Schedule of Events
Science Showcase Poster Session
Friday, May 2, 2025
11:00 am - 12:30 pm
Swenson Science Center, 2nd floor lobby
Students from the Division of Natural Sciences will present their research and class projects in an interactive poster session. This event highlights student work across disciplines like biology, chemistry, physics, and environmental science, offering a closer look at the scientific questions they’re exploring. Attendees can walk through the posters, ask questions, and engage directly with student researchers to learn more about their discoveries and methodologies.
Student Abstracts
Mapping the Racing Stripe (RS) Transgene in Drosophila Melanogaster
The novel Racing Stripe (RS) phenotype in Drosophila melanogaster is characterized by the expression of a white+ (w+) transgene in a thin equatorial stripe along the dorsal-ventral (D-V) midline of the compound eye. This striking expression pattern suggests that the w+ reporter has come under the transcriptional control of a gene active at the eye equator, a critical site for cell-cell signaling during early eye development. In addition to this spatially restricted expression, the RS insertion is homozygous lethal, indicating that the transgene may have disrupted a gene essential for embryogenesis. To identify the gene affected by the RS insertion, we performed a series of complementation tests by crossing RS/Balancer flies to eight lethal deficiency stocks spanning a region previously mapped for RS. These crosses revealed a failure to complement lethality within the region spanning base pairs 4,805,800 to 4,950,800 on chromosome arm 3R. Closer analysis of RS/RS embryos revealed segmentation defects, consistent with disruption of a gene involved in early patterning. Within the identified genomic interval lies the segmentation gene odd paired (opa), a transcription factor required for establishing parasegment boundaries during embryonic development. Examination of RS/RS embryos revealed a canonical opa mutant phenotype, further supporting the hypothesis that RS disrupts opa function. We are currently conducting direct complementation tests between RS and known opa alleles to definitively determine whether the RS insertion is located within or near the opa locus. If confirmed, RS may represent a novel opa allele or enhancer trap, providing a unique tool for studying opa regulation and its role in eye and embryonic development.
Student(s):
Nasharli Abeygunawardene
Faculty Mentor:
Dr. David Marcey
Attuned
Attuned is a mobile social media app focused on podcasting. Attuned brings a greater social aspect that I believe podcasting currently lacks. Attuned has all the features of a traditional social media platform. Including account creation, customization, comment sections, direct messaging, etc. However, it also features an exceptional media view where users can watch or listen to podcasts. This app combines a well designed user interface, a robust back-end, and an AI processor to bring an exceptional user experience to podcast listening.
Student(s):
Connor Adams
Faculty Mentor:
Dr. Craig Reinhart
A Comparative Study Between Microplastic Contamination in Channel Islands Harbor and Santa Clara Riverbed
Microplastic pollution is a significant environmental concern, particularly incoastal and agricultural regions where human activities contribute tocontamination. This study compares microplastic and microfiber pollutionbetween Channel Islands Harbor, a recreational and commercial hub, and theSanta Clara Riverbed, which runs through agricultural land before emptyinginto the Pacific Ocean. By analyzing sediment samples from both sites, weaim to identify the sources and distribution of microplastic contamination inthese contrasting environments.Sediment samples were collected at high tide marks from both locations andprocessed using vacuum filtration and Rose Bengal staining to distinguishsynthetic microplastics from organic material. Samples were examined undera Nikon dissecting microscope for quantification, and Fourier TransformInfrared Spectroscopy (FTIR) was used for material identification.Preliminary results indicate higher concentrations of microfibers in theharbor, likely due to boating activity, urban runoff, and marineinfrastructure, whereas the riverbed samples contained more microplastics,likely originating from agricultural runoff, textiles, and irrigation systems.The findings highlight distinct pollution sources in coastal and inlandenvironments, emphasizing the need for further research on pollutionmitigation and regulatory measures.
