suggesting that the biofertilizers act synergistically, generating compound effects. Future research should focus on developing a framework for creating crop-specific synergistic biofertilizer combinations, both to identify high-productivity combinations for otherwise low efficiency crops as well as expand unexplored agricultural niches, such as in localized crop growing systems that increase food supply diversity.

Extracellular Vesicles From Activated T Cells As A Novel Early Diagnostic Approach For Type 1 Diabetes

BIOLOGY, MICROBIOLOGY, & GENETICS

Type-1 Diabetes is an autoimmune disorder common in children and adolescents. 17.2% of premature deaths are attributed to non-diagnosis soon after onset, but current markers used for diagnosis become detectable too late for practical treatment. Thus, there is a need for more thoroughly researched early diagnostics for T1D.
Autoimmune diseases like Type-1 diabetes are caused when the body’s cells are mistakenly recognized by white blood cells as foreign substances and destroyed. This response involves communication between T-cells and other cells, which is facilitated by extracellular vesicles (EVs).
Found in body fluids, EVs are excreted from cells and transport biomolecules. Because extracellular vesicles play a role in the immune response, EVs can indicate the activation of white blood cells, specifically killer T-cells, in autoimmune diseases.
Thus, the question asked in this research project was “What is the effect of activation by anti-CD3 antibody of a Jurkat T-cell line on the yield and size of extracellular vesicles from these cells?” It was predicted that if EVs are released from T-cells during the immune response, and Jurkat T-cells are activated and their EVs are isolated using size-exclusion chromatography and examined using nanoparticle tracking analysis, then the yield and size of the EVs will increase. The results were that the concentration of particles for the activated T-cell group was higher than the control, but the particle size did not differ between groups. The hypothesis was therefore partially supported. Further research for this experiment includes experimenting with actual T-cells from a patient with T1D.

Regulation Of Membrane Repair Pathways By Snare Proteins In Dmd

BIOLOGY, MICROBIOLOGY, & GENETICS

Duchenne’s Muscular Dystrophy (DMD) is an X-linked neuromuscular disorder caused by the loss of dystrophin, ultimately leading to progressive muscle damage. Repeated sarcolemmal damage places a high burden on vesicle-mediated membrane repair pathways. SNARE proteins, such as SNAP23, STX4, and VAMPs, play key roles in membrane repair; however, their regulation in DMD is not fully understood. This study investigates whether alterations in SNAP23 and the broader SNARE system contribute to impaired membrane repair in human DMD muscle. I conducted a systematic literature review following PRISMA guidelines and used a PICO framework to identify quantitative evidence linking SNARE proteins to membrane repair. In parallel, I reanalyzed publicly available human datasets to compare gene expression between DMD and healthy muscle. Differential expression analyses targeted SNAP23 and related repair proteins, while weighted gene co-expression network analysis (WGCNA) explored SNAP23-associated modules involved in vesicle trafficking and membrane repair. Gene set enrichment analysis (GSEA) further identified pathways impacted in DMD, focusing on membrane fusion, vesicle transport, exocytosis, and cytoskeleton organization. Together, these approaches aim to clarify the role of SNAP23 in DMD pathology and provide insight into SNARE-mediated membrane repair as a potential therapeutic target.

Effect Of Adding Probiotics To E.Coli Biofilm Bacteria

BIOLOGY, MICROBIOLOGY, & GENETICS

The first key study of this research comes from a study by Augusto Montiel-Castro et al., who reviewed evidence on the ability of the gut microbiota to communicate with the brain. They argue for a dynamic gut–brain-axis , involving neural and metabolic pathways, and further state that microbiota can influence brain function and behavior. While existing research has shown strong links between the gut microbiota, the gut-brain-axis, and depression, there is a significant gap in experimental studies that examine the effects of probiotics on neurotransmitter activity. The testing population of the study is Escherichia coli biofilm, which is one of multiple different microbial populations that influence the gut-brain axis, playing a significant role in bacterial balance in the gut. Also, the study will be using L. rhamnosus as the probiotic. The reason why I used this substance as a probiotic is that it has been tested on E. coli biofilm bacteria and has shown little to no inhibition of bacterial biofilm growth. I grew the E.coli biofilm bacteria in a petri dish for 24 hours and then transferred the bacteria to an electrode in order to measure the current produced by the neurological signals of the biofilm bacteria. I then added the probiotic to half of the samples so that I had a control group and an experimental group. Finally, I did a bell curve t-test to determine if there was a significant difference between the currents of the control and experimental group.

In Silico Identification Of Latent Tb Biomarkers For Crispr Cas13a Detection

BIOLOGY, MICROBIOLOGY, & GENETICS

Latent tuberculosis infection (LTBI) represents a significant reservoir for potential future active disease. However, current diagnostic methods rely on host immune responses and invasive sampling, which limit both accuracy and accessibility. Direct detection of RNA biomarkers from Mycobacterium tuberculosis presents a promising alternative, but this approach requires careful target selection and rigorous guide design to ensure specificity and assay robustness.
In this study, we present a comprehensive in silico framework that links latent and non-replicating M. tuberculosis RNA biomarkers to suitable CRISPR-Cas13a guide RNA candidates for amplification-based detection. We compiled a literature-supported shortlist of dormancy-associated biomarkers, including DosR/hypoxia-regulated genes and small RNAs, and evaluated their transcript conservation across the M. tuberculosis complex. Using BLASTN analysis, all prioritized targets showed complete nucleotide conservation across over 500 strains.
Cas13a guide RNAs were designed using the NY Genome Center platform and filtered according to stringent sequence metrics, such as a GC content of 40–60%, exclusion of homopolymers longer than 3 bp. Guides were using a conservative query coverage threshold (≤61%, corresponding to ≤13 matched bases per 23-nucleotide guide). This process resulted in a highly restricted shortlist of three high-confidence guides: crRNA068 (positions 68-90, % query cover ≤57), crRNA316 (positions 316-338), and crRNA317 (positions 317-339), which were selected as backup guides (% query cover ≤61).
Collectively, this work establishes a scalable blueprint for RNA-targeted CRISPR diagnostics.
In conclusion, this study establishes a rigorous in silico framework for prioritizing conserved RNA targets of latent Mycobacterium tuberculosis and for designing high-specificity Cas13a guide RNAs.

The Effect Of Solution Type On Osmosis In Decalcified Chicken Eggs

BIOLOGY, MICROBIOLOGY, & GENETICS

Osmosis and diffusion are important processes that happen in living things. Diffusion is the movement of particles from an area of high concentration to an area of low concentration. My experiment involves osmosis, a type of diffusion. Osmosis is described as the movement of a solvent, for example water, across a semipermeable membrane from an area of high concentration to an area of low concentration. Osmosis is a form of passive transport, this means it doesn’t require energy, and continues until the concentration on both sides are balanced.
This project investigated how different types of solutions affect osmosis in chicken eggs. Vinegar was used to dissolve the eggshells, exposing the semipermeable membrane. The initial mass of each egg was measured before being left in vinegar for 48 hours. After the shells were removed the eggs were placed in different solutions such as corn syrup, salt solution and water for 24 hours. The mass of each egg was measured again after the experiment to see how each solution affected the movement of water across the membrane. Changes in mass were analyzed to determine the effect of each solution on osmosis.

Influence Of Uv Light On The Susceptibility Of Lactobacillus To Garlic

BIOLOGY, MICROBIOLOGY, & GENETICS

This project investigated whether exposure to ultraviolet light could induce mutations in Lactobacillus reuteri bacteria that would alter their susceptibility to garlic, a natural antimicrobial agent. Lactobacillus cultures were exposed to controlled UV radiation for different durations to see how the exposure time influenced the results compared to the unexposed control group.

The Silent Treatment: How Goldenrod Reacts When Pollinators Disappear

BIOLOGY, MICROBIOLOGY, & GENETICS

Pollination is well known for its role in seed and fruit production, but it can also influence other plant traits. Recent studies have shown that pollination can affect both the amount and chemical composition of essential oils in aromatic plants. For example, studies on lavender found that pollination increased essential oil yield and altered oil chemistry (Jackson et al., 2025). Research on mint has also shown that essential oil composition can vary depending on biological and environmental factors (Nazem et al. 2019; Radev 2022).

Essential oils play important ecological roles in plants, such as attracting pollinators and deterring herbivores (Pichersky & Gershenzon, 2002). Because these compounds are involved in interactions with pollinators, changes in pollination may affect how much essential oil a plant produces.

Goldenrod (Solidago canadensis) is a native Utah plant that attracts a wide variety of pollinators, making it a useful system for studying pollination effects. Because pollination has been shown to influence essential oil production in other aromatic plants, this project tested whether pollination affects the yield and chemical composition of essential oils produced by goldenrod.

The Effect Of Different Stiffness Matrices On Melanoma Cells

BIOLOGY, MICROBIOLOGY, & GENETICS

Acral Melanoma is an aggressive form of skin cancer that appears on the palms and plantar surfaces of the feet, as well as under the nails. It commonly appears on pressure points such as the heel and the ball of the foot. Previously in the Holmen lab, DF-1 cells were delivered, producing RCAS-Cre virus to newborn mice to either behind the ear or to the footpad of DCT::TVA:Braf-CA/CA;Pten-flox/flox;Cdkn2a-flox/flox mice. Their studies showed that tumors developed in all mice, regardless of injection site. With mice injected in the footpad, brain and lung metastases developed 67% and 44%, respectively. In contrast, only 8% of mice injected behind the ear developed lung metastases. We hypothesize that stiffness influences metastasis development rates.
The extracellular matrix gives biochemical and mechanical information that regulates
melanoma cell behavior. In regions with high pressure, there is increased mechanical stress and
increased stiffness. These can activate mechanotransduction pathways, which increase the
aggressiveness of the cells.
To study this effect on metastatic potential, we created batches of polyacrylamide gels with two different stiffnesses to mimic a soft versus a stiff matrix, mimicking the different tumor microenvironments of cutaneous and acral melanoma. We believe that in a stiffer acral tumor environment, mechanotransduction via integrin/FAK signaling will be promoted, and there is a higher metastatic potential.

Un Folding Cancer: Evaluating Alpha Fold's Ability To Simulate Models That Predict Pathogenic Mutations

BIOLOGY, MICROBIOLOGY, & GENETICS

Alphafold, a protein-folding prediction tool created by Google DeepMind, has shown unprecedented ability to accurately simulate the 3D structure of proteins given an amino acid sequence. Based on this technology, tools such as AlphaMissense have been developed to predict the pathogenicity of single-nucleotide mutations through a combination of AlphaFold structural predictions and the output from a machine learning model trained on protein language. However, while tools such as these currently exist that evaluate pathogenic likelihood through the effect the mutation produces within the single protein, an evaluation has yet to be made that considers the protein structure within the larger context of the specific protein-protein interactions that the protein will be partaking in. This study addresses this gap to evaluate whether the correlation between pathogenicity and the quantitative increase in disorder within the Alphafold-simulated protein structure can be improved by simulating the entire protein-protein interaction rather than just the individual protein. To do so, this study uses the p53-p300/CBP protein pathway as a model to run experimental trials on. While one trial evaluates the effect of introducing mutations to p53 on the confidence rating of the p53 model, the other evaluates the effect produced by those same mutations on the entire simulated pathway. By comparing the correlation produced by each trial between the effect on the confidence rating and the clinically determined pathogenicity, the potential of this as a method to develop more targeted and accurate tools can be assessed.

Fame Biodiesel Fuel: A Biochemical And Thermodynamic View Into The World Of Renewable Energy.

CHEMISTRY & BIOCHEMISTRY

FAME (Fatty Acid Methyl Ester) fuel is a biobased, renewable fuel made through the transesterification reaction of oil and an alcohol. This project focuses on the biochemical reactions and reaction mechanisms used to synthesize FAME biofuel and the processes in determining its energy output using thermodynamics. Multiple types of FAME can be made using different methods, oils, and alcohols. These variations of biofuel are then combusted in an alcohol burner to determine it's energy output. This energy output, along with other values, are then compared to traditional diesel fuel to determine its viability as an everyday fuel.

Developing A Dishcloth Using Biowaste: An Environmentally Friendly Paper Towel Alternative

CHEMISTRY & BIOCHEMISTRY

Excessive use of paper towels contributes significantly to environmental pollution. Chemical treatments used in the manufacturing of paper towels hinder decomposition, causing them to persist for months to years, depending on humidity. The process for making paper towels involves bleaching harvested wood pulp with chlorine, which releases dioxins into wastewaters and can lead to soil contamination, underscoring the need for sustainable alternatives, such as biodegradable biofabrics. In this work, a symbiotic culture of bacteria and yeast (SCOBY) and coconut coir were used to create a cellulose pad with sponge-like properties that absorbs similarly to a paper towel. These specific materials were selected due to their viability as byproducts of the kombucha and coconut water industries, respectively. These derived byproducts are often thrown in landfills due to their limited applications, while emitting methane during decay. In addition to minimizing environmental impact, biofabric maintains high water absorbency due to the woven structure of cellulose nanofibrils (NFs), making it a more sustainable and efficient option for everyday cleaning needs compared to old rags, kitchen towels, and traditional paper towels. This novel biofabric is expected to function comparably to regular paper towels while being compostable, offering a practical and environmentally responsible alternative to single-use paper towels. The cloth will be evaluated for absorptivity, moisture retention, dry weight, tensile strength, abrasion resistance, density, and burst strength to assess its performance relative to paper towels. Future work will determine whether a binder, such as starch, is necessary during cultivation to improve wet- and dry-mechanical strength.

The Effect Of Different Common Salts On Electrolysis

CHEMISTRY & BIOCHEMISTRY

My research investigates the effect of different common salts—NaCl, KCl, and MgCl₂—on the stability and efficiency of electrolysis. Electrolysis is a process that uses electrical energy to drive non-spontaneous chemical reactions, and the type of electrolyte used can significantly influence both the rate of reaction and energy consumption. I hypothesized that KCl would provide the most efficient electrolysis due to its high ionic mobility and monovalent nature, which reduces disruptive interactions in solution. To test this, I prepared aqueous solutions of each salt at equal molarity and conducted electrolysis under controlled conditions. Ionic mobility and conductivity were analyzed to determine how each salt affects the flow of charge and the overall reaction rate. This study aims to identify how the choice of electrolyte influences electrolysis performance, providing insights that could improve efficiency and reliability in both industrial and laboratory applications. By understanding the relationship between salt type and electrolysis behavior, this research contributes to optimizing processes that rely on electrolytic reactions.

Machine Learning Optimization Of Conducting Polymer Thin Films With Magic Blue

CHEMISTRY & BIOCHEMISTRY

Conducting polymers are an emerging class of materials within the field of bioelectronics that have advantageous properties which offer uses in many applications such as: biological imaging, disease detectors, organic transistors, and more. Conducting polymers possess the unique ability to conduct both electricity and ionic flow when doped with an electrolyte solution, as well as an exceptional ability to interact with biological interfaces and directly to cells. One such conducting polymer, P3MEEET, has an exceptionally high ability to swell when in contact with water, and a high crystallinity, which makes it promising for uses within bioelectronics. However, the key issue with preparation of conjugated polymers (often prepared as thin films) is their volatility in performance to many different factors during preparation, such as temperature, dopant solution concentration, and more. In this project, I optimized the conductivity of P3MEEET thin films using a new cheap dopant known as "Magic Blue", which has shown promise in other similar polymers. The conductivity was optimized through an iterative process tweaking the preparation parameters with a machine learning algorithm until a maximum conductivity was reached. Because of the sensitivity of preparation, I utilized a robot-automated preparation technique to reliably produce thin films with greater precision. In the end, I obtained a maximum conductivity of 14.7 S/cm over 37 data points, which is greater than the conductivity of 12 S/cm that was obtained when doping with the previous dopant.