Student(s):
Kiana Adli
Faculty Mentor:
Dr. Robert Richards
Overexpression Methos for DesD to Investigate Binding Kinetics
DesD is an NIS synthase involved the the synthesis of stealth siderophores in virulent bacteria. Previous work in the Hoffmann Lab established a binding assay for DesD with the goal to construct a full cofactor binding profile of DesD. Currently issues with overexpression and purification of DesD held up the further development of a DesD cofactor binding profile. Methodically changing factors that could negatively contribute to optimal overexpression and purification of DesD, we were able incrementally improve overexpression. Further directions include confirming correct sequence in the stored vector samples. Repairing the issues in the central protocol is key to continue any future work.
Student(s):
Ashlee Bacconi, Dr. Katherine Hoffmann
Faculty Mentor:
Dr. Katherine Hoffmann
Generation of a Novel Canine Tumor Cell Line for In Vitro Cancer Studies
Rates of cancer across the world have steadily increased throughout the past decades, leading to an increased need for research aimed at diagnosing and treating cancer. While mouse models have led to significant advances in our understanding of cancer due to their physiological similarities to humans, widespread availability, and ability to be modified, such models have their limitations. These limitations include differences in tumor development between mice and humans, variety in genetic complexity, and overall differences between environmental factors that may play a role in the development of cancer. This has led to the rise of comparative oncology focused on companion animals, specifically canines, as a way to better understand cancer in humans. The rise of canines as a model system as an alternative to mice in cancer studies stems from its similarities in genetic complexity, tumorigenesis, and everyday environment as humans. Our study aims to generate a novel canine tumor cell line, using an enzyme called telomerase, for in vitro cancer research. Cell deterioration is associated with the shortening of a cell’s chromosome ends (telomeres) due to decreasing telomerase levels over time. However, when telomerase is reactivated, it can lengthen telomeres indefinitely, achieving immortalization, as observed in cancer cells. We hypothesize that by transfecting primary canine cells with telomerase, we will activate immortalization and generate a novel canine cell line that allows us to study various aspects of canine and human cancer and provide a new cell line model for research in comparative oncology.
Student(s):
Nevign Angelique Besas
Faculty Mentor:
Dr. Chad Barber
Optimizing Event Selection Search for Supersymmetric Top Quark Pair Production & Dilepton Decay
In the search for Supersymmetric (SUSY) Top Quarks produced through the Di-Lepton channels from collisions in the Large Hardon Collider (the highest energy particle accelerator), event data was processed through an event slimming preselector for electron-electron (ee), electron-muon (emu), and muon-muon (mumu) decay channels. This selection predominantly left SUSY, TT, WZ, WW, and DY events, which were further filtered using “cuts” on the kinematical variables associated with each event for each channel. The ee channel cuts were completed by previous students under this research project, this project added the emu, mumu, and METs filters. These filtered results were then used to train and optimize the machine learning program used to automate the filtering process on additional events. Significant work remains before publication, including, but not limited to, immediate steps such as running studies on the sources of different systematic errors associated with signal identification and continuing the jet energy scale errors. This research must continue to find the never-observed SUSY, which is essential for unifying all the fundamental forces and providing the identity for dark matter.
Student(s):
Khusanjon Bobokhojaev, Carys Garvey
Faculty Mentor:
Dr. Sebastian Carron Montero
The Analysis of Microfibers in Sediment at Silver Strand Beach in Oxnard, CA
The purpose of this research project is to analyze the concentration of microfibers in sediment at Silver Strand Beach in Oxnard, CA. The study focuses on whether there are differences in microfiber concentrations along the north and south ends of open coast beach. Microfibers, a category of microplastics originating from synthetic textiles, are abundant in marine environments, mainly due to laundry emissions. The study aims to examine the impact of microfiber pollution in marine environments by collecting and analyzing sediment samples from both ends of the same beach. Using previously established practices used in the Huvard lab, the project will count and compare microfiber concentrations from different ends of the same beach. This is to assess how environmental factors contribute to the presence of microfibers. The expected outcome is that microfiber concentrations will be higher at the south side of Silver Strand Beach. This research displays the amount of microfiber pollution at Silver Strand Beach.