Combining Soybean Induced Carbonate Precipitation (Sicp) And Hydroxypropyl Methylcellulose (Hpmc) To Prevent Wind Erosion

CHEMISTRY & BIOCHEMISTRY

How can we combat wind erosion in semi-arid agricultural regions? Following the Dust Bowl—a series of wind events in the 1930s that destroyed thousands of acres of farmland—researchers have attempted to mitigate the impacts of soil degradation in the face of intense weather conditions. Through a myriad of infrastructural techniques, conservative-cropping methods, and chemical solutions, today’s scientists have turned to employing bio-cementing agents that solidify soil aggregates and protect against wind storms. Our research investigates how the combination of two bio-cementing solutions, soybean induced carbonate precipitation (SICP) and hydroxypropyl methylcellulose (HPMC), can improve soil stability while preserving root integrity.

Restoring Native Riparian Vegetation In Utah: Halophilic Rhizobacteria To Improve Black Cottonwood Establishment In Saline Soils

CHEMISTRY & BIOCHEMISTRY

Riparian forests play a key role in increasing the resilience of surrounding ecosystems to climate change by hosting a diverse range of endangered species, decreasing sediment erosion, and sequestering carbon from the atmosphere. However, riparian areas are especially susceptible to warming since they depend on reliable water access. In Utah, these conditions have favored Tamarix, an invasive tree species that is more resilient to drought and increased temperature than many native species. Tamarix invasions negatively affect riparian soil by using up water and shedding salt-laden foliage, resulting in dry, saline soil. This research examines how to encourage native plant growth in Tamarix-altered riparian soils. Root cuttings of Black Cottonwood, a prominent native species found in riparian areas that is highly sensitive to salinity, were inoculated with bacteria derived from the halophyte Allenrolfea occidentalis prior to planting. The bacteria were isolated from the rhizosphere and root endophytes of the halophilic species and grown on Petri dishes in a lab, then in a lysogen broth, a nutrient-rich liquid used to grow bacteria. The sample of bacteria was mixed in an Arabic gum slurry solution to promote the colonization of the bacteria onto the Black cottonwood. Finally, the roots were planted in sterile potting soil with a salinity level mimicking that of Tamarix-invaded soil and monitored during germination and early seedling growth. It is expected that the inoculation of the Black cottonwood root cuttings will promote plant growth and increase the germination rate, seedling survival, and total biomass of the saplings. Future research should focus on applying these findings in the field by inoculating Black cottonwood root cuttings with halophilic bacteria and planting them in riparian areas invaded by Tamarix and examining how the bacteria interact with the local microbial community.

The Azide Free Electrochemical Synthesis Of Aza Wittig Products

CHEMISTRY & BIOCHEMISTRY

The formation of an imine, or a carbon double bonded to a nitrogen, is of great
importance to medicinal and synthetic chemistry. Imines, commonly found in
heterocyclic compounds, are found in 70% of FDA approved drugs. Due to their
importance to chemistry, there has been investigation on how to synthesize imines on a
large scale. One such method to reliably and efficiently synthesize imines is via an aza-
Wittig reaction. The aza-Wittig reaction uses an iminophosphorane to form a carbon-
nitrogen double bond. However, the iminophosphorane material requires multiple
reactions to synthesize. Two steps in synthesizing the starting material are
the Sandmeyer and Staudinger reactions. The former forms an azide from an amine, the
latter forming an iminophosphorane from the azide. However, the azide functional
group is hazardous to work with, given its high shock sensitivity. Avoiding azides in
chemistry is a growing concern due to the danger working with them poses. A method
to synthesize the iminophosphorane without azides could be the answer to a safe and
efficient method to produce imine-containing heterocycles.

The Effect Of Fermentation Level On Lactose Derived Glucose Concentration In Dairy Products

CHEMISTRY & BIOCHEMISTRY

Many dairy products contain lactose, must be broken down by the enzyme before it can be digested. Fermentation reduces lactose content, which affect how much glucose is produced after lactase treatment. The purpose of this experiment was to investigate how the level of fermentation in dairy products affects glucose concentration after lactase treatment.
Four dairy products with different fermentation levels were tested: milk, cottage cheese, Greek yogurt, and Parmesan cheese. Two samples of each were prepared. One sample served as a control group without lactase, while the other served as an experimental group with lactase. The control samples were used to confirm that any detected glucose resulted from lactase activity rather than being naturally present in the dairy product. Each sample contained 5g of dairy mixed with 20mL of water. The experimental samples received 1/8 of a lactase tablet and were left for one hour. Glucose concentration was measured using glucose test strips.
The results showed that fermentation level strongly affects lactose content and the amount of glucose produced after lactase treatment. Parmesan cheese, which undergoes long-term fermentation, showed no detectable glucose, indicating that most of its lactose had already been broken down during fermentation. In contrast, lightly fermented products such as Greek yogurt and cottage cheese still contained significant amounts of lactose and produced measurable glucose after lactase treatment, nearly equal to whole milk. These findings suggest that fermentation influences glucose concentration, but short-term fermentation does not significantly reduce lactose levels in dairy products.

Squeeze Or Pour: Which Do You Perfer

CHEMISTRY & BIOCHEMISTRY

Vitamin C is an essential nutrient that plays an important role in maintaining a healthy immune system and supporting overall health. Since the body cannot produce vitamin C, it must be obtained from foods such as oranges and orange juice. This project investigates how the amount of orange juice affects the reaction between vitamin C and iodine. In this experiment, three clear cups will be prepared. Two of the samples will contain equal amounts of orange juice, while the third sample will contain a smaller amount. The same amount of cornstarch solution will be added to each cup. Iodine will then be added drop by drop to each sample until a dark color appears. Vitamin C reacts with iodine, and once all the vitamin C has been used up, the iodine reacts with the starch to produce the color change. The number of iodine drops required to cause the color change will be recorded for each sample. By comparing the number of drops needed, this experiment will show how the amount of orange juice, and therefore vitamin C, affects the reaction. Overall, this project demonstrates how chemical reactions can be used to measure nutrient content and how changing one variable can influence experimental results.

Promoting Lung Branching Morphogenesis Using Xyloside, A Small Molecule That Promotes Priming Of Endogenous Heparan Sulfate Glycosaminoglycans

CHEMISTRY & BIOCHEMISTRY

Embryonic morphogenesis is a complex yet highly organized process to develop each organ. Preterm birth interrupts the completion of this process, contributing to extremely low survival rates among infants born at less than 23 weeks of gestation—mainly because of immature lung development. However, there is currently no treatment that advances the process of lung branching. Thus, innovative interventions are urgently needed to promote lung growth in preterm babies.
Lung branching morphogenesis—the process of epithelial buds continuously diverging—is regulated by a signaling network of growth factors, which in turn are regulated by heparan sulfate. Yet, endogenous heparan sulfate is a limiting factor during early stages of lung development. Small-molecule xylosides are promising potential therapeutics for catalyzing enhanced lung growth, as they can easily diffuse into cells to manipulate glycosaminoglycan biosynthesis and increase heparan sulfate production. It is hypothesized that xyloside-primed heparan sulfate chains can bind growth factors and facilitate growth factor signaling in a manner similar to endogenous heparan sulfate, thereby enhancing lung growth.
To investigate this process, an embryonic mouse lung explant culture was utilized. Lungs were collected, cultured on membranes, and imaged to follow lung growth microscopically and analyze the number of peripheral lung buds. At low concentrations of xylosides, a statistically significant increase in the number of distal tips (the termini of the airway branches) was observed in comparison to control cultures. These findings demonstrate that xylosides promote lung growth in an embryonic mouse model, indicating the potential use of this approach for enhancing premature lung development.

Lithium In Salt Lake City Waterways: An Analysis Of Epidemiological And Extraction Potential

CHEMISTRY & BIOCHEMISTRY

Lithium is an antidepressant and mood-balancing drug for bipolar disorder. The concentration of lithium naturally in water is inversely correlated with suicide, homicide, and rape (Schrauzer, 1990 and Kohno, 2020). It’s used in batteries, especially in electric vehicles and phones, so mining lithium is increasingly relevant (Bhutada, 2023). I collected water samples from multiple sources in partnership with the SLC Water Reclamation Plant and tested the samples for lithium in an ICP mass spectrometer at the U of U. I compared our year’s data with data from the past 13 years collected by chemistry classes at my school, the Salt Lake Center for Science Education. My most interesting findings include that the lithium concentration in our tap water is lower than optimal for epidemiological benefits and that the lithium concentration in the runoff going to the GSL is showing an increasing trend, which could be useful for future mining.

Selective Oncolytic Replication Of Human Rhinoviruses In Mc Py V Associated Merkel Cell Carcinoma

CHEMISTRY & BIOCHEMISTRY

Merkel Cell Carcinoma is a rare and aggressive skin cancer and about 80% of cases are associated with Merkel Cell Polyomavirus (MCPyV). In cases where the virus is present, tumor growth occurs due to truncated viral T antigens (tT), which disrupt important tumor suppressor pathways, including p53 and RB, which normally help control abnormal cell growth. This allows mutated cells to keep growing and transforming into cancerous cells. In this project, we studied to see if two human rhinoviruses, HRVA2 and HRVA45, could target cells with MCPyV oncogenes while leaving normal epithelial cells unharmed. To replicate the viral gene expression found in patient tumors, we introduced individual MCPyV genes (ST, LT, tT, and S/LT) into semi-normal keratinocytes (HaCaT) and HEK293 cell lines which helped us assess how changes from MCPyV affect viral susceptibility. We then introduced HRVA2 and HRVA45, which we specifically developed for their cancer-targeting ability and compared their effectiveness to the well-known oncolytic Coxsackievirus A21 (CVA21). This targeted viral replication causes tumor cell death and releases tumor-associated antigens. This boosts immune response, enabling T- and B-cells to recognize and eliminate any remaining cancer cells. This shows how engineered or cancer-specific rhinoviruses like HRVA2 and HRVA45 could act as targeted treatments for virus-driven cancers like MCC.

Guanine Nitrogenous Base Dna Damage Due To Oxidative Stress Induced By Metals Inhaled From Great Salt Lake Dust

CHEMISTRY & BIOCHEMISTRY

Over the past few decades, the water level of the GSL (Great Salt Lake) has been decreasing at an alarming rate, due to a collection of factors like prolonged drought, climate-driven evaporation, and nearly 95% of the freshwater flowing into the GSL being diverted for agricultural and industrial purposes. As water levels continue to recede, the previously submerged lakebed is now exposed to the atmosphere. Because of the GSL's location at the bottom of a watershed, all sediments and heavy metals are transported downstream. Further, the Kennecott Copper Mine, one of the world's largest mining operations, is located directly upstream from the GSL, allowing trace metals like iron, copper, and arsenic to enter and accumulate in the exposed lakebed. When winds or storms occur, this dust is lifted into the atmosphere, having the potential to travel hundreds of miles and risk being inhaled. This iron-rich dust poses significant health risks if inhaled, since it serves as a catalyst for the
formation of reactive oxygen species (ROS), which have the potential of damaging DNA and creating health issues. The focus of this project is to analyze how DNA bases, such as guanine, are altered or damaged when GSL dust is inhaled, which provides a metal catalyst, such as iron, that induces the Fenton reaction in a bicarbonate setting. When induced, this Fenton Reaction produces a carbonate radical anion, that initiates damage through one-electron abstraction, damaging the DNA. The findings of this project highlight the serious potential health risks, such as lung cancer, associated with inhaling GSL dust.

Ph Levels Of Different Liquids.

CHEMISTRY & BIOCHEMISTRY

The project is about Ph Levels Of Different Liquids. The project describes how acidic or basic drinks are scaling from 0 to 14, where lower numbers mean stronger acidity , higher numbers mean stronger numbers mean stronger alkalinity, and 7 is neutral . Common liquids span a wide range on this scale: battery acid is extremely acidic with a Ph close to 0 , while stomach acid is also very acidic at round PH 1 to 2. The research Question is Which liquids are acidic , neutral , or basic based PH levels. The Hypothesis is i vinegar will have low Ph and apple juice will have high PH. The procedure is first i will get 5 cups and i will get juice coke and apple juice and then i will buy ph strips atleast 10 and i will dip strips into them to see if its acidic or basic or neutral and i will have my notepad to record the data

Investigating Structure Activity Relationships Of Synthetic Cone Snail Peptide Analogs For N A Ch R Inhibition

CHEMISTRY & BIOCHEMISTRY

Modern medicine faces growing limitations due to drug resistance, side effects, and the addictive nature of many current pain treatments. This creates an urgent need for safer and more targeted therapies. Marine organisms, particularly cone snails, produce a wide variety of bioactive peptides that interact with highly specific biological targets and represent a promising source of novel therapeutics. This study investigates the structure–activity relationships of synthetic analogs of the cone snail peptide RgIA, an α-conotoxin known to inhibit α9α10 nicotinic acetylcholine receptors (nAChRs), which play a key role in pain signaling.

Guided by prior structural and functional studies, including work from the Schmidt lab, simplified cyclic RgIA analogs were designed with targeted amino acid substitutions, specifically involving arginine and citrulline residues, to assess how subtle changes in peptide structure influence receptor inhibition. Peptides were synthesized using solid-phase peptide synthesis, followed by Alloc deprotection, on-resin cyclization, cleavage, and purification via high-performance liquid chromatography. Molecular identities were confirmed using mass spectrometry, and purified peptides were prepared for biological testing.

By systematically modifying peptide sequences and analyzing their chemical and biological properties, this project aims to clarify which structural features are critical for nAChR inhibition. Understanding these relationships provides insight into how cone snail peptides modulate neural pathways and supports the rational design of more selective, potent, and non-addictive pain-relief therapies derived from marine peptide scaffolds.

Utilizing A Monte Carlo Particle Transport Simulation To Evaluate Lithium‑Based Materials For Enhanced Tritium Breeding, Shielding, And Energy Deposition Performance In Nuclear Fusion Reactor Systems.

CIVIL & ENERGY ENGINEERING

Nuclear fusion has the potential to sustainably meet the world’s energy demands with minimal impact. However, the efficiency of the breeding blankets still remains a challenge. In this experiment, the focus is to find out what material makes the best lithium blanket, using Li4SiO4, Li2TiO3, Li17Pb83, Li2BeF4, and LiF-NaF-KF. Measuring the tritium breeding ratio, energy multiplication, leakage fraction, heating density, and specific tritium breeding ratio, the initial hypothesis was that either Li4SiO4 or Li2BeF4 would be the most efficient overall (Li4SiO4 having a higher specific TBR, whereas Li2BeF4 having a higher heating density). Using OpenMC, a Monte Carlo simulation created by the CRPG from MIT, to model and measure the qualities of millions of neutrons inside of a nuclear reactor, with all of the materials, the findings were clear: the material with the highest TBR is Li17Pb83 (1.54729), the material with the highest energy multiplication is Li2BeF4 (0.78440), the material with the lowest leakage fraction is Li2TiO3 (0.007), the most heat-density effective material is Li2BeF4 (0.47764), and the most specific TBR friendly material is Li2BeF4 (3.1279e-5). Overall, it is evident that Li2BeF4 takes the lead for many of these, proving itself as one of the best materials for a lithium blanket (partially supporting the hypothesis). Engineers who choose the material based on how it aligns with the design are more likely to effectively breed tritium, protect the poloidal, toroidal, and correction coils from neutron bombardment and heat, absorb byproducts of the fusion reaction, and regulate temperature.

An Early Warning System For Quantum Algorithm Failure: Detecting And Classifying Decoherence In Grover Search Using Distribution Based Metrics

COMPUTER SCIENCE & APPLIED COMPUTATIONAL METHODS

Quantum algorithms can provide major advantages over classical algorithms, but current quantum computers are susceptible to corruption, decoherence, and gate noise due to physical factors. These errors distort quantum states long before failure is observable through standard metrics like accuracy of the final answer. This project develops a distribution based early detection algorithm for an unstructured search quantum algorithm, Grover's algorithm. Grover's algorithm was implemented in Q# with 8 qubits, searching for a single state out of 2^8 = 256 states. Different types of noise were simulated through controlled random phase errors being added to the circuit during the iterations, modelling dephasing and decoherence, along with additional interference models such as amplitude disturbances and mixed noise conditions. Repeated trials were performed to estimate the algorithm success probability and the measurement distribution. Instead of relying on success probability, multiple distribution based metrics were used, including KL-Divergence, JS-Divergence, and L1 distance, computed directly from measurement results at each noise level, in real time. Results show that these metrics rise significantly earlier than accuracy degradation, and different noise models produce different distributions, enabling classification of interference type. This experiment demonstrates that real-time distribution monitoring can provide early detection of decoherence and noise before Grover's algorithm visibly fails. Early warning systems could support real time error-mitigation in fields like cryptography, optimization, and medicine discovery.