Student(s):
Hanna Booth
Faculty Mentor:
Dr. Andrea Huvard
Utilizing Foraminifera to Reconstruct Paleo-Methane Seeps and Paleoceanographic Changes in the Norwegian Sea
We studied six taxa of benthic and planktonic foraminifera (i.e., Cassidulina spp., Cibicides lobatulus, Elphidium spp., Melonis barleeanus, Neogloboquadrina pachyderma, Nonionella labradorica) from one gravity core (GC01; 220 m water depth) collected from the Hola trough, off the coast of the Lofoten Islands (68°55.123’ N 14°17.194’ E). We selected this core because it was from an area where seep carbonate occurrence indicated a long history of methane emissions (Sauer et al., 2017). By analyzing benthic foraminiferal assemblages, studying the possible impact of methane on the preservation of foraminiferal shells (Foram Preservation Index (FPI); Poirier et al., 2021), and analyzing the stable isotope composition of the isolated specimens (𝛿¹⁸O and 𝛿¹³C) we aimed to understand how benthic foraminifera responded to potential methane activity and, possibly, paleoceanographic changes during the most recent geological past (~16 kyr). Stable isotope data do not suggest methane seepage close to the site because there is no clear excursion in 𝛿¹³C in any of the genera. Notably, Cassidulina and Elphidium experienced a marked change in their preservation with more poorly preserved assemblages at the base of the core. These results, combined with chemical data (X-Ray Fluorescence) collected on the whole core, suggest paleoceanographic changes.
Student(s):
Emma Caputo, Megan Fung, Giuliana Panieri, Claudio Argentino, Chiara Borrelli, Norrie Ling, Lara Sullivan, Robert Poirier
Faculty Mentor:
Dr. Megan Fung
LuxSpot
LuxSpot is an IOS mobile application that helps individuals find date spots according to their established preferences such as location, budget, facilities, and individual preference. The app helps individuals search for date spots such as restaurants, parks, cafes, and more, with an established filter to limit the options. It features a user-triggered input system, whereby users will be able to submit new date locations which will then go through an approval system before being added to the app. The app also includes AI-based suggestions, where users will be given customized date suggestions and itineraries through their profiles and interests. Users will be prompted to give feedback on their experience once they have been to a spot, providing useful input that assists the app in making recommendations. A list and map view make access easy, and users can save a favorite location to a wish list. Businesses can even promote their locations as a featured location for a fee. This app has been developed using Swift and Xcode with the implementation of a Firebase database for storing and managing data. By providing a comprehensive date planning tool, this project enhances the user's experience and makes it easier to do so while finding enjoyable, niche date spots.
Student(s):
Adam Cartozian
Faculty Mentor:
Dr. Craig Reinhart
Personalized Learning Platform for Golf Using Machine Learning
With recent advancements in sports technology, there have been significant improvements to training simulators and detailed automated skill analysis programs. In golf, there are numerous programs designed and developed to analyze players’ swings and assess their strengths and weaknesses across various types of shots and clubs, from driving to putting. However, these tools often fall short in integrating the collected player data with automated and personalized training programs tailored to an individual's statistics and training availability. This project utilizes a web-based app with a machine learning algorithm to bridge this gap, providing players and coaches with automated, customized training schedules. By streamlining this process of data analysis and training program development, this learning platform aims to increase training efficiency and accessibility for golfers at all levels.
Student(s):
Ellie Cavanagh
Faculty Mentor:
Dr. Craig Reinhart
Fueling the Fire: Data-Driven Insights into Heat and Wind Effects on Ventura County Wildfires
Between January 1, 2013 and December 15, 2024, Ventura County experienced 31 wildfires exceeding 10 acres, 11 of which burned over 1,000 acres. Many of these large fires occurred following significant heat events and during late-season Santa Ana Wind conditions. This project aims to investigate the role of meteorological conditions in driving wildfire behaviors across the region, using historical weather observations from the National Weather Service and CAL Fire.
Machine learning models (random forest and lasso regression) were applied to identify key meteorological predictors of wildfire risk. Heat events and relative humidity were revealed as the strongest predictors of air temperature variation throughout the year (R² = 0.54), while wind speed and gusts were the dominant drivers of wildfire spread given ignition (R² = 0.87). Pearson correlation analysis revealed that years with stronger inverse relationships between air temperature and relative humidity, and stronger alignment between temperature and wind (e.g., 2018 and 2024), corresponded to larger fire burn scars.