Early Detection Of Skin Cutaneous Melanoma With Artificial Intelligence Techniques

COMPUTER SCIENCE & APPLIED COMPUTATIONAL METHODS

Skin Cutaneous Melanoma (SKCM), arising from pigment-producing melanocytes, is the most prevalent melanoma subtype and the fifth most frequently diagnosed cancer in the United States. Given the rising incidence rates, developing methods for early detection is critical for improving patient prognosis and survival. However, the high dimensionality of genomic datasets—characterized by thousands of genes—presents a significant computational challenge for accurate artificial intelligence (AI) modeling.
This project proposes a novel diagnostic framework designed to identify SKCM earlier, faster, and more cost-effectively by leveraging RNA-seq transcriptomic data. Central to this approach is a feature selection methodology utilizing Kullback-Leibler (KL) Divergence. Based on the KL divergence-based feature selection, three distinct machine learning architectures were implemented and compared: Support Vector Machines (SVM), Deep Neural Networks (DNN), and Random Forest Classifiers (RDF). Empirical results demonstrate that this method is highly robust; all three models achieved an Area Under the Receiver Operating Characteristic Curve (AUC) of approximately 0.99. These findings suggest that integrating KL divergence with AI models provides a highly accurate and scalable solution for the early diagnosis of SKCM, potentially streamlining clinical workflows and reducing diagnostic costs.

Statistical Analysis Of Osimertinib Usage In Nsclc

COMPUTER SCIENCE & APPLIED COMPUTATIONAL METHODS

Cancer is a group of diseases characterized by uncontrolled cellular growth and spread to tissues, interrupting bodily function. Specifically, lung cancer is one of the most common forms of cancer, and affects millions of people each year. Lung cancer treatments are separated between small cell lung cancer (SCLC), and non small cell lung cancer (NSCLC), which makes up 87% of all lung cancer cases.
NSCLC treatments primarily consist of chemotherapy, radiation therapy, immunotherapy, and targeted therapies. Targeted therapies, relatively new treatment methods, target specific genetic mutations to kill cancer cells. They generally regarded as the standard of care for patients with targetable mutations, because they target the cancer, killing them off without as many side effects. Of targetable mutations, EGFR is the most common. When targeting EGFR, osimertinib is considered one of the most effective agents. However, many patients that are eligible for this drug may not receive it, due to treatment disparities related to age, race, socioeconomic status (SES), or other factors. To ensure that patients receive the recommended therapies, minimizing these disparities is vital.
This project uses R programming to isolate groups from a dataset that are eligible for the drug, and if they received the drug or not. Then, statistical tests such as Student's t-test and a chi-square test were run to determine the statistical significance of these categories.
Of the tested variables, SES, race, and sex were related to treatment disparities, and steps must be taken to improve equity in medical care in these areas.

Using 3 D Models To Predict Alzheimer's In Adults: Brains Of Change

COMPUTER SCIENCE & APPLIED COMPUTATIONAL METHODS

The objective of this engineering project was to create a fully animated model of the brain undergoing the stages of Alzheimer's disease with activity gains. This project will help in many aspects of medical research and diagnoses, educating others on the versatile effects of Alzheimer's disease on the brain as well. Alzheimer's is a neurodegenerative disease that disrupts communication between the brain's neurons due to tau proteins and beta-amyloid plaques. Activity levels in the brain heavily decrease as the stages of Alzheimer's progresses. Decreases in activity can lead to severe dementia, changes in personality, or a decline in mental abilities. With the creation of this 3-D model of Alzheimer's disease slowly encroaching on an adult brain, doctors can identify and diagnose Alzheimer's in its early stages. This way, treatments to clear the blockage of tau tangles and beta-amyloid plaques can be administered for slowing the process. Since this model was derived from PET scans, the model can be compared to real PET scans to diagnose patients with risk of Alzheimer’s. Faster detection can save many years on a patient’s life and potentially make way for scientists to find more treatment methods. The result of this project was a 3-D brain shifting and undergoing several changes in activity in only ten seconds. The Alzheimer’s Association believes life expectancy after the first diagnoses will soon change with the first survivor, and many are optimistic about their opportunities to cure or prevent Alzheimer’s disease.

Optimizing Epilepsy Neurostimulation Therapies Through Analysis Of Electrode Distance And Regional Neural Firing Dynamics

COMPUTER SCIENCE & APPLIED COMPUTATIONAL METHODS

Electrical stimulation is widely used to study and modulate brain activity, yet how stimulation influences large-scale neural population dynamics is not fully understood. In this project, micro-electrode intracranial EEG (iEEG) data were analyzed to investigate how stimulation in different brain regions affects coordinated patterns of neural firing across distributed neural populations. Peri-stimulus firing-rate series were organized into structured matrices of neuron × time, with grouping by brain region and by distance between the stimulation site and the recording site. Principal component analysis (PCA) and other methods were applied to these matrices to extract dominant population-level activity from high-dimensional firing-rate data. The results reveal low-dimensional population dynamics that vary across brain regions and with increasing distance from the stimulation site, indicating that stimulation produces organized and brain-wide changes rather than purely local effects. The differences found in the principle component analyses and trajectories put in perspective how distance and brain region affect how neural populations respond to stimulation. This can be applied to neurostimulation therapies in epileptic patients by indicating the frequency and intensity of stimulation necessary to affect the abnormal neurons.

Employing The Synthesis Machine Learning Algorithm To Synthesize New Novel Crystal Structures

COMPUTER SCIENCE & APPLIED COMPUTATIONAL METHODS

Over 90% of solid materials used in modern technology (silicon chips, lithium-ion batteries, smartphones, etc.) are crystalline, and differences in crystal structures can change a material’s strength, conductivity, or chemical behavior. Because the energy sector accounts for about 4/5 of global greenhouse gas emissions, discovering and synthetically creating new crystal structures could both improve energy storage and reduce environmental damage at a large scale. However, research costs and time of development prohibit efficient discovery of new crystals. To generate crystal structures in an affordable and efficient manner, this project presents Synthesis: a state-of-the-art algorithm for the synthesis of new, novel crystal structures never previously imagined by scientists. First, the widely researched Wasserstein GAN’s architecture was improved by employing both the M3GNET and CHGNET formation energy calculator in a novel cost function. Afterwards, Synthesis was improved using an attention-based network, large language model, StyleGAN and diffusion-enhanced generative adversarial network, after which the algorithm successfully produced 1000+ semi-stable crystals. To prove the viability of Synthesis, one crystal was made in a polymer-fusion technique in the lab. While more tests still have to be done to prove the viability of these materials, a computational algorithm creating new, novel, completely hypothetical crystal structures that are stable enough to be synthesized without the environmental damage associated with discovering crystals currently is a major step in reducing greenhouse gas emissions and combatting the world’s most pressing energy issues.

Understanding The Feasibility Of Streaming Black Box Data To The Cloud

COMPUTER SCIENCE & APPLIED COMPUTATIONAL METHODS

Every commercial passenger aircraft carries two black boxes to store aircraft information, such as its altitude, pitch, and speed, and audio recordings from inside the cockpit. In the event of a plane accident, this data is extremely crucial to investigators in determining how and why the plane went down, and how to take steps to prevent it from happening again. This is a large reason why air travel is so safe today. However, the data stored by black boxes can be lost if the black box itself is damaged during the crash, the aircraft loses power before crashing, is simply never recovered, or many other factors. The purpose of this project is to simulate live-streaming the data of black boxes to a separate device in order to evaluate and determine the effectiveness for real-world applications in the aviation sector. To help determine the feasibility, data was run through a test while being subjected to various factors such as latency, low signal, and distance from a receiving device. These were conducted in order simulate the actual factors an aircraft would experience, which could interfere with the transfer of data. The results showed that while possible, the act of streaming the data itself will cause some data loss, even in ideal circumstances. Further environmental factors would also further limit the amount of data transmitted before the data transmitter stopped function. However, this data would be extremely crucial to investigators in the scenario both black boxes are unrecoverable. Even with a few seconds absent, any data at all is better than no data.

Topological Optimization Of Multi Layer Perceptrons: A Comparative Analysis Of Architectural Symmetry And Computational Efficiency Across Diverse Data Domains

COMPUTER SCIENCE & APPLIED COMPUTATIONAL METHODS

My project analyzed the "Red AI" problem, where the central focus is model scaling to achieve higher accuracy. My goal was to determine the datasets that each architecture functioned best on, to reduce wasted computational time and resources and provide an efficient basis for smaller scale projects. Most current AI models rely on modular architectures, which leads to using exclusively Rectangles. I tested differing topologies (Funnels, Rectangles, Diamonds, etc.) across Iris, Wine, and Digit datasets from the UCI Machine Learning Repository, running 30 trials per data point. My findings were that the standard Funnel and Rectangle shapes were consistently the best options. They trained faster, handled noise better, and generalized well. Specifically, I found that, unexpectedly, the Rectangle architecture performed the highest on average, outperforming the Funnel. When testing wine analysis, the Rectangle yielded a 3.70% increase in accuracy compared to the Funnel while both performed 20 iterations. The topologies that included expansion, like the Reverse Funnel and Diamond, were generally inefficient and tended to overfit the data. The results showed a big difference for different architectures across the different data dimensionalities, revealing the importance of the initial architecture decision. The general rule discovered is that performance is maximized by maintaining topological continuity, meaning that sharp differences between any layers are minimized, and decreasing the large data compression (shrunk to 20 nodes in the final hidden layer for the Funnel), which removes too many details.

Ai Based System For Detection, Classification, And Removal Of Microplastics: A Novel Approach Using Artificial Intelligence, Machine Learning And Magnetic Oil Based Nanoparticles For The Detection, Identification, And Removal Of Primary, Polyethylene, And Polypropylene Microplastics Aged With Uv And Hydrogen Peroxide From Water

ENVIRONMENTAL SCIENCES & ENGINEERING

Microplastics are pieces of plastic that are less than 5mm in size. They have serious health effects on humans, such as entering the lungs and bloodstream. The objective of the project was to make a Machine Learning model that would detect microplastics in water and to discover if polymer weathering and microplastic type affected the efficiency of removal. It was hypothesized that weathered Polypropylene would be removed the most effectively from the water. This is because the weathering process uses chain scission which makes the surface full of jagged edges. These edges allow for effective removal because the oil and magnetic nanoparticles can anchor into the edges securely. This anchor creates an oil-bridge that makes the bond between the microplastics and the magnetic compound stronger, leading to more microplastics being removed. To test this, a ResNet-18 Convolutional Neural Network was created which took an image of water and classified the microplastics as Primary, Polyethylene, or Polypropylene. For the removal phase, Oleic Triglyceride and magnetic nanoparticles were added to the water. This mixture used the hydrophobic nature of the plastics to form a removable clump. The plastics were weathered using hydrogen peroxide and UV light for 48 hours. The trials with weathered Polypropylene had the most amount of microplastics removed. This is because the increased roughness and the presence of tertiary carbons in the Polypropylene allowed the magnetic compound to bond easier, leading to more microplastics being removed. This system provides a cost-effective and non-toxic method to help clean the environment.

Role Of Nitrogen Form In Regulating Cyanobacteria Growth And Chlorophyll A Production Under Controlled Conditions

ENVIRONMENTAL SCIENCES & ENGINEERING

Harmful algal blooms (HABs), caused by cyanobacteria such as Microcystis, represent a significant ecological and public health issue in freshwater systems, yet the influence of different nitrogen sources on bloom formation is not yet understood. The purpose of this study was to determine how inorganic nitrogen, organic nitrogen, and the absence of nitrogen affects the growth and photosynthetic biomass of Microcystis over time. The samples were grown under controlled laboratory conditions, and growth was monitored over an 18-day period using optical density to estimate cell density and chlorophyll-a concentration to assess photosynthetic activity. The results showed cultures supplied with inorganic nitrogen exhibited the greatest increases in both optical density and chlorophyll-a, indicating rapid growth and a high biomass. Cultures with organic nitrogen exhibited very minimal growth, and cultures with no added nitrogen exhibited moderate growth, suggesting that nitrogen availability limits Microcystis proliferation. These findings demonstrate that inorganic nitrogen most strongly supports cyanobacterial growth and suggest that managing inorganic nitrogen inputs may help reduce HAB formation in freshwater environments.

Seeing Drought Sooner: Comparing Sif And Ndvi For Vegetation And Drought Monitoring In The Wasatch Front From 2013 2024

ENVIRONMENTAL SCIENCES & ENGINEERING

Vegetation health can be monitored from a variety of different indices and benchmarks, one of the most common being the Normalized Difference Vegetation Index (NDVI). This index measures the relative “greenness” of an area, but more recent works suggest that Solar-Induced Chlorophyll Fluorescence (SIF), which measures photosynthetic activity directly, is a more accurate metric for measuring vegetation/ecosystem health and drought stress. This project compares these two metrics to benchmarks including Gross Primary Product (GPP), which measures the total amount of carbon dioxide produced over a certain area, and Standardized Precipitation Evapotranspiration Index (SPEI3), to learn which index is more accurate overall and for drought prediction. SIF showed a higher degree of accuracy when compared to both GPP and SPEI3 relative to NDVI, as well as showing an earlier indicator of drought. This suggests that SIF is a more accurate and earlier benchmark for evaluating drought stress and vegetation levels compared to NDVI.

Foaming Out Forever Chemicals: Air Water Partitioning Of Pfas In Wastewater Treatment And Implications For Human Exposure

ENVIRONMENTAL SCIENCES & ENGINEERING

This project investigates the potential partitioning of fluorinated compounds into the air through the process of foaming. This project exemplifies the crossroads of environmental engineering, chemistry, water engineering, and health to address one of the most notable environmental and health challenges of the 21st century. Perfluoroalkyl and polyfluoroalkyl substances (PFAS) are persistent “Forever Chemicals” that have attracted global concern due to their toxicity and ecological risks. Foaming results from surfactants in municipal treated wastewater and other surface water bodies separating into the air, allowing PFAS chemicals to escape the liquid phase in a more mobile form. Wind events have the potential to transport foam into the air and eventually be ingested by humans who live far removed from the original body of water. In my research, I investigated the presence of PFAS in foam collected from a wastewater treatment plant. High concentrations of several PFAS compounds were present in the foam. Additionally, I conducted bench-scale experiments to evaluate the effect of water quality parameters and surfactant concentration on PFAS partitioning. Results indicate that foam is a major removal mechanism for PFAS from the water matrix to the air phase. These findings demonstrate an intersection where environmental engineering and public health converge. The results will help health professionals develop risk models for PFAS cycling into airborne foam while providing a tool to the wastewater community for developing improved PFAS management strategies to decrease environmental and human health impacts.

Plastic Plant Pots

ENVIRONMENTAL SCIENCES & ENGINEERING

The project that I am working on is how to reduce plastic bags and convert them into something more useful. This led to my scientific question: how many layers can we compress to make a plastic pot sturdy enough to hold a plant with its necessities? I did it in sections to get a good general idea. The first part was knowing the layers I can use without burning holes or having it not sturdy, which was 6 layers. The second part of the project was to know what pattern I would have to use to make it a strong and sturdy pot.

Development Of Novel Water Treatment Technologies For Enhanced Removal Of Nutrients And Trace Contaminants In Aquatic Systems And Formation Of Freshwater And Fertilizer.

ENVIRONMENTAL SCIENCES & ENGINEERING

What does your daily routine consist of? When you wake up, brush your teeth, shower, or wash dishes, all these tasks require water. After being used, this water becomes contaminated and flows into the municipal sewage system. But what is in this water, and how can we reuse it again? Municipal wastewater comes from everyday activities like washing your dishes, showering, and using the toilet or bathroom, all of which end up in municipal sewage. This wastewater contains high concentrations of nutrients, organic carbon, nitrogen, and phosphorus. It also contains many harmful substances, such as estrogen, PFAS, and other persistent micropollutants. If not properly managed, these substances can harm human health and the environment and pollute our rivers. For decades, wastewater treatment plants have used the activated sludge process (ASP) to treat wastewater. ASP uses microorganisms to biologically remove nutrients and micropollutants. However, conventional ASP systems were designed for low concentrations of micropollutants and are not efficient at accommodating the increasing volume of wastewater generated by rapid population growth and climate change. As a result, more advanced methods are necessary. My research aims to improve the ASP using granular activated carbon (GAC). The GAC provides a surface for microorganisms to attach, grow, and form biofilm, enabling them to treat the water more effectively and remove nutrients and micropollutants. This improved system is more efficient than conventional ASP and can be easily retrofitted into existing wastewater treatment plants. By enhancing wastewater treatment, this process helps protect water bodies, ecosystems, and the environment while allowing cleaner water to be reused.

Investigating The Impact Of Varied Biodegradable Treatments On Enhancing Hydration Retention And Ignition Resistance In California Wildland Biomass

ENVIRONMENTAL SCIENCES & ENGINEERING

Recent climate changes are causing droughts and wildfires and severe damage throughout California. Wildfires are destroying thousands of miles of greenery and causing damages to properties and other human assets. Droughts are causing severe health issues in people resulting them to move across country. To address this issue, I wanted to design a biodegradable alternative to reduce the intensity of drought and wildfire. In my experiment, I combined the hydroxyethyl cellulose(HEC) with crushed mussel shells which is considered as the natural fire hydrant material to develop a biodegradable treatment to address this escalating issue.I then compared this treatment to a commercial alternative and an uncontrolled group and analyzed the moisture retention capabilities of soil and wood samples. The wood samples were then exposed to direct flame to assess how effectively each treatment delayed ignition and reduced overall burn time. The results showed that the biodegradable treatment outperformed all other groups in both moisture retention and fire prevention. I also investigated the treatment effects on soil pH and tested its performance across different soil types and wood species. Additionally, I created a new friendly treatment from compostable household waste as another potential method for reducing the impacts of drought and wildfires in California.

Modeling Pm2.5 Air Pollution During Temperature Inversions In The Salt Lake Valley Using Machine Learning

ENVIRONMENTAL SCIENCES & ENGINEERING

Air pollution is a major environmental and public health concern in the Salt Lake Valley, where winter temperature inversions frequently trap pollutants near the ground and cause dangerous increases in PM2.5. Short-term prediction of air quality during these events could help reduce health risks and give more information to the public. My project investigates how accurately air quality can be predicted using weather variables and python/machine learning techniques.

Publicly available PM2.5 data was obtained from the U.S. Environmental Protection Agency (EPA), while corresponding with weather data, temperature, humidity, and wind speed which was collected from meteorological databases. These variables were selected because temperature inversions are characterized by cold surface air, low wind speeds, and stable atmospheric conditions that prevent pollution from happening. The dataset was focused on winter months when inversion is most common.

I developed my machine learning model using Python and Googlecolab. Linear regression was used as a baseline model, and its performance was evaluated using time-based validation, in which the model was trained on earlier data and tested on later periods. Predicted PM2.5 values were compared to actual measurements over time to evaluate real-world performance.

Space To Sea: Decoding Whale Migration With Satellite Data

ENVIRONMENTAL SCIENCES & ENGINEERING

Can I predict where and when whales will feast using satellites? This project investigates the environmental drivers of baleen whale migration in the North Pacific by synthesizing multiple satellite-derived ocean variables.

Traditionally, cetacean migration has been tracked through unique fluke and dorsal fin pattern identification and acoustic movement to infer whale presence. However, such methods rely on sparse sightings, unreliable ecological proxies, and expensive tracking efforts. Additionally, global warming of the hydrosphere has caused phytoplankton, and thus krill populations, to become limited and sporadic.

I propose that whale migratory behavior is a direct response to a hierarchy of ocean conditions: chlorophyll-a initiates the food web, particulate backscattering signals the presence of zooplankton prey (like krill), and Sea Surface Temperature (SST) & Photosynthetically Available Radiation (PAR) act as environmental stressors that modulate the system's timing and quality. Using publicly available NASA and NOAA satellite data alongside collaborative whale sighting and tracking data, I will analyze these variables against whale speed (a behavioral indicator: slow=feeding, fast=searching/migrating).

In warmer years, whales spend more time searching for sustainable food sources. Therefore, I further test how thermal anomalies disrupt the baleen whales’ feast and famine migration system. By connecting space-based oceanography to whale movement, this project aims to reveal the critical, climate-sensitive link between microscopic marine life and ocean giants. The goal is to build a more predictive model of whale distribution, demonstrating how integrating multiple data "layers" from space provides a deeper understanding of marine megafauna ecology in a changing ocean.

Air Powered Electricity Harvesting Air Drag

ELECTRICAL & COMPUTER ENGINEERING

This project investigates the feasibility of capturing ambient kinetic energy, specifically air pressure resistance (drag) and ambient acoustic vibrations to generate supplemental electricity for vehicular systems. As a vehicle moves, it encounters significant air resistance and generates high-decibel ambient noise. The objective of this effort is to determine whether these "wasted" environmental factors can be effectively transduced into usable DC voltage to power small devices or eventually augment electric vehicle (EV) battery life.
The project goal is to design and test a low-cost energy harvesting circuit that utilizes an electret microphone as an acoustic-to-electric transducer. By capturing the pressure fluctuations from moving air, the signal is processed through an LM386 low-voltage audio amplifier circuit. To ensure the output is sufficient for practical use, the amplified signal is fed into a voltage quadrupler, which steps up the voltage to a level capable of powering a light-emitting diode (LED).
Controlled testing was conducted using a multi-speed handheld fan to simulate varying levels of air resistance and wind velocity. The varying fan speeds served as the independent variable to correlate air velocity with electrical yield. Initial results demonstrated that the circuit could successfully illuminate an LED, suggesting that with further scaling, aero-acoustic energy harvesting could provide a sustainable method for recovering energy typically lost to the environment during transit.

A Computational Framework For Transmission Line Theory Using Physics Informed Neural Networks

ELECTRICAL & COMPUTER ENGINEERING

This project demonstrates a proof of concept for using Physics-Informed Neural Networks (PINNs) to model electrical transmission lines. Through coding a custom PINN governed by the two first-order telegrapher's equation, creating a physically accurate transmission line simulation.

The network predicts voltage and current distributions along the length of a transmission line while simultaneously learning additional physical parameters required to produce realistic and physically consistent behavior. By enforcing the underlying differential equations as constraints during training, the model captures both spatial variation and wave propagation effects natural to transmission line theory.

This approach highlights how PINNs can bridge analytical physics and machine learning, enabling accurate system modeling without the need for real world data. The project serves as a foundational demonstration of how PINNs can be extended beyond proof-of-concept toward practical engineering applications, such as simulating transmission lines under varying conditions and optimizing design parameters for performance, stability, and efficiency.

Overall, this work demonstrates how physics-informed learning can provide a more accessible, adaptable, and interpretable approach to modeling electromagnetic and circuit-based systems, with potential to improve the design, reliability, and efficiency of modern technologies.

Economically Scalable Advance Wild Fire Alert System

ELECTRICAL & COMPUTER ENGINEERING

Wild fires destroy millions of acres every year. I wanted to examine how early detection could prevent or slow the spread of these fires. Because much of this forest land is remote and hard to monitor, fires may have a chance to spread and be harder to contain. I thought of the tsunami early detection devices and how they can alert people before the tsunami arrives. I wanted to see how I could apply that same idea over the large space of the forests. So I created an Esp32 device that uses a smoke dector and high rated temperature sensor to send a signal to a node which would be connected to the internet to alert the wild fire department. All of devices would be interconnected to nodes to spread out across the forest. Because of the scale of covering such a large area, affordability was a focus of mine to create these devices so that they would be reliable but at an affordable price so that a town could adopt these senors.

Securing The Edge Securing Operational Technology Networks In A Adversarial Cyber Environment

ELECTRICAL & COMPUTER ENGINEERING

A increasing convergence between Information Technology and Operational Technology has resulted in security incidents throughout the country. This project aims to determine attack vectors, common bad actor TTPs (Techniques, Tactics and Procedures), then utilize this information to harden a live Operational Technology environment against common vectors of attack, and reduce both internal and external attack surfaces.

Trash Robot

ELECTRICAL & COMPUTER ENGINEERING

This robot utilizes AI to detect trash using a Raspberry Pi camera. When trash is detected, the robot will move towards the trash and pick it up using a Robotic arm we have designed. It will then put the trash in the trash bin that we have attached to the top of the robot.

Signal Processing Optimization For Universal Bioamplifier Systems

ELECTRICAL & COMPUTER ENGINEERING

Medical devices which include EEG (brain) and ECG (heart) and EMG (muscle) monitors depend on specialized chips which operate only for one type of signal. The electronic characteristics of these signals may allow them to use identical hardware. This study will evaluate whether a single bioamplifier system can capture three different biosignals through testing their electronic properties. I will record EEG and ECG and EMG signals through the NPG Lite Explorer bioamplifier kit by using identical hardware settings except for changing the electrode placement. I will use Fast Fourier Transform (FFT) analysis to examine the frequency components of each signal type and check for shared frequency ranges between them. I will evaluate different amplifier gain settings to find the best signal-to-noise ratios (SNR) for each biosignal while testing whether one gain setting can be used across all three signals. I will use digital notch filtering to eliminate 60 Hz electrical interference and then evaluate how well it works. The research will show that a single bioamplifier system can capture multiple biomedical signals through its proper configuration which will advance the creation of universal programmable medical devices that decrease expenses while enhancing healthcare progress and reveal essential engineering decisions needed for designing multi-signal systems.

Backdrivable, Motor‑Actuated Knee Exoskeleton Using Imu‑Based Gait Detection For Adaptive Assistance And Haptic Feedback

ELECTRICAL & COMPUTER ENGINEERING

The human knee joint is one of the most complex and essential lower-limb joints. Knee ligament injuries often require long recovery periods and specialized support not available in commercial knee braces. The purpose of this project is to develop an assistive rehabilitation device that supports knee stability while promoting active movement to aid recovery from ligament related injuries. To accomplish this, a motor‑actuated knee brace was modeled with CAD and iterated upon continuously to create a compact, comfortable, and secure fit around the thigh and calf. Following the base brace design, an actuator was developed to provide active assistance throughout the walk-cycle. After evaluation, an outrunner motor was selected due to its high torque-output and scale.

The actuator design consists of a rotor and stator housing, a rotating shaft, and a stationary base, with the motor controller mounted to the stator. A neodymium magnet was integrated into the shaft to align with the motor controller for precise motion feedback. Connector modules were designed to mechanically link the actuator to the thigh and calf restraints using a dual‑locking screw mechanism, guaranteeing structural reliability. These connectors also house embedded IMU sensors within protective cavities, allowing gait and motion data to be collected and mapped to capture the user's unique instantaneous point of rotation.

Components were 3D-printed using flexible TPU and rigid PLA, with adjustable Velcro straps for universal fit. The final design integrates mechanical actuation, sensing, and structural support into a compact knee exoskeleton intended to improve stability and assist rehabilitation.

Signal Transmission Through Aqueous Salt Solutions

ELECTRICAL & COMPUTER ENGINEERING

Non-invasive biosensors are becoming very popular. They find many applications in “smart” devices like smart-watches, smart-phones, smart mirrors, and more. They have many advantages over traditional sensor versions like faster measuring times, less corrosion, longer service life, no need to clean between uses, etc. In this project, I attempt to improve on the non contact sensor I first designed last year for the science fair. This sensor detects ions in solution through magnetic fields generated by high frequency signals. My goal is to be able to determine concentration and types of ions in a given solution. By using a diode switch, it should be possible to evaluate resonance after the sine wave has been turned off. This should enable detection of ion type. By measuring both resonance and loss in different solutions, it is theoretically possible to detect changes in ion type and concentration in solution. The project also used a high current signal to be able to detect small signals in solution. A high resolution signal frequency range was also tested.

Automated Table Recognition And Data Extraction Using Hybrid Cnn Yolo Architectures

ELECTRICAL & COMPUTER ENGINEERING

Vast amounts of data are contained in datasheets, scientific papers, and scanned PDF documents, but manually extracting this information is time-consuming, error-prone, and inefficient. This Datasheet Scrubber addresses the need for an automated, accurate, and robust solution for extracting tabular data from PDF files. The tool combines image processing, deep learning, and Optical Character Recognition to detect and extract tables, using a hybrid CNN/YOLO-based architecture for automated table recognition. This work explores how AI-based table extraction can be optimized to reliably extract analog circuit parameters and how the resulting data can be structured and prepared for use in analog circuit design.

Electrostatically Modified Collagen Hybridizing Peptides For Noninvasive Molecular Imaging And Tissue Targeting In Pulmonary Fibrosis

BIOMEDICAL ENGINEERING & HEALTH TECHNOLOGIES

Over five million people worldwide suffer from pulmonary fibrosis, an incurable progressive lung disease that leads to respiratory failure and a 3–5 year life expectancy. With limited and costly treatment options, early detection is paramount to improving survival and preventing disease progression. However, current diagnostics such as CT scans and lung biopsies are invasive or detect only established scar tissue, delaying diagnosis by up to 12 years. To address this gap, this project presents the first electrostatically modified collagen hybridizing peptide (CHP) for early, noninvasive detection of pulmonary fibrosis.

Fibrosis is characterized by excessive accumulation of denatured collagen within the extracellular matrix due to dysregulated remodeling, an early molecular target not detected by conventional imaging agents. CHPs consist of repeating Glycine-Proline-Hydroxyproline (GPO) sequences that self-assemble into a collagen-mimetic triple helix and selectively bind denatured collagen. However, prior CHP probes showed nonspecific hepatic uptake in vivo, limiting imaging specificity and translational potential.

To improve targeting, a neutral G₃ linker was replaced with a negatively charged E₃ linker to reduce nonspecific macrophage interactions via electrostatic repulsion. The resulting CF-E₃-(GPO)₉ peptide was compared to the neutral control and evaluated for triple-helix stability using CD spectroscopy, collagen binding affinity in vitro, and selective targeting to pathological tissue using ex vivo fluorescence imaging of fibrotic lung tissue sections with scrambled peptide negative controls and heat-denatured collagen positive controls. The results demonstrate improved translational utility of CHP-based probes for noninvasive disease monitoring and can be conjugated to drug molecules for future targeted therapeutic delivery.

Aspect Ratio Driven Control Of Magnetic Particle Retention And Transport In Synovial Like Fluid

BIOMEDICAL ENGINEERING & HEALTH TECHNOLOGIES

Intra-articular drug delivery is widely used to treat joint diseases such as arthritis; however, therapeutic particles are often rapidly cleared by synovial fluid motion, limiting treatment effectiveness. Magnetic particles offer a potential method for externally controlled retention, but particle geometry is rarely optimized. This project investigates how particle aspect ratio affects magnetic retention and transport behavior in a synovial-like fluid.

Computational fluid dynamics (CFD) simulations were performed using SimScale to model particle motion under an external magnetic field. A Newtonian fluid approximation was used as a first-order representation of synovial behavior, and identical boundary conditions were applied to isolate the effect of geometry. Three particle aspect ratios were evaluated: spherical (1:1), moderately elongated (3:1), and highly elongated (5:1). Velocity magnitude and residence behavior near a target region were analyzed to predict retention trends.

A physical validation experiment was conducted using glycerin as a synovial analog and magnetically actuated steel bead assemblies representing each geometry. Retention time was measured as the time required for particles to reach and remain near a magnetic target zone. Both simulation and experimental results showed a consistent increase in retention time with increasing aspect ratio.