Spatial Mapping with ArcGIS and analysis further show that wildfire occurrences and severity have increased, with fire seasons extending to later in the year. These findings emphasize the increasing influence of climate change on wildfire dynamics while supporting integration of data-driven modeling into localized fire risk assessments and hazard zone planning.
Machine learning models (random forest and lasso regression) were applied to identify key meteorological predictors of wildfire risk. Heat events and relative humidity were revealed as the strongest predictors of air temperature variation throughout the year (R² = 0.54), while wind speed and gusts were the dominant drivers of wildfire spread given ignition (R² = 0.87). Pearson correlation analysis revealed that years with stronger inverse relationships between air temperature and relative humidity, and stronger alignment between temperature and wind (e.g., 2018 and 2024), corresponded to larger fire burn scars.
Spatial Mapping with ArcGIS and analysis further show that wildfire occurrences and severity have increased, with fire seasons extending to later in the year. These findings emphasize the increasing influence of climate change on wildfire dynamics while supporting integration of data-driven modeling into localized fire risk assessments and hazard zone planning.
Student(s):
Anthony Delgadillo Salas, Dr. Christopher Brown
Faculty Mentor:
Dr. Christopher Brown
Investigating the Chemical Surface Inhomogeneities of the Magnetic B-type Star HR 3042
To search for magnetic fields in evolved OBA stars, the Large Impact of magnetic Fields on the Evolution of hot stars (LIFE) project used the spectropolarimeter ESPaDOnS@CFHT to observe the B8 II star, HR 3042. Initial observations showed the average line shape was asymmetric. Such asymmetries can be caused by chemical inhomogeneities, pulsations, and/or binary companions. An additional 16 spectropolarimetric observations were taken and the data were normalized. The equivalent width (EW) was calculated using python for several lines of He, H, Fe, Si, Mg, and C. For single elements, spectral lines for different transitions sometimes matched well, while others matched poorly. For example, Fe II 4584 Å and Fe III 5127 Å experience a peak at a rotational phase of 0.5, while Mg II 4481 Å experiences a peak and Mg II 6347 Å experiences a minimum at a rotational phase of 0.5. We used Specpolflow and the multi-line technique Least Squares Deconvolution (LSD) to determine average Stokes I and Stokes V line profiles from which the surface-averaged longitudinal magnetic field strength (Bl) was calculated. We observed both the positive and negative poles of the magnetic field with the maximum value of the dipolar magnetic field strength, Bd, of estimated value of 629.4 G occurring at a rotational phase of ~0.5. We note that the EW extremes generally occur when the magnetic field is at its maximum; thus, the variation of chemical abundance and magnetic field are likely connected.
Student(s):
Leah Derrick
Faculty Mentor:
Dr. Mary Oksala
Calendar Scheduler Desktop Application
This project implements a functional and user-friendly personal calendar management system. This application features a user authentication system that requires username and password verification before allowing users access to their personal calendar. After authentication, users are able to straightforwardly create and remove events specific to their accounts. A highlight to this system is the synchronization in real-time between the frontend user interface and the backend database. This allows the application to provide constant changes to the interface in the calendar and login process. As this project prioritizes simplicity, it offers a more intuitive learning curve, while maintaining the essential event management functionalities. Technical challenges addressed in this project include implementing a secure authentication process, designing efficient database schemas to store user and event data, and creating a functional user interface featuring all the necessary aspects. This project demonstrates the integration of security, database management, and user interface design to create a practical solution for personal time management.