These findings demonstrate that particle geometry is a controllable engineering parameter for improving magnetic retention in synovial-like fluids.

Generating A Pd 1 Car Construct For Ms Therapy

BIOMEDICAL ENGINEERING & HEALTH TECHNOLOGIES

Autoimmune diseases occur when the immune system attacks healthy tissue, causing chronic inflammation and organ damage. Current treatments use broad immunosuppressants, which increase infection and cancer risk, underscoring the need for targeted therapies.

Regulatory T cells (Tregs) naturally suppress immune responses and maintain immune balance. Chimeric antigen receptors (CARs) are synthetic proteins that redirect immune cells; engineered into Tregs, they enhance precision in suppressing harmful immune activity. PD-1 is an immune checkpoint receptor that regulates immune activation and is often dysregulated in autoimmunity. It is highly expressed on pathogenic autoreactive immune cells.

This project explores whether PD-1-targeting CAR Tregs improve localization to inflamed tissues and suppress autoimmune responses. By comparing medium- and high-affinity PD-1 CAR Tregs to unmodified Tregs, this study evaluates their potential as a targeted autoimmune therapy.

To conduct this experiment, a CAR Construct plasmid was developed using techniques of E.coli amplification, double enzyme digestion, gel separation, and ligation. The plasmid was then cultured using mammalian cell lines, verified through flow cytometry and fluorescent microscopy. A mammalian cell line was then inoculated with both the CAR plasmids and helper plasmids for lentiviral packaging. The resulting lentivirus was transfected into T cells whose function was tested through flow cytometry.

Results are still in progress, but preliminary flow cytometry results show the successful expression of the CAR construct on the surface of inoculated mammalian cells, which is crucial for proper CAR function.

Optimizing Microbial Culture Conditions To Enable A Large Scale Production Of Heparin Precursors

BIOMEDICAL ENGINEERING & HEALTH TECHNOLOGIES

Heparin is one of the most widely used medicines in the medical industry due to its blood-thinning abilities, which is used for numerous surgeries and procedures. However, heparin is currently sourced from pig intestine, which causes many problems, such as contamination, impurities, ethical concerns, and many more. Additionally, the harvesting process is extensive, and the yield from each pig is very low. Making heparin in a lab, or bioengineering it, would be a practical solution to increase the yield and minimize the costs of production. However, when bioengineering heparin, the first step is to bioengineer heparosan, which is the precursor for heparin. The purpose of this project was to compare the growth of cell cultures in different culture types to maximize the yield of heparosan. This was done by growing two different cultures, one in flask cultures and one in a bioreactor culture, to help optimize the conditions and produce the highest yield of heparosan, and compare which method would be best. The results showed that the bioreactor produced a higher yield than the flask cultures did. This was because the bioreactor has more controlled conditions than the flask cultures do. This demonstrates that using a bioreactor for the production of heparosan is the most efffective step for the bioengineering process of heparin that helps to minimize the costs and increase the yield.

Mito Q, A Mitochondrial Targeted Antioxidant, Rescues Impaired Synaptic Plasticity In A Mouse Model Of Epilepsy

BIOMEDICAL ENGINEERING & HEALTH TECHNOLOGIES

Epilepsy is a neurological disorder characterized not only by recurrent seizures but also by persistent cognitive impairments, particularly in learning and memory. These deficits suggest underlying disruptions in hippocampal synaptic plasticity that extend beyond seizure activity alone. Previous research has identified mitochondrial dysfunction and elevated oxidative stress as central contributors to epileptic pathology; however, the specific impact of oxidative stress on long-term potentiation (LTP) in the CA1 region of the hippocampus remains insufficiently researcged. Furthermore, it is unclear whether targeting mitochondrial oxidative stress can restore normal synaptic function in epileptic tissue.

This study investigates whether reducing oxidative stress through the mitochondria-targeted antioxidant MitoQ can restore CA1 LTP in a corneal-kindled mouse model of temporal lobe epilepsy. Adult C57BL/6 mice undergo corneal kindling to induce epileptic activity while preserving overall neuronal structure. Acute hippocampal brain slices are prepared and maintained in oxygenated artificial cerebrospinal fluid to preserve neuronal viability. Electrophysiological recordings of field excitatory postsynaptic potentials (fEPSPs) are conducted in the CA1 region, and LTP is induced using theta-burst stimulation to assess NMDA receptor–dependent synaptic strengthening. Synaptic plasticity is quantified by measuring changes in fEPSP slope relative to baseline and analyzed using one-way ANOVA in GraphPad Prism.

By directly examining CA1 synaptic plasticity under controlled redox conditions, this study aims to clarify the mechanistic relationship between mitochondrial oxidative stress and learning-related plasticity in epilepsy and to evaluate the potential of mitochondria-targeted antioxidants as a therapeutic strategy for epilepsy-associated cognitive dysfunction.

An Adaptive Closed Loop Vibrotactile Neuromodulation Device For Reducing Tremor Related Propranolol Dosing Ramification

BIOMEDICAL ENGINEERING & HEALTH TECHNOLOGIES

Pathological tremor is a neurological condition that impairs fine motor tasks, affecting 1% of the general population and 4% of the elderly. Tremors arise when muscles’ micro-oscillations synchronize and phase lock, typically within a 4–12 Hz frequency range. Administering beta-blockers can reduce tremor severity, but doses are hard to personalize and limit, with heavy doses of propranolol correlating with low blood pressure, dizziness, and nausea. In this project, tremor was modeled, and a closed-loop control framework was developed to suppress tremor amplitude while minimizing pharmacological dependence.

Because using pharmacological suppression alone is constrained by many side effects, noninvasive neuromodulation was investigated as an alternative strategy. VTS was used to disrupt pathological tremor synchronization and reduce oscillatory amplitude. The procedure consisted of constructing a feedback-driven propranolol-reduction simulation by implementing a Hill curve, where tremor reduction, HR drop, and BP drop are calculated and utilized. Multiple dosing strategies were modeled, revealing that the most optimal dosing was a step-wise function regression. A device was then built to capture predictive physiological tremor-related data and create vibration using two LRA coin motors. It was then connected to a microcontroller, where optimal vibrotactile neuromodulation frequency and amplitude are determined through a feedback loop. The change in tremor amplitude is measured thereafter. Using the reduction percentage, the propranolol simulation determines the amount needed with the corresponding blood pressure penalization (representing side effects). This study demonstrates a closed-loop neuromodulation system that achieves equivalent dose tremor suppression while reducing modeled pharmacological side effect risk.

Development And Testing Of A Simulator For Multiple Types Of Blindness And An Accompanying App To Help Spread Awareness

BIOMEDICAL ENGINEERING & HEALTH TECHNOLOGIES

The project developed and performed a simulation of blindness to define requirements for visual prosthesis to better understand the visual information required in prosthetic eyes and an app that simulated many variations of blindness to spread awareness. A self-built VR-headset was used with python-coded software to adjust image resolution, convert video images into dotted outlines of shapes at varying resolutions, and create simulations of varying types of blindness. Human test subject volunteers were recruited and carried out two sets of tasks while observing their surroundings through the VR headset at different image resolutions and dotted outlines. The first test was a visual acuity test using a standard eye chart, counting the number of characters successfully read at 10 ft of distance from the chart with normal vision and reduced resolutions. The second test was the completion of an obstacle course with large cardboard obstacles, measuring time for completion and number of collisions with obstacles. Even at the highest resolution, the ability to read characters drops to 25%, course completion time increases by a factor >6, and the number of collisions increases. The app simulated six types of blindness (myopia, hyperopia, retinitis pigmentosa, glaucoma, cataracts, AMD) with the ability to adjust each to varying levels of severity. To determine its ability to spread awareness, the amount of downloads and user surveys will be taken into account. An expert in the visual-stimulation field verified that the app correctly depicted each type of blindness successfully and the different degrees of severity within each.

Applying Machine Learning To Predict Chemotherapy Drug Response

BIOMEDICAL ENGINEERING & HEALTH TECHNOLOGIES

In this project, we used a supervised machine learning model to predict the effectiveness of a specific lung cancer drug by using gene expression data. We also used regression models to get our desired output which was to predict a continuous value to be able to understand what the relationship between each drug response and gene was. The input values for our model were gene expression levels, which we obtained from the GDSC and DEPMAP database. As mentioned before, the output of the ML model were the drug sensitivity response values. After going on to match the NSCLC cell lines with both datasets, the data was cleaned and modified. After, the dataset was split down into training and testing groups. The training data was first used to teach our model how the changes in gene expressions can relate in the drug sensitivity response values. The model learnt the coefficients for each gene, going on to show how largely each gene contributed to the cisplatin's resistance. The trained model was then tested by comparing the amount of predicted drug sensitivity response values to the actual drug sensitivity response values from the datasets. A relation was seen, which was positive, going on to show that the model succeeded in finding a real biological relationship. This shows us that the machine learning model identified actual patterns between the gene expression lines and the chemotherapy drugs.

Comparing Transfection Efficiency Of Mc3 And Sm 102 Based Lipid Nanoparticles

BIOMEDICAL ENGINEERING & HEALTH TECHNOLOGIES

Chronic myeloid leukemia (CML) is a slower-growing cancer of the bone marrow that affects early forms of white blood cells (myeloid cells). Gene therapy in CML has been studied to improve the treatment of the cancer. Lipid nanoparticles (LNPs) have been widely investigated for a wide range of gene therapy strategies since the success of the COVID-19 vaccine utilizing LNPs. There are a number of ionizable lipids (ILs) available to form LNPs. In this study, we tested two different LNPs that are used in 2 FDA-approved drugs. One is MC3-LNPs, which is used in Onpattro (Patisiran), the other is SM-102, which is used for the Moderna COVID-19 vaccine. In this study, we tested their ability to transfect the CML cells as well as their cytotoxicity on these cells. We synthesized approximately 100 nm LNPs and determined their size, polydispersity index (PDI), and zeta potential values. Then, the intracellular uptake and cytotoxicity of the LNPs were investigated. Our results showed that both LNPs were successful in transfecting the cells and did not cause significant toxicity. On the other hand, the viability of the cells treated with MC3-LNPs was moderately higher. According to these results, the MC3-LNPs might be a better carrier for CML. However, more testing, such as mRNA encapsulation and protein translation efficiency, is needed to find whether specific characteristics of each IL are better.

Developing Anti Epileptic Drugs (Ae Ds) Using Bonnevillamides Found In The Great Salt Lake

BIOMEDICAL ENGINEERING & HEALTH TECHNOLOGIES

Approximately 1 in 26 people will develop epilepsy at some point in their life. Epilepsy is a chronic neurological disorder that causes repeated seizures as a result of abnormal electrical signals produced by damaged brain cells leading to a higher rate of physical and psychological problems and causing a significant increase in the risk of premature death. To prevent such risks, the development of various anti-epileptic drugs, both artificial and natural, is important in finding treatment options for everybody. Fortunately for those diagnosed with epilepsy, an estimated 70% of people can live seizure-free if properly diagnosed. Medication is prescribed based on the type of seizures the person has and which drugs they respond best to. While most drugs, like carbamazepine and valproic acid, are artificially created, drugs such as cannabidiol are derived from natural products. Recently, researchers discovered a new natural compound in the Great Salt Lake called bonnevillamides which was isolated from a strain of streptomyces bacteria that is known for producing bioactive compounds linked to antibiotics. Testing revealed the bonnevillamides possessed many of the same characteristics and initial testing on zebra fish have demonstrated potential in anti-seizure properties. This study will test the bonnevillamides potential as anti-seizure medication through a drug screening process using the Maximal Electrical Shock model, which is geared towards testing the efficacy of compounds that prevent generalized tonic-clonic seizures. This research will allow for better understanding of the bonnevillamides’ properties and will contribute to their potential use in developing new epilepsy treatments.

Evaluation Of The Efficacy Of Atorvastatin In Epileptogenesis Using The Corneal Kindling Mouse Model

BIOMEDICAL ENGINEERING & HEALTH TECHNOLOGIES

Epilepsy is a severe neurological disorder affecting over 50 million people globally, often developing through acquired causes such as traumatic brain injuries or infections. While current antiepileptic drugs (AEDs) manage symptoms, there are no approved treatments to stop epileptogenesis, the process by which the brain develops epilepsy. This project investigates atorvastatin, a common cholesterol-lowering drug, for its potential anti-epileptogenic properties.
The study utilizes the corneal kindled mouse model, an electrical stimulation model that simulates the progression of epilepsy. Twenty-eight adult CF-1 mice were split into two groups. A control group and a group receiving 10 mg/kg/day of atorvastatin. Mice received twice-daily electrical stimulations to the cornea, and seizure intensity was recorded using the Racine scale.
Results indicate that atorvastatin successfully delayed the onset of seizures. Mice treated with atorvastatin required significantly more stimulations to reach their first seizure and their first generalized seizure compared to the control group. Furthermore, the atorvastatin group displayed slightly lower average seizure scores throughout the process. However, both groups eventually reached a fully kindled state after a similar number of total stimulations. These findings suggest that atorvastatin may slow the rate of epileptogenesis and reduce seizure severity, though it may not stop the eventual development of epilepsy in this specific electrical model. This research addresses a gap, as previous studies had only evaluated atorvastatin using chemical kindling models.

Development And Analysis Of Mass & High Speed Effectiveness Of A Polyester Based Underride Mitigation System

MECHANICAL & MATERIALS ENGINEERING

On a majority of semi-trailer configurations, there exists a gap from the ground to the undercarriage of the trailer. This side clearance’s height allows the majority of a vehicle's hood to travel under it, while simultaneously being at the exact level to potentially intrude into the passenger compartment of a vehicle upon collision. Federal estimates state an annual average of 89 fatalities and 409 serious injuries in this manner, referred to as side underride accidents.

Aftermarket rail-based solutions exist to mitigate the fatal risk of side underride accidents, though they have received significant pushback due to concerns of increased weight, costs associated with such weight, and interference with utilities among other reasons. Polyester Strip-based Underride Mitigation System (PSUMS) addresses these concerns while maintaining integrity against an impact.

The development process began with sketches of concepts, which would evolve later into perceptible 3D models in the Autodesk Fusion 360 Computer-Aided Design (CAD) software. Testing utilized Fusion 360’s simulation features, particularly a static load study and dynamic event simulations.

PSUMS transitioned through three different renditions, the final of which proved to withstand significant static forces of more than 450,000 N. However, simulations also raised the question of biomechanics, and how structural integrity did not equate to survivability. Further implications would require a serious study and development of a design with the purpose of increasing survivability through increasing the time interval of a collision.

The Effect Of Tire Inflation Pressure On The Rolling Resistance Of A Mountain Bike On Rough Terrain

MECHANICAL & MATERIALS ENGINEERING

Rolling resistance is a significant resistive force during mountain biking and it can amount to 69% of the total resistance on a rough surface. Achieving a low rolling resistance is crucial to race performance. No known studies have used the reliable virtual elevation method to find how tire inflation pressure affects the rolling resistance of a modern mountain bike tire on rough terrain surfaces such as category 4 gravel. The rolling resistance of 2 different tire combinations, each inflated to 3 different tire pressures were tested: 10 psi, 15 psi, and 25 psi. The virtual elevation method was used to determine the average coefficient of rolling resistance over the course of the test lap for each inflation pressure. The lowest coefficient of rolling resistance was achieved during the 10 psi test. There was a 3.9% increase in rolling resistance between the 10 and 15 psi tests. Increasing the pressure from 15 psi to 25 psi resulted in an additional 5.5% increase in rolling resistance. On a 16 mile climb segment of a typical race course, decreasing tire pressure from 25 psi to 10 psi would save the rider 2 minutes and 9 seconds. Although this 10 psi pressure increases the risk of a flat tire, the additional mass of a tire insert only costs 14 seconds in time and allows the rider to ride with these low tire pressures.

Applying Principles Of Electromagnetism And Resistive Heating To Develop A More Accessible Method Of Metal Additive Manufacturing

MECHANICAL & MATERIALS ENGINEERING

Current methods of metal additive manufacturing (AM) typically require high-power laser or arc technology to melt metal into desired forms (powder bed fusion and wire arc AM, respectively). These methods introduce dangers in toxic metal powders and intense UV radiation exposure. Following a sharp increase in demand for low-cost, accessible, and precise production of increasingly complex part geometries, there is an urgent need for machines capable of meeting higher standards. This project aimed to evaluate a new method for melting down metal filaments, combining chamber heating with electromagnetic modulation to control outward flow velocity and direction. The combination was deemed theoretically capable of generating structures with higher resolution, enabling software to enhance precision when more advanced mechanical parts are unavailable. Furthermore, the heating method allowed for greater efficiency, requiring only enough electrical energy to maintain chamber temperature during the extrusion procedure. To test the novel arrangement, a toolhead and its internal components were constructed. A dual-drive extruder was designed, allowing for moving parts to be replaced in anticipation of many filament types and diameters, while simultaneously reducing slip and increasing extrusion speed range relative to other mechanisms. To protect delicate plastic components from heat, insulating and cooling regions were implemented between the hot-end and extruder. To demonstrate movement, x-axis and y-axis linear motion gantries were built. The stepper motors, extruder motor, heater, and electrical inverter modules, alongside numerous sensors and integral devices distributed around the machine, were controlled and powered continuously via an electronics system.

A Model For The Distribution Of Cadmium Interstitials In Cadmium Telluride Photovoltaics

MECHANICAL & MATERIALS ENGINEERING

Cadmium Telluride (CdTe) based solar cells are the leading commercialized thin film photovoltaic technology. But there is poor efficiency when trying to dope microscopic CdTe with elements such as Phosphorus, Arsenic, and Tin when attempting to increase its conductivity. This research project attempts to take a look at a possible reason why doping efficiency is so low. A simulation that models the distribution of Cadmium interstitials (extra atoms of Cadmium in the CdTe crystal structure) was used to model their distribution within the grains of the solar cells using differential equations. Matlab was used to numerically compute the concentration of Cadmium interstitials within the grains of the crystal with position as an independent variable. Once that was completed, the project was taken in related directions by more thoroughly examining how doping the solar cell with different elements affects the interstitials, and how time and temperature affect the interstitials at different positions. This research may not directly lead to a solution to the problem with doping CdTe solar cells, but it will give more insight into how they function and where research could head next.

Welding Optimization

MECHANICAL & MATERIALS ENGINEERING

How does the voltage you choose affect the efficiency and quality of your weld?
In my experiment I believe the higher the voltage the more material is used, and the integrity of the weld will stay about the same.
Prepare metal:
Cut c-channel to 6-inch bars
Grind down to 350g +/- 2g
Cut in half to about 3-inch pieces

Weld Pieces:
Use magnet and ground clamp to hold together tightly
Weld in a C pattern, varying speed maintaining a good pool

Analyze Final Product:
Dunk in cool water to expose flaws in weld (qualitative)
Record final weight (quantitative)

Repeat steps 3 times for each voltage setting

I found that level 6 on Miller welders, or the lower range for 1/8 inch gauge steel will create the cleanest weld, and the most efficient weld.

Send It Sweetie!

MEDICINE & HEALTH SCIENCES

Researching the differences in mountain biking race times as boys and girls progress in category and skill from 7th to 12th geade in an effort to determine at which point boys start to naturally further surpass girls

Endocrinopathies And The Heart: A Data Driven Cardiovascular Analysis

MEDICINE & HEALTH SCIENCES

The endocrine and cardiovascular systems form a tightly integrated regulatory axis essential for maintaining physiological homeostasis. The heart not only responds to hormonal signals that regulate metabolism, vascular tone, and cardiac performance, but also functions as an endocrine organ itself, secreting factors that influence distant tissues. Disruptions in hormonal balance, particularly involving thyroid hormones in the endocrine system, can therefore exert profound effects on cardiovascular structure and function, contributing to a broad spectrum of pathologies.

As cardiovascular disease remains the leading cause of mortality worldwide, understanding non-hemodynamic drivers of cardiac pathology has become increasingly important, especially thyroid dysfunction. This study investigates the hypothesis that thyroid dysfunction, including subclinical hypothyroidism, overt hypothyroidism, and low triiodothyronine (T3) syndrome, acts as an independent and statistically significant predictor of adverse cardiovascular outcomes, specifically heart failure and mortality. Building upon established physiological and epidemiological evidence, this project examines how hormonal excess and deficiency alter cardiac contractility, vascular integrity, metabolic regulation, and neurohormonal signaling.

To evaluate this hypothesis, a comprehensive, data-driven analysis using major U.S. public health and clinical databases, including the National Health and Nutrition Examination Survey (NHANES), the Medical Information Mart for Intensive Care (MIMIC-IV), the eICU Collaborative Research Database (eICU-CRD), and the All of Us Research Program was conducted. The study establishes standardized computational phenotypes across disparate electronic health record systems and provides a detailed framework for large-scale statistical analysis of cardiovascular outcomes.

The findings suggest that while traditional cardiovascular risk factors remain critical, endocrine status represents a potent and underutilized dimension of cardiovascular risk stratification. These results underscore the importance of integrating endocrine biomarkers into predictive models of heart failure and highlight the potential for earlier detection and targeted intervention through a more holistic understanding of cardiometabolic regulation.

Liposomal Encapsulation To Reduce Propofol Adsorption In Extracorporeal Membrane Oxygenation (Ecmo) Systems

MEDICINE & HEALTH SCIENCES

Extracorporeal Membrane Oxygenation (ECMO) is a life-saving cardiopulmonary support system for critically ill patients with severe lung and/or heart failure. Its use has risen sharply since the COVID-19 pandemic, with the 200,000th ECMO-supported hospitalization reported in 2022. These patients require multiple medications, yet clinicians lack reliable dosing guidelines because ECMO circuits introduce complications such as drug adsorption. Propofol, a commonly used sedative, is vulnerable to sequestration within hydrophobic circuit components due to its high log P (3.8). Studies report that up to 70% of administered propofol is lost within the circuit during the first 30 minutes, increasing the risks of toxicity, treatment failure, and mortality. To address this challenge, it was hypothesized that encapsulating propofol within liposomes could reduce adsorption while preserving its sedative function. Two liposomal formulations composed of 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) and hydrogenated-soy-phosphatidylcholine (HSPC) were synthesized to compare stability, ensuring the formulations were stable enough to limit circuit adsorption while enabling the encapsulated drug to achieve the desired pharmacological activity. The formulations were characterized for encapsulation efficiency using high-performance liquid chromatography; size, polydispersity index, zeta potential, and colloidal stability using dynamic light scattering; cytocompatibility using Cell Counting Kit-8; and morphology using transmission electron microscopy. Both formulations showed high encapsulation efficiency, stability at 37 °C and 4 °C, and favorable cytocompatibility. Ex vivo ECMO circuit studies further confirmed reduced drug adsorption. These findings warrant work to assess complement activation, validate sedative efficacy in an in vivo model, and use physiologically based pharmacokinetic modeling to predict dosing in pediatric ECMO populations.

Long Term Synaptic Potentiation In The Dentate Gyrus In Mice With A Traumatic Brain Injury

MEDICINE & HEALTH SCIENCES

In young adults, nearly 30% of diagnosed epilepsy develops from trauma to the brain. Depending on the severity of a traumatic brain injury, epilepsy can stem from a disruption of the synaptic plasticity in the hippocampus of the brain, which will result in impaired memory and learning. If a person is diagnosed with post-traumatic epilepsy, they will struggle with the limitations of this disease for the rest of their life; however, if a model is developed that creates a framework of the brain disruptions occurring between the time of the injury and the time it takes for the epilepsy to develop, drugs can be tested to prevent epilepsy in these patients. Yet, many parts of post-traumatic epilepsy have yet to be studied, and by looking into the electrical activity, scientists will be able to understand further how epilepsy can be detrimental to learning and memory. To create this model, CD1 and C57Bl/6 mice were tested using electrophysiology to study the electrical activity of neurons in the hippocampus. By testing how neurons change after the injury, an identification of a specific change was aimed for to alter the function of post-traumatic epilepsy. This leads to the question: How has the electrical excitability of the dentate gyrus changed in the CCI mice 3-4 weeks after injury?

From Risk To Action: Time Series Prediction Of Respiratory Support Requirements In Sepsis Patients

MEDICINE & HEALTH SCIENCES

Background: Sepsis commonly progresses to respiratory failure, requiring timely escalation or de-escalation of support. Reliable next-day predictions for specific respiratory-support modalities could improve bedside decision-making and anticipatory resource planning.

Methods: We performed a single-center retrospective study using de-identified MIMIC-IV v2.0 data. Adult ICU stays meeting Sepsis-3 criteria were converted into day-level sequences (ICU days 1–14); records with >30% missingness were excluded, yielding 10,566 stays. The outcome was next-day respiratory-support state (none, oxygen, HFNC, NIV, IMV, or tracheostomy). To enforce causality, only same-day or prior information was used, with labels shifted from day t to t+1 and a patient-stratified 70/15/15 train/validation/test split. Features included temporal lags, trends, rolling statistics, exponential moving averages, and task-oriented history (recent support and time since last event). XGBoost with Bayesian hyperparameter optimization was the primary model, compared with LightGBM, random forests, and RBF-SVMs. Class imbalance was addressed via training-only weights. Per-class thresholds were optimized on validation, followed by Hidden Markov Model decoding to impose temporal coherence. Performance was assessed using macro-AUROC, macro-F1, and a severity-weighted AUROC; calibration and SHAP-based interpretability were applied.

Results: The selected XGBoost model achieved macro-AUROC of 0.856 (validation) and 0.839 (test), with severity-weighted AUROC of 0.839 and 0.795, respectively. Predictions were well-calibrated in low-risk ranges and physiologically interpretable, with higher risk driven by worsening oxygenation and ventilatory burden over time.

Conclusions: Leakage-controlled time-series modeling enables accurate, interpretable next-day prediction of respiratory-support needs in sepsis, supporting real-time clinical monitoring, prospective evaluation, and multicenter validation.

Molecular Docking And In Silico Admet Approach To Identify Potential Phytochemical Lead Molecules Targeting Ezh2 For The Treatment Of Ar Low Triple Negative Breast Cancer And Endometriosis

MEDICINE & HEALTH SCIENCES

Our project aims to identify orally bioavailable phytochemical lead compounds that inhibit the canonical functions of the epigenetic regulator EZH2 as a targeted strategy for treating both EZH2-high triple-negative breast cancer and endometriosis. As the catalytic subunit of the Polycomb Repressive Complex 2, EZH2 trimethylates histone H3 at lysine 27, thereby inhibiting tumor suppressor pathways and promoting invasion, stemness, immune escape, and therapy resistance. In endometriosis, EZH2 suppresses estrogen receptor alpha, creating an oxidative environment that allows estrogen receptor beta to bind to the TNF-α promoter, driving inflammation and cell proliferation. In this project, we are screening a library of ~500 phytochemicals that have demonstrated activity against breast cancer cells, and have been filtered for oral bioavailability (using Lipinski's rule of five), for binding to EZH2 crystal structure (PDB ID: 5WFC). We utilized Schrödinger’s Maestro workflow for molecular docking to identify compounds with favorable binding poses and key interactions within the receptor grid/catalytic pocket. Top-ranked hits will then be refined using MM-GBSA calculations to more accurately estimate relative binding affinities and improve confidence. For promising lead molecules, we will use QuikProp to perform in silico ADMET profiling—predicting physicochemical properties and potential liabilities such as poor metabolic stability, excessive CNS penetration, and other risks. Our project addresses a major translational constraint for epigenetic therapies—the need for sustained exposure to maintain remodeled transcriptional states. Innovation lies in early safety de-risking and a faster, lower-cost path to commercialization (e.g., as adjuvant supplements).

Tumor Microenvironmental Signatures Associated With Low Darc/Ackr1 Expression In Solid Tumors, And Potential Mechanistic Insights

MEDICINE & HEALTH SCIENCES

Atypical Chemokine Receptor 1 (ACKR1/DARC) is a non-signaling chemokine receptor expressed on erythrocytes and endothelial cells. In tumor cells, ACKR1 modulates inflammation, chemokine gradients, and immune cell trafficking in the tumor microenvironment (TME). ACKR1 underexpression in bulk tumors is clinically associated with worse prognosis in breast cancer (BC). Black women with BC have a higher proportion of ACKR1-low tumors, which has been hypothesized to contribute to the 40% higher mortality observed in Black women with BC compared to their White counterparts. The aim of this study was to understand why low-ACKR1 breast tumors are associated with poorer prognosis. We began with a literature review of potential mechanisms regulating ACKR1 expression, complemented by mapping transcription factor (TF) binding sites within the ACKR1 promoter. bc-GenExMiner v5.2 pairwise correlation analyses were utilized to identify TFs likely to drive ACKR1 intratumoral expression. TIMER2.0 platform was employed to evaluate immune cell infiltration patterns associated with expression of (a) ACKR1, TFs that regulate ACKR1, chemokines, and angiogenic regulators, and (b) proteins whose overexpression drives proliferation and chromosomal instability in BC. These analyses found that low-ACKR1 tumors are “immune-cold,” have poor angiogenic support, and are depleted of cancer-associated fibroblasts, hematopoietic stem cells, and endothelial cells, while also exhibiting high proliferation and chromosomal instability. Compellingly similar trends emerged upon extending these analyses across seven different cancer types. Thus, DARC is an important regulator of chemokine-mediated immune and vasculogenic dynamics in numerous cancers, harboring potential implications for therapeutic targeting of low-ACKR1 tumors with presently dismal outcomes.

Molecular Docking And In Silico Admet Approach To Identify Potential Phytochemical Lead Molecules Targeting The Oncoprotein Atad2 For Treatment Of Ar Low Triple Negative Breast Cancer

MEDICINE & HEALTH SCIENCES

The oncogene ATAD2 is highly overexpressed in several cancers, including triple-negative breast cancer (TNBC), and its overexpression is associated with poor prognosis. ATAD2 regulates chromatin dynamics and is a transcriptional co-regulator that binds to acetylated histones through its bromodomain to enhance expression of downstream genes. Bromodomains tether oncogenic regulators to acetylated chromatin and sustain transcriptional programs that drive proliferation, survival, and treatment resistance. Additionally bromodomains have a defined binding pocket that can bind small molecule inhibitors, meaning bromodomain inhibition offers a pathway to selectively shut down tumor-enriched transcriptional dependencies while minimizing toxicity in healthy adult tissues. This project aims to screen a library of phytochemicals with established activity against BC cells, for orally bioavailable ATAD2 bromodomain inhibitors to identify lead molecules that are suitable for long-term use as adjuvant supplements alongside (or following) standard chemotherapy—an especially important need in TNBC, where recurrence risk remains substantial. Using the crystal structure of the ATAD2 bromodomain (PDB ID: 5R4V), a series of known compounds that have activity against BC tumors were put through in silico molecular docking using Schrodinger’s Maestro suite of tools to identify hits with favorable binding poses/interactions. The most promising compounds will be used in MM-GBSA analyses to calculate binding energies. Top-performing compounds will then be subjected to in silico ADMET testing via the QuikProp workflow to identify those most likely to achieve high oral exposure with manageable liabilities. The outcome of this project is a ranked set of candidate molecules with high predicted efficacy and safety profiles that could serve as lead compounds for future experimental validation.

Toxicological And P H Induced Mortality: Metabolic Stress In Drosophila From Artificial Food Additives

MEDICINE & HEALTH SCIENCES

This project investigates whether common artificial food additives and organic acid preservatives cause a higher rate of mortality and greater metabolic stress in Drosophila melanogaster compared to standard mineral salts. By testing compounds like Potassium Sorbate and Sodium Citrate alongside control groups such as Sodium Chloride and pH-balanced food, the study aims to differentiate between simple osmotic stress and the effects of the food additives caused. The hypothesis was that these preservatives create a significant metabolic burden by forcing cells to use excessive ATP beyond maintaining ionic balance, causing increased mortality.
The experimental results confirmed that chemical additives are significantly more lethal than mineral salts. Potassium Sorbate was the most toxic substance, resulting in 0% survival for both male and female flies by Day 15. While mineral salt groups followed a more typical survival trend, the preservative groups experienced rapid mortality. The high mortality in the Potassium Sorbate group was not caused by the pH of the food, as mortality remained similar even in pH-matched controls. This suggests that these additives interfere with biological factors beyond simple salt-induced stress, likely disrupting the tricarboxylic acid (TCA) cycle or other energy-production pathways. Future research will focus on measuring systemic water loss and testing varying concentrations to identify specific toxicological thresholds for these common chemicals.