Student(s):
Chelsea Duong
Faculty Mentor:
Dr. Craig Reinhart
Investigating the Role of the Phage Shock Protein System in Stress-Induced Motility in Escherichia Coli
The phage shock protein (Psp) system is a set of proteins that function as a response mechanism in bacteria, primarily to stress conditions that damage the bacterial membrane or disrupt its normal function. This system plays a role in antibiotic resistance and virulence, but its characterization is limited, even though it exists in all three domains of life. This study investigates the connection between Psp system and motility in Escherichia coli, the purpose of which is to determine if motility is a general response to cellular stress. The psp operon consists of 5 genes, pspA-E. We plated psp mutant strains with single gene knockouts in 0.35% motility agar and incubated to determine if they showed enhanced motility.The motility of psp-deficient strains was compared to that of an isogenic wild-type strain. Our results showed that there was increased motility in 4 out of 5 mutant strains. These findings suggest that membrane stress can cause increased motility. Paired with findings on hypermotility in efflux pump-deficient strains, further research can assess whether there is a connection between motility, metabolism, and efflux pump function. Through this work, we hope to clarify the causes of hypermotility in efflux pump deleted mutants, and determine the role of the psp operon in efflux pump assembly and function.
Student(s):
Ashley Escamilla
Faculty Mentor:
Dr. Dana Harmon
A 225-Year Retrospective Analysis and Future Risk Assesment of Earthquakes in the Ventura-Santa Barbara Region
The Ventura - Santa Barbara region is one of California's most seismically active areas, characterized by a complex geomorphological setting of highly active thrust faults that have shaped the east-west Transverse Ranges. This region has a long history of seismic activity, including large M7.0+ earthquakes, triggering subsequent hazards such as mass wasting events, which historically have also resulted in regional tsunamis. This study consolidates recent and historical scientific data, as well as accounts from local Indigenous peoples and Spanish missionaries, to assess seismic hazards. Using historical data and USGS projections, a data analysis was conducted to better understand historic seismicity, potential faults of high concern, and future earthquake scenarios. Data analysis was performed using R programming in Collab constructed by Anthony D. Salas, which focused on the distribution, frequency, potential fault origins, and locations of earthquakes in the region since 1800. These findings were then compared to historical seismic events to identify seismic “hotspots” and assess future risks. A review and analysis of multiple high-risk fault systems in the Ventura-Santa Barbara region was conducted alongside the examination of historical seismic activity to more precisely assess earthquake potential and narrow realistic future earthquake scenarios. Due to a lack of reliable records prior to 1800, many seismic events and their associated hazards remain poorly constrained, underscoring the need for further paleoseismic research to enhance our understanding of the region’s seismic history and the extent of existing hazards.
Student(s):
Cameron Fetter, Anthony D. Salas
Faculty Mentor:
Dr. Megan Fung
Is Male Seasonal Fattening in Squirrel Monkeys (Saimiri Collinsi) Associated with Male-Male Competition?
Sexual dimorphism describes differences between males and females within a species. Many primates show body size dimorphism, which can help with male-male competition. Adult male squirrel monkeys (genus Saimiri) undergo a hormonally-induced “fattening” in their 8-week breeding season. Possibly, male fattening is a sexually-selected trait that allows fatter males to have higher access to mates through competition. This study examined whether the male fattening trait is linked to advantages during male-male competition events. We hypothesized that "fatter" males would gain more access to females due to higher competitive abilities. We collected behavioral data on two social groups of squirrel monkeys, during two mating seasons (2023 & 2024; N=272 contact hours), in Eastern Amazonia, Brazil. Groups averaged 33 individuals, nine were adult males. Males were classified visually from 1 to 3 (1=least fat, 3=most fat). Male focal samples were used to construct nearest neighbor budgets, while recording: the initiator, recipient and winner of male-male contests, along with copulation occurrences. Grade 3 males were responsible for 90% of the copulations (N=20). During male-male contests, females were within 3 meters 69% of the time (N=93 observations). Grade 3 males won 71% of competitive interactions. Grade 2 males received the most aggression; grade 1 males did not win any fights. These results show that fatter males are more competitively dominant and gain more access to females, suggesting that the degree of fattening of male S. collinsi affects their competitive ability and mating success. Funded by NSF (Award no. BCS-2140666).
Student(s):
Bella Fuentes, Vanessa Lopez
Faculty Mentor:
Dr. Anita Stone
CLU Women’s Soccer Performance Analysis
This project explores the integration of data science and multimedia in sports analytics by analyzing player performance data from the CLU women’s soccer team using both R and Python. The primary goal is to evaluate and compare the effectiveness of each programming language in visualizing and interpreting key performance metrics such as top speed, distance covered, sprint distance, energy expenditure, and player load. By leveraging data collected from PlayerTek tracking devices, this project aims to enhance the understanding of individual and team performance trends. Through statistical analysis and interactive visualizations, the study will assess which tools provide the most effective insights for coaches and players. The project will also address challenges in data handling, visualization techniques, and interactivity, while exploring the strengths and limitations of R and Python in sports data analysis. Ultimately, this research will contribute to the broader conversation on how data-driven decision-making can improve athletic performance and training strategies.
Student(s):
Brissa Garcia-Sandoval
Faculty Mentor:
Dr. Craig Reinhart
Machine Learning for Detecting SUSY Events in CERN Collision Data
In the search for Supersymmetric (SUSY) top quarks produced via di-lepton channels from proton-proton collisions in the Large Hadron Collider (LHC)—the world's most powerful particle accelerator—event data was processed through an event slimming preselector. This preselection focused on the electron-electron (ee), electron-muon (eμ), and muon-muon (μμ) decay channels. After preselection, SUSY events, along with background events such as top quark pair production (TT), WZ, WW, and Drell-Yan (DY) processes, remained. To isolate potential SUSY signals, kinematic "cuts" were applied to the variables associated with each decay channel. Previous work in this project focused on the ee channel, while our summer research expanded this effort by introducing cuts for the eμ, μμ channels, as well as implementing MET filters to further refine the dataset. These refined event selections were then used to train and optimize a machine learning algorithm designed to distinguish SUSY events from background processes. Despite progress, significant work remains before results can be published. Ongoing efforts include studying sources of systematic error in signal identification by the neural network and continuing to implement jet energy scale corrections. The ultimate goal of this research is to find evidence of SUSY, a crucial step toward unifying the fundamental forces of nature and identifying the nature of dark matter, which remains one of the most significant mysteries in modern physics.
Student(s):
Carys Garvey, Khusanjon Bobokhojaev
Faculty Mentor:
Dr. Sebastian Carron Montero
Variation in Health of Juvenile Hawaiian Songbirds Across Elevation and Comparisons with Adults
Stressors can have negative effects on the health of wildlife. During the last century, many Hawaiian honeycreeper bird species have seen a drastic decline in populations due to the introduction of avian malaria, particularly at low elevations. Avian malaria is prevalent in lower elevation habitats and virtually absent at higher elevations. Remarkably, the Hawaii Amakihi (Chlorodrepanis virens) honeycreeper species has been able to maintain population levels at high and low elevations, and evidence suggests that low elevation populations have evolved an increased ability to cope with avian malaria. While these populations are able to survive infection, they may experience negative effects related to chronic infections and/or increased immune function. These consequences remain poorly understood in Amakihi, particularly in juveniles. We examined measures of health in juvenile Amakihi across elevation and in comparison with adults. We predicted that individuals living at lower elevations would have lower overall health than higher elevation birds because the latter do not experience stressors related to avian malaria infection. To quantify health, we measured body condition, heterophil to lymphocyte ratio (H:L), hematocrit, and stress hormone (corticosterone) levels. As expected, hematocrit was significantly higher in individuals at higher than lower elevations (p = 0.0126). Surprisingly, body condition was significantly better in birds at lower than higher elevations (p = 0.0003). Our results suggest that negative effects related to avian malaria are trait specific, and further investigations will be necessary to determine if the relationship between hematocrit and avian malaria prevalence is causational.
Student(s):
Faith Imber, Brittany Perez, Dr. Gabrielle Names
Faculty Mentor:
Dr. Gabrielle Names
Chasing a Substrate Complex for NIS Synthetase DesD
NIS synthetases are promising drug targets due to their role in siderophore synthesis and their increasing association with virulence in bacteria such as Staphylococcus aureus and Bacillus anthracis. However, structure-based drug design has been hampered by statistical disorder in the oligomeric crystal structures, resulting in key interactions remaining unclear and limiting inhibitor design. Streptomyces coelicolor, a soil organism producing NIS siderophores, contains model NIS synthetase DesD, a key enzyme in this pathway, but its full structural interaction with substrates is incompletely described. This study attempted to solve a model of the DesD-substrate complex structure, utilizing a key variant preventing catalytic turnover, predicting that this will reveal critical molecular interactions for inhibitor design. We co-crystallized the complex with additional 3 mM soaking, collected X-ray crystallographic data, solved and refined the structure. Since our results unexpectedly revealed no ATP or intermediate in the binding site, we analyzed the unusually high kinetic turnover of the variant with limiting ATP and suggest that the variant, rather than inhibiting turnover, is resulting in possible ATP autolysis. Future efforts will focus on maintaining Glu467's binding role while refining the kinetic function through more conservative mutations to Gln rather than Ala.