Model Dependent Effects Of Local Sustained Release Estradiol/Testosterone On Fracture Healing And Screw Osseointegration In Ovariectomised Rodent Models

MEDICINE & HEALTH SCIENCES

Sex hormone deficiency is associated with impaired musculoskeletal healing, yet systemic estrogen/testosterone therapy carries risks that limit long-term use. Prior work links sex steroids to tendon to bone healing and motivated testing whether local sustained release delivery modulates bone repair.

We evaluated local estradiol/testosterone delivery in two rodent bone repair settings. In a femoral osteotomy model in C57BL/6J mice, WT and ovariectomised animals received local vehicle or combined estradiol/testosterone and underwent micro-CT at days 14, 21, and 28 (male/female analyzed: 4/5, 11/12, and 5/4). Outcomes included BV/TV, BS/TV, Tb.Th, Tb.N and Tb.Sp. In an osseointegration model, ovariectomized Sprague-Dawley rats (n=24) received a screw plus a 1.1x5 mm PCL implant loaded with estradiol:testosterone (3:1) or control. The samples were harvested at 14 days. Peri-Implant ROIs at 1.5 and 3.0 mm were analyzed, and histology quantified bone implant contact (BIC) and bone area ratio (BAR). Local estradiol/testosterone was associated with poorer fracture-callus microarchitecture (lower BS/TV with reduced Tb.N and increased Tb.Sp). Peri-implant micro-CT metrics showed no detectable treatment effect, but histology showed higher BIC (78±9% vs 41±13%) and BAR (36±4% vs 25±4%; p<0.001). These model dependent results motivate optimization of dose/timing and functional testing (torsion/pullout) to link structure to performance.

Arf6 As A Regulator Of Amyloid β Trafficking And A Potential Therapeutic Target In Alzheimer’s Disease

MEDICINE & HEALTH SCIENCES

Alzheimer’s disease (AD) is the leading cause of dementia, affecting about 6.9 million people in the United States aged 65 and older. A hallmark of AD is the accumulation of amyloid β (Aβ) in the brain, forming plaques that contribute to neuronal loss and brain atrophy. Although no cure exists, Aβ homeostasis is regulated by its bidirectional transport across the blood–brain barrier (BBB). Most Aβ is cleared from the brain through endothelial cell–mediated efflux into the bloodstream, followed by degradation in the liver and kidneys, while peripheral Aβ can enter the brain through BBB influx. ARF6, a small GTPase linked to neurodegenerative pathways, has an unclear role in Aβ trafficking. This study examined how ARF6 activity regulates Aβ transcytosis using an invitro BBB model. Human brain microvascular endothelial cells (HBMECs) were cultured on transwell inserts to form a monolayer, and adenoviral vectors were used to express wildtype (WT) ARF6, dominant-negative ARF6T27N, or fast-cycling ARF6T157N. Fluorescently labeled Aβ was added to either the apical or basal chamber to model efflux and influx, and transported Aβ was quantified. ARF6T27N significantly increased Aβ efflux (p = 0.0011), while ARF6T157N produced a marginal increase in Aβ influx (p = 0.0845). These results suggest that inhibiting ARF6 activity may reduce brain Aβ burden by promoting clearance and limiting peripheral entry. Ongoing studies are assessing the involvement of critical transport receptors LRP1 and RAGE in ARF6-mediated Aβ movement. Overall, this work identifies ARF6 as a potential therapeutic target for future AD treatment.

Ai Guided De Novo Engineering Of Bispecific Nanobodies Targeting Cd47 And Cd163 To Enhance Macrophage Mediated Tumor Phagocytosis

MEDICINE & HEALTH SCIENCES

Current cancer immunotherapies often struggle to penetrate dense tumor microenvironments or fail to effectively recruit the innate immune cells. This project proposes a novel solution: the development of bispecific nanobodies designed to physically bridge cancer cells and macrophages to induce tumor-specific phagocytosis. Using an RF‑antibody AI program linked to a thermal stability prediction program, we will develop nanobodies specific to cancer cells, using epitopes like CD47 or Her2. Similarly, the second nanobody able to bind to macrophages will be developed using macrophage-specific epitopes like CD163. Following this in silico design, both nanobodies will be expressed as a single bispecific nanobody where two functionalities will be linked in a single protein. We will express each bispecific nanobody in modified bacteria and purify them, enabling dual functionality.

We will characterize the specificity of these nanobodies by performing immunohistochemical staining of cancer cells. After evaluating specificity and selecting the most specific nanobody, we will add the purified bispecific nanobodies to a co‑culture of peripheral blood mononuclear cell‑derived macrophages and cancer cells.

To measure therapeutic efficacy, we will evaluate anti-cancer phagocytic activity of macrophages with and without the nanobodies, using a point-mutated nanobody as a negative control. Phagocytic activity will be evaluated using the Incucyte S3 platform and pH-sensitive dye to label cancer cells. This dye becomes fluorescent upon exposure to the acidic environment of lysosomes, indicating successful phagocytosis and digestion by macrophages. This will allow us to assess the ability of the nanobodies to effectively engage macrophages in phagocytic activity against cancer cells.

Development Of An Ex Vivo Human Retinal Model To Investigate Inflammation Mediated Neurodegeneration In Retinal Disease

MEDICINE & HEALTH SCIENCES

Retinal inflammation and ischemia are major factors in irreversible vision loss in diseases such as retinal stroke, diabetic retinopathy, and central retinal artery occlusion(CRAO) when blood flow to the retina is interrupted, depriving neurons of oxygen and triggering an inflammatory cascade that leads to neuronal death in over 80% of patients leading to vision impairment in hundreds of millions of individuals worldwide. Despite the severity of these outcomes, effective anti- inflammatory treatments are lacking, also due to the lack of human-based experimental models that recapitulate physiology and structure, including inflammatory response.This study established an ex vivo human donor retina platform capable of maintaining retinal and RPE tissue viability and inflammatory responsiveness up to a week.

This project investigated differences between the mouse retina and inflammation-based retinal injury using an ex vivo human retina model with lipopolysaccharide(LPS)-induced inflammation. A model to recover the Human Donor retina was created. Retinal tissues were exposed to LPS to simulate acute inflammatory conditions observed following ischemic injury. Samples were then cryosectioned and imaged using immunohistochemistry and fluorescence imaging to assess inflammatory signalling, neuronal damage, and glial activation, including IL-1β, TUNEL, and IBA1 expression. Qualitative image analysis revealed a significant increase in inflammatory markers and localized activation of glial cells following LPS exposure compared to untreated controls.

This platform provides a clinically relevant tool for studying acute human retinal inflammation and highlights key differences between mouse and human physiology. It enables future testing of targeted therapies aimed at reducing inflammation-driven neuronal damage.

Hard To Crack: Does Fluoride Really Work?

MEDICINE & HEALTH SCIENCES

The tooth enamel is the outermost layer of our teeth and is mainly composed of calcium based minerals. Even though it's one of the hardest substances in the human body, it can easily be weakened when exposed to acidic foods and drinks. This process is known as demineralization, when acid removes surface level minerals such as calcium, making it more vulnerable to cavities and other damage. Fluoride is commonly used in dental products because it protects the enamel and helps in restoring lost minerals. My project is determining whether fluoride is effectively reducing mineral loss after repeated acid exposure. It used eggshells to represent the enamel since they are both mostly composed of calcium carbonate. In my experiment, I exposed 2 eggs to acid using vinegar then treated 1 of the eggs with a diluted fluoride solution and left the other untreated. This process was repeated for three cycles to simulate the repeated pattern of consuming acidic foods and then brushing teeth. The mass of the eggs was measured before and after each cycle to quantify mineral loss. After each cycle, I recorded and analyzed the data to track patterns in mineral loss and better understand how fluoride affected the enamel.

(Imbalances In The Endocrine): Feedback Loop System In The Thyroid

MEDICINE & HEALTH SCIENCES

My project investigates the endocrine system by modeling the thyroid gland's negative feedback loop to understand how changing hormone levels influence overall system stability today. To represent hormonal communication, I designed an experiment using plastic bottles, balloons, yeast, sugar, and warm water to simulate how cells react to varying signals. Each bottle contained equal amounts of yeast and water but different sugar concentrations, including two tablespoons, one tablespoon, and one-half tablespoon for comparison purposes. Balloons were stretched over bottle openings to capture carbon dioxide produced during fermentation, symbolizing the body's visible response when hormone signals increase or decrease dramatically. Yeast activity represented target cells, while sugar symbolized hormone levels, allowing observation of how balanced, excessive, or insufficient signals could influence reactions within modeled systems. When sugar amounts increased, fermentation accelerated, and balloons expanded more, demonstrating how elevated hormone levels may push feedback loops toward overactivity and reduced regulatory control. Conversely, lower sugar concentrations slowed yeast metabolism, resulting in smaller balloon inflation, which modeled conditions where limited hormones cause underactive responses or weakened communication patterns. This setup provided a simple physical analogy for thyroid disorders, where imbalances disrupt normal signaling, preventing the endocrine system from maintaining steady internal conditions effectively. Careful measurement of balloon size over time helped compare reactions between groups, highlighting how variations in simulated hormones directly affected stability within feedback processes clearly. My hypothesis predicted that higher sugar concentrations would create greater balloon inflation, supporting the idea that excessive hormone levels can overwhelm negative feedback mechanisms severely.

Battle Against The Rays

MEDICINE & HEALTH SCIENCES

This project investigates, how different sun protection factors or SPF levels in sunscreen affect the amount of protection against the suns UV light rays. This experiment uses UV sensitive color changing beads to measure UV exposure, by applying different SPF’s of sunscreen onto the surface of each Ziplock bag the beads are in and placing them in the sunlight; the intensity of color change will indicate the amount of protection each SPF has. A lighter color change indicates greater protection while a darker color indicates more exposure. The results will help me understand the effectiveness of various SPF levels. This experiment will also explain the types of skin cancer having too much UV exposure can result resulting.

The Role Of Cdp Choline In The Metabolism Of Ctnnb1 Driven Hepatocellular Carcinoma

MEDICINE & HEALTH SCIENCES

Hepatocellular Carcinoma (HCC) is one of leading types of cancer in the world, with just over one million diagnosed cases worldwide and a low survival rate of around 10-20%. The main challenge in targeting HCC is the fact that there is no known root cause of the disease. Recent studies have pointed to evidence that lipid metabolism deregulating in the body is correlated with higher incidence rate of HCC. One of the most common mutations in HCC is the mutation of ctnnb1, which results in an unregulated production of Beta-catenin that accumulates and leads to tumor development. I used a transgenic line of zebrafish that had the Activated Beta Catenin gene to compare against a wild-type line of fish. The Kennedy pathway controls the production of cells mainly in the liver area through a series of phosphorylation reactions, but when deregulated, can overproduce the energy by-product of CDP-Choline that is used to build more cells at an unregulated rate. Through extensive experimentation on dosing these two different lines of zebrafish, I was able to find a statistically significant relationship in CDP-Choline and transgenic liver size growth in the Activated Beta Catenin fish, showing that CDP-Choline has the potential to worsen the conditions of HCC as a driver in metastasis. Further investigation of CDP-Choline can help us understand how its metabolic pathway works in human beings, and can help us work toward finding an eventual treatment in the future.

Bifenthrin Disrupts Neuroimmune Cell Viability And Bmp 4 Expression In Human And Invertebrate Models: Emerging Public Health And Ecosystem Safety Risks

MEDICINE & HEALTH SCIENCES

Bifenthrin (BF) is a synthetic pyrethroid used in insecticides, replacing DDT. Although BF is banned in the EU, it remains in use in the US. However, BF’s impacts on mitochondrial function, membrane integrity, and gene expression related to neurodevelopment and immune function remain poorly understood. This study investigates BF’s cytotoxic and molecular effects on human and invertebrate neuroimmune cells by assessing cell viability and BMP-4 expression. Human U937, HTB-11, and COLO320 as well as Mytilus edulis hemocytes and ganglion cells were treated with BF concentrations of 0.1, 1, 10, and 100 µM. MTT assays, ELISAs, and cell counts revealed dose-dependent decreases in viability for U937 and HTB-11 cells up to 41.08%(p<0.001), while COLO320 viability increased by 27.41% (p<0.001), suggesting a pro-proliferative effect in cancer cells. BF reduced marine invertebrate hemocyte viability by up to 43.41%(p<0.001) and increased ganglion death by ~165 cells per image(p<0.001). BF overexpressed BMP-4 in HTB-11 and COLO320 cells, with peak increases of 165.56% and 106.06% respectively (p<0.001), indicating neurodevelopmental and oncogenic signaling pathway disruption. These findings stress BF’s capacity to compromise neuroimmune cell function and gene regulation in human and marine systems, raising urgent concerns about its environmental safety and long-term public health impact. Further research is warranted to evaluate BF’s mechanisms and inform regulatory action.

How Exercises Affect The Human Body Heart Rate

MEDICINE & HEALTH SCIENCES

My project was the human heart rate because of how people exercise everyday. I wanted to know how the human heart rate would change as we did them. My hypothesis was based off of how people put more effort into stuff it can increase your heart rate more than if u barely put in any effort into it. I used my phone to check his heart rate during the experiment and after I put in all the exercises into their average form all 3 trials. My conclusion was me finding out that some exercise that use less effort and muscle can increase the heart rate more than other ones but the one that used the most muscle and effort increased it the most proving my hypothesis correct.

How Would The Sun's Heliosphere Interact With A Strong Galactic Cosmic Radiation?

PHYSICS, ASTRONOMY & MATH

This science experiment shows how the Sun’s heliosphere interacts with strong galactic cosmic radiation, such as a relativistic jet from a black hole. The heliosphere acts as a protective bubble around the solar system, shielding it from cosmic rays and interstellar particles. The goal of this project is to model and determine whether the heliosphere can still protect the Sun and its planets from extreme radiation beyond normal cosmic conditions. To simulate this, lasers of different power levels (30 mW, 140 mW, and 170 mW) were used to represent intensities of galactic cosmic radiation, and fish tanks filled with water modeled the heliosphere. The experiment measured the change in laser energy after passing through the tanks to calculate the percentage of change, mean, and standard deviation. Results are expected to show that the heliosphere, like the water barrier, significantly reduces radiation intensity but has limits when exposed to extreme levels of cosmic energy. This project helps demonstrate the importance and limits of the Sun’s heliosphere in protecting our solar system from intense cosmic forces.

Zenith Angle Versus Coincidence Rate For Cosmic Ray Muons

PHYSICS, ASTRONOMY & MATH

The purpose of this project is to determine the correlation between zenith angle of incoming muons to how often they occur. This is measured using coincidence rate, that is the rate at which a muon passes through all the detectors at once, ensuring accurate measures for zenith angle. I used Quarknet Cosmic Ray Muon detectors at varied Zenith angles to collect coincidence rate data. This science fair project is an analysis of that data, along with evaluating it compared to past data. This data was collected with the detectors facing east- west, and used a collection time of 12 hours for each data point. This experiment was collected at my home, with the data being collected on my laptop. The zenith angle was varied using a custom rig assembled out of Tslot supported by household items such as textbooks. The angle was confirmed using trigonometry. The results matched with previous groups' data, matching the trend of a negative correlation.