Student(s):
Smitha Janet Joseph, Dr. Katherine Hoffmann
Faculty Mentor:
Dr. Katherine Hoffmann
Building Generalizable and Transparent Neural Networks for AccurateChemical Toxicity Prediction
This research project focuses on developing transparent and generalizable neural networks for accurate chemical toxicity prediction using MATLAB and Python. By leveraging an expanded dataset of over 50 xenobiotics and reactive metabolites, the study addresses the limitations of commercial "black box" models, which often lack interpretability and struggle with diverse chemical groups. The enhanced framework incorporates automated data filtering, MinMax normalization, and modular neural networks with three hidden layers. MATLAB employs Levenberg-Marquardt optimization, while Python utilizes Keras/Adam. Results show that Python models achieve slightly better generalizability (0.91 vs. 0.89), while MATLAB trains 30% faster, with both outperforming commercial tools in transparency and achieving low MSEs (near 1e-9). The study provides accessible open-source tools and visualizations, enabling researchers to predict toxicity across a wide range of chemical structures. This work bridges the gap between performance and interpretability, offering a valuable resource for future environmental and biological analyses
Student(s):
Yassine Kefi
Faculty Mentor:
Dr. Grady Hanrahan
Mutual Fund Similarity Through Graph Machine Learning
Given the rapid increase in mutual funds and ETFs, understanding fund similarities is essential for tailored investment strategies, effective risk management, and informed decision-making. Traditional methods of quantifying fund similarity often fall short due to subjectivity and the complexity of nonlinear relationships among assets.
Introducing Fund2Vec: a graph machine learning approach to evaluate mutual fund similarity. By representing funds and assets as a weighted bipartite graph and using the node2vec algorithm, we gain a nuanced understanding of fund similarities. The previous authors used a k-means clustering approach to tune the hyperparameters of the embedding. We proposed to replace k-means clustering with Gaussian model-based clustering for better separation of funds and assets in the embedding space. Our method outperforms traditional k-means clustering in identifying anomalous funds, aiding in risk management and investment strategy.
Introducing Fund2Vec: a graph machine learning approach to evaluate mutual fund similarity. By representing funds and assets as a weighted bipartite graph and using the node2vec algorithm, we gain a nuanced understanding of fund similarities. The previous authors used a k-means clustering approach to tune the hyperparameters of the embedding. We proposed to replace k-means clustering with Gaussian model-based clustering for better separation of funds and assets in the embedding space. Our method outperforms traditional k-means clustering in identifying anomalous funds, aiding in risk management and investment strategy.
Student(s):
Binderiya Khurtsbaatar, Dr. Christopher Brown
Faculty Mentor:
Dr. Christopher Brown
Uncovering Efflux Regulators that Control the Expression of the EmrAB-TolC
The EmrAB-TolC efflux pump found in Escherichia coli is involved in the removal of toxic molecules and regulation of cellular processes. Transcription of emrA and emrB is controlled by EmrR, and is cotranscribed with emrA and emrB in the emrRAB operon. It is not known if transcriptional regulators involved in regulating the expression of other efflux pumps also control the expression of emrRAB. We will investigate four regulators – SoxR, SoxS, Rob, and AcrS – for their role in controlling emrRAB expression. These regulators also control the expression of acrAB, the main efflux pump in E. coli. This study aims to characterize the EmrAB-TolC regulatory network and determine if there is overlap with AcrAB-TolC. Using an emrRAB promoter gfp fusion construct (emrRABp-gfp), we will measure the expression of emrRAB and measure the promoter activity when soxS, soxR, rob,and acrS are deleted. To this end, we’ve acquired strains from the KEIO collection, a library of E. coli strains containing deletion mutants of these regulators, replaced by a kanamycin resistance cassette, which must be removed to facilitate transformation with a reporter plasmid carrying the same resistance. We will transform our markerless mutants with a plasmid containing the emrRAB-gfp fusion. In total, we have cured the kanR gene from all four KEIO mutants, creating the strains ∆soxS, ∆soxR, ∆rob, and∆acrS, and will transform these strains with our emrRABp-gfp reporter construct. This study will provide insight into the roles of these regulators in EmrAB-TolC expression and in antibiotic resistance and metabolism.