Airplane Wing Proposal

PHYSICS, ASTRONOMY & MATH

Our project proposes and evaluates new innovative airplane wing designs with the objectives of achieving greater structural strength and improving fuel efficiency. Our research investigated many different primary wing configurations, including a control (conventional), forward-swept, and hybrid types, as well as various wingtip devices, such as blended winglets, raked tips, and spiral designs. Our testing was deployed in both simulation and scaled wind tunnel models to analyze aerodynamic performance, structural integrity, and air displacement. The data from these tests will be used to determine the optimal geometry of the wing and wingtip treatment to better balance weight, durability, and drag reduction. We hope that research from this project will contribute meaningful data to aeronautical engineering and support the development of next-generation aircraft. This experiment uses a comparative method, where each unique wing and wingtip combination undergoes identical testing protocols to ensure consistent and reliable data on stress distribution, vortex displacement, and lift-to-drag ratios. The project's approach is trying to analyze both traditional and new design elements to discover solutions for modern aviation challenges.

Effect Of Charon’s Mass In The Stability Of The Pluto Charon System

PHYSICS, ASTRONOMY & MATH

Pluto and Charon have been the focus of much research into the modeling and evolution of tidally locked systems. Pluto exists in two distinct orbits with both Charon, its moon, and Neptune. The goal of this study is to model the effect of mass on the stability of the Pluto-Charon system. This study uses a RK4 simulation to model the orbits of the Pluto-Charon system. The mass of Charon is decreased by a set percentage of its total mass until it reaches zero. This study measures the average separation of Pluto and Charon as the mass of Charon decreases. The separation is compared to both the initial separation and the previous separation. The study records at what decrease in mass there are noticeable changes in separation and rate of increasing separation. This study is relevant to studies on both predicting the evolution of binary systems, such as those present in the asteroid belt and to understanding how Pluto’s permanence in the Kuiper Belt is dependent on the stability of its binary orbit.

Predicting Cycling Aerodynamics: Can Statistical Models Replace Cfd And Wind Tunnel Testing?

PHYSICS, ASTRONOMY & MATH

Aerodynamic drag accounts for 80% of resistance in competitive cycling, making coefficient of drag area (CdA) optimization critical for an athlete’s performance. Current methods for calculating CdA rely on expensive computational fluid dynamics (CFD) simulations or wind tunnel testing, limiting accessibility for both athletes and coaches. This study investigates whether multivariate regression modeling can accurately predict and minimize CdA from a cyclist's body position measurements, offering a practical alternative to simulation-based approaches.
Using 469 aerodynamic simulations with the top 5 postural and anthropometric variables (P < 10-12), the ordinary least squares (OLS) regression model predicts CdA with an R2 of 0.839. The model identified elbow width, weight, and hip bend as the three strongest predictors of aerodynamic drag (r > 0.4). Cross-validation reveals moderate overfitting (R2 gap between training and CV: 0.0354), indicating that additional data would improve reliability. Tests show a 5.71% error compared to simulation values, suggesting that the model captures the primary aerodynamic relationships needed to reduce CdA.
These findings demonstrate that regression models can replace CFD simulations for cycling position optimization while maintaining the precise accuracy necessary for cyclists and their coaches. Future work should expand the data set to reduce overfitting and continue to develop models for real world application. This approach equalizes aerodynamic optimization, making performance gains accessible beyond elite athletes with access to testing facilities and programs.

Stellar Simulations: How Fuel Density And Temperature Affect Fusion Reaction Rates

PHYSICS, ASTRONOMY & MATH

Fusion energy, the same energy that powers the sun, is a promising avenue for sus-
tainable, green energy. However, current fusion reactors face two major challenges: high
energy usage and fast material degradation. Unlike reactors, stars operate efficiently at
much lower temperatures with reactions far more difficult to achieve. They achieve this
efficiency with high temperatures and fuel density, so understanding how stellar fusion
can improve fusion reaction rates is key to discovering new principles in reactor design.
No empirical models exist that directly relate temperature and fuel density to reaction
rates. However, based on the reaction rate equation, I hypothesized that density would
have a quadratic relationship with reaction rates, independent of temperature. To test
this hypothesis, I analyzed stellar cores using stellar simulations to distill this complex
physical phenomenon into a direct model of how temperature and density affect fusion
reaction rates. These simulations showed that at low temperatures, increases in density
either reduced reaction rates or had no effect. However, as temperatures increased,
density had a greater and greater positive effect on reaction rates. This trend shows
that, contrary to the hypothesis, temperature and density interact to produce the most
fusion reactions. If this trend holds in more controlled environments, new fusion reactor
designs could increase efficiency by exploring this new avenue: mimicking stars.

Applying Black Holes As Cosmological Standard Rulers

PHYSICS, ASTRONOMY & MATH

Standard rulers are astronomical objects with known intrinsic sizes whose apparent angular sizes can be used to determine distances. Black holes (BHs), regions of space defined by gravity, have intrinsic sizes that can be calculated from their masses. Supermassive black holes (SMBHs) have observable angular sizes, allowing them to be applied as standard rulers.

My project uses the SMBH in the galaxy Messier 87 (M87*) as a research subject. Its mass is estimated using the distance-independent M-sigma relation, which correlates SMBH mass with stellar velocity dispersion. I refit this relation using Kelly (2007)’s LINMIX algorithm, excluding M87* from the sample to avoid circularity. The intrinsic size of M87* is calculated from this mass, and the Event Horizon Telescope’s measured angular diameter of 42.3 microarcseconds is used to compute M87*’s distance. From this distance, I calculate the Hubble Constant (the expansion rate of the universe) which is a function of distance.

I obtain a distance of approximately 4 (+150%, - 50%) e23 meters to M87* (23.6 ± 0.4 dex in logarithmic scales), corresponding to a Hubble Constant of 125 (+190, −76) km/s/Mpc. Uncertainties are rigorously propagated using quadrature addition. Accepted distances to M87 are about 5e23 meters, indicating substantial error.

I conclude that using black holes as standard rulers is currently experimental due to large uncertainties in distance-independent mass estimates, and lacking candidates. As far as I am aware, this is the first such scientifically valid and rigorous application of this concept.

A Novel Biochar Based Approach To Reduce Runoff Toxicity And Modulate Epigenetic Stress Memory In Crops

PLANT SCIENCES

Agricultural runoff water carries the pollutants as it flows through the fields and is often contaminated with harmful substances like heavy metals, pharmaceutical wastes, industrial wastes and micro plastics, which can be fatal to the agricultural crops. My project investigates whether different bioorganic biochars mixed with potato starch would be efficient in reducing the toxicity of the agricultural runoff water. It also focuses on mitigating the environmental stress responses of the Triticum aestivum crop to regulate epigenetic responses. In my project, two different biochars from cilantro, sugarcane bagasse was prepared and mixed with known amount of potato starch and then applied to runoff water prepared and then applied to the soil with different amount of salt stress to initiate the epigenetic stress. The amount of acetaminophen and microplastics in the filtered runoff water was found. Wheatgrass growth and the percentage of stress recovery of the wheat grass leaves were measured as indicators of plant stress. My experiment showed that the different bioorganic biochars mixed with potato starch can mitigate the contaminant toxicity in the agricultural runoff water and also have an influence in controlling the epigenetic stress responses of wheatgrass. I think this is an emerging field, and my study can serve as a steppingstone for future studies in the field can of epigenetics and toxicology research.

Ameliorating The Farming Process With Mycorrhizae By Improving Plant Growth And Soil Remediation

PLANT SCIENCES

With my project, I aim to study how mycorrhizae can help improve the farming process by testing if it can improve soil remediation, enhance plant growth, the amount of water needed for growth, and help with drought resistance. This project will be using Alfalfa. Cover crops can enhance soil fertility, reduce the need for fertilizers, and increase microbial activity. Furthermore, I have tested two different seeds (Mustard seed and Red Clover seed). In four different soils to test how affective the plants can grow with the mycorrhizae and if there is a change in soil nutrients in the pots with the different seeds and different inoculation status. Legume cover crops can help with nitrogen fixation by fixing atmospheric nitrogen in the soil by forming symbiotic relationships with Rhizobia bacteria. I want to test if the mycorrhizae can increase the amount of nitrogen that is fixed into the soil. Before conducting any experiments, I will test the soil for the nitrogen and Rhizo-bacteria levels. In the tests with the Alfalfa, there will be four sections that receive different amounts of water per week. I will have two trials of each condition. In each section, half will be inoculated, while the other half will not be. After the plants have reached maturity, I want to test the dry biomass of the plants and roots, the nutrient levels in the plants, and the number of Rhizobia bacteria. Then test the nutrient levels in the soil.

University of UtahScience & Engineering Fair

SAVE THE DATE: March 10-13, 2026

Senior Project Abstracts

Evaluation Of The Table Rock Petroglyph

Public Health Intervention: A Novel Hybrid Framework For Early Suicide Risk Detection And Prevention Using Real Time Linguistic And Socioeconomic Indicators

School Uniform Perceptions And Violations

How Ai Speaks Changes How People Think: Framing Ai Advice To Keep Moral Reasoning Human

Short Term Memory Accuracy

An Ensemble Learning Framework For Frailty Classification In Older Adults Using A Dual Screening Approach

The Impact Of Mouse Whisker Stimulation On C Fos Expressions In Neuronal Activity

Reliability Of Eeg To Identify Neural Signatures Of Cognitive Overload During Complex Task Performances

The Power Of Pressure

Scrutinizing Sound: Associations Between Phonemes And Game Provoked Emotions

Humanized Vs. Dehumanized Opponents In First Person Shooter Games: How Do Different Designs Impact Adolescent Aggression?

Ph It Mediated M Kate2 Fluorescent Tagging Of The Unc 25 Gene In Caenorhabditis Elegans

Evaluating The Necessity Of Bone Morphogenetic Protein Binding Endothelial Cell Precursor Derived Regulator (Bmper) In Modulating Beige Adipocyte Differentiation.

Crop Specific Microbial Biofertilizer For Indian Ricegrass: A Sustainable Strategy To Supplement Alfalfa Cultivation

Extracellular Vesicles From Activated T Cells As A Novel Early Diagnostic Approach For Type 1 Diabetes

Regulation Of Membrane Repair Pathways By Snare Proteins In Dmd

Effect Of Adding Probiotics To E.Coli Biofilm Bacteria

In Silico Identification Of Latent Tb Biomarkers For Crispr Cas13a Detection

The Effect Of Solution Type On Osmosis In Decalcified Chicken Eggs

Influence Of Uv Light On The Susceptibility Of Lactobacillus To Garlic

The Silent Treatment: How Goldenrod Reacts When Pollinators Disappear

The Effect Of Different Stiffness Matrices On Melanoma Cells

Un Folding Cancer: Evaluating Alpha Fold's Ability To Simulate Models That Predict Pathogenic Mutations

Fame Biodiesel Fuel: A Biochemical And Thermodynamic View Into The World Of Renewable Energy.

Developing A Dishcloth Using Biowaste: An Environmentally Friendly Paper Towel Alternative

The Effect Of Different Common Salts On Electrolysis

Machine Learning Optimization Of Conducting Polymer Thin Films With Magic Blue

Combining Soybean Induced Carbonate Precipitation (Sicp) And Hydroxypropyl Methylcellulose (Hpmc) To Prevent Wind Erosion

Restoring Native Riparian Vegetation In Utah: Halophilic Rhizobacteria To Improve Black Cottonwood Establishment In Saline Soils

The Azide Free Electrochemical Synthesis Of Aza Wittig Products

The Effect Of Fermentation Level On Lactose Derived Glucose Concentration In Dairy Products

Squeeze Or Pour: Which Do You Perfer

Promoting Lung Branching Morphogenesis Using Xyloside, A Small Molecule That Promotes Priming Of Endogenous Heparan Sulfate Glycosaminoglycans

Lithium In Salt Lake City Waterways: An Analysis Of Epidemiological And Extraction Potential

Selective Oncolytic Replication Of Human Rhinoviruses In Mc Py V Associated Merkel Cell Carcinoma

Guanine Nitrogenous Base Dna Damage Due To Oxidative Stress Induced By Metals Inhaled From Great Salt Lake Dust

Ph Levels Of Different Liquids.

Investigating Structure Activity Relationships Of Synthetic Cone Snail Peptide Analogs For N A Ch R Inhibition

Utilizing A Monte Carlo Particle Transport Simulation To Evaluate Lithium‑Based Materials For Enhanced Tritium Breeding, Shielding, And Energy Deposition Performance In Nuclear Fusion Reactor Systems.

An Early Warning System For Quantum Algorithm Failure: Detecting And Classifying Decoherence In Grover Search Using Distribution Based Metrics

Early Detection Of Skin Cutaneous Melanoma With Artificial Intelligence Techniques

Statistical Analysis Of Osimertinib Usage In Nsclc

Using 3 D Models To Predict Alzheimer's In Adults: Brains Of Change

Optimizing Epilepsy Neurostimulation Therapies Through Analysis Of Electrode Distance And Regional Neural Firing Dynamics

Employing The Synthesis Machine Learning Algorithm To Synthesize New Novel Crystal Structures

Understanding The Feasibility Of Streaming Black Box Data To The Cloud

Topological Optimization Of Multi Layer Perceptrons: A Comparative Analysis Of Architectural Symmetry And Computational Efficiency Across Diverse Data Domains

Role Of Nitrogen Form In Regulating Cyanobacteria Growth And Chlorophyll A Production Under Controlled Conditions

Seeing Drought Sooner: Comparing Sif And Ndvi For Vegetation And Drought Monitoring In The Wasatch Front From 2013 2024

Foaming Out Forever Chemicals: Air Water Partitioning Of Pfas In Wastewater Treatment And Implications For Human Exposure

Plastic Plant Pots

Development Of Novel Water Treatment Technologies For Enhanced Removal Of Nutrients And Trace Contaminants In Aquatic Systems And Formation Of Freshwater And Fertilizer.

Investigating The Impact Of Varied Biodegradable Treatments On Enhancing Hydration Retention And Ignition Resistance In California Wildland Biomass

Modeling Pm2.5 Air Pollution During Temperature Inversions In The Salt Lake Valley Using Machine Learning

Space To Sea: Decoding Whale Migration With Satellite Data

Air Powered Electricity Harvesting Air Drag

A Computational Framework For Transmission Line Theory Using Physics Informed Neural Networks

Economically Scalable Advance Wild Fire Alert System

Securing The Edge Securing Operational Technology Networks In A Adversarial Cyber Environment

Trash Robot

Signal Processing Optimization For Universal Bioamplifier Systems

Backdrivable, Motor‑Actuated Knee Exoskeleton Using Imu‑Based Gait Detection For Adaptive Assistance And Haptic Feedback

Signal Transmission Through Aqueous Salt Solutions

Automated Table Recognition And Data Extraction Using Hybrid Cnn Yolo Architectures

Electrostatically Modified Collagen Hybridizing Peptides For Noninvasive Molecular Imaging And Tissue Targeting In Pulmonary Fibrosis

Aspect Ratio Driven Control Of Magnetic Particle Retention And Transport In Synovial Like Fluid

Generating A Pd 1 Car Construct For Ms Therapy

Optimizing Microbial Culture Conditions To Enable A Large Scale Production Of Heparin Precursors

Mito Q, A Mitochondrial Targeted Antioxidant, Rescues Impaired Synaptic Plasticity In A Mouse Model Of Epilepsy

An Adaptive Closed Loop Vibrotactile Neuromodulation Device For Reducing Tremor Related Propranolol Dosing Ramification

Development And Testing Of A Simulator For Multiple Types Of Blindness And An Accompanying App To Help Spread Awareness

Applying Machine Learning To Predict Chemotherapy Drug Response

Comparing Transfection Efficiency Of Mc3 And Sm 102 Based Lipid Nanoparticles

Developing Anti Epileptic Drugs (Ae Ds) Using Bonnevillamides Found In The Great Salt Lake

Evaluation Of The Efficacy Of Atorvastatin In Epileptogenesis Using The Corneal Kindling Mouse Model

Development And Analysis Of Mass & High Speed Effectiveness Of A Polyester Based Underride Mitigation System

The Effect Of Tire Inflation Pressure On The Rolling Resistance Of A Mountain Bike On Rough Terrain

University of Utah
Science & Engineering Fair