Student(s):
Takoda Lakpour, Dr. Dana Harmon, Raphael Poveda
Faculty Mentor:
Dr. Dana Harmon
Metabolic Elucidation and GC-MS Characterization of Green Tea Extract for Catechin Identification
Green tea extract is widely studied for its natural antioxidant properties, primarily due to its catechin content. This study aimed to evaluate the extract's chemical profile and identify specific catechins using gas chromatography–mass spectrometry (GC-MS) after solvent extraction with hexane, methanol, toluene, and ethanol. Across all solvent conditions, only one catechin—((2R,3R)-2-(3,4,5-trihydroxyphenyl)-3,4-dihydro-2H-1-benzopyran-3,5,7-triol)—was detected with a quality match of 35. Mass spectrometric analysis confirmed the compound’s identity and demonstrated consistent extractability regardless of solvent. These findings suggest limited catechin recovery under the tested conditions and support future optimization of extraction protocols for natural product profiling.
Student(s):
Luke Larson, Bahar Keyvani
Faculty Mentor:
Dr. Grady Hanrahan
The Role of EmrAB-TolC in Oxidative Stress in Escherichia Coli
Efflux pumps found in Gram-negative bacteria have been implicated in resistance toantibiotics and other toxic molecules. Recent studies in Escherichia coli have alsodemonstrated a role for efflux pumps in control of metabolic processes and themitigation of metabolic stresses. Previously, our lab investigated the transcriptionalchanges that occur when the EmrB subunit of the pump was deleted, resulting in anonfunctional pump. Those studies revealed that genes involved in oxidative stresswere upregulated in the emrB mutant, suggesting a role for EmrAB-TolC in mitigatingoxidative stress. To investigate if the EmrAB-TolC pump is involved in protectingbacteria from oxidative stress, we performed a minimal inhibitory concentration (MIC)assay to H 2 O 2 using our wild-type strain, ΔemrB and ΔemrR mutants at 37°C and 30°C.EmrR encodes a transcriptional repressor, that when deleted, should lead to increasedproduction of the EmrAB components of the pump. We predicted that the ΔemrB wouldhave a significantly lower MIC to both bleach and H₂O₂ reflecting its putative role inprotecting E. coli from oxidants and the ΔemrR mutant would have the opposite effectdue to overexpression of the EmrAB system. Our preliminary results show that at both30 and 37˚C we observed that the emrB mutant showed enhanced susceptibility to H 2 O 2versus the WT whereas the emrR mutant was similar to the WT. These results suggestthat EmrAB-TolC does play a role in protecting cells from oxidative stress. How exactlyEmrAB-TolC does this is unclear and will be the subject of follow up studies.
Student(s):
Alexis Lopez, Mina Ahmadi
Faculty Mentor:
Dr. Dana Harmon
Linking a Novel Transgene in the Eye of Drosophila Melanogaster to an Odd Paired Mutation
This research aims to better understand the factors that affect expression of transgenes and how they are related to mutations. It also demonstrates the possibility of safe genome modification and the effectiveness of correct expression of transgenes.
The purpose of this project is to map the transgene presented in the eye of the Drosophila Melanogaster and to discover if it is influencing other genes. Specifically, if the Racing Stripe (RS) white+ transgene is related to the odd paired mutation which disrupts embryonic development.
This project began with complementation tests to narrow down the location of the RS white+ transgene on the 3R chromosome. The complementation results identified the transgene to be located in the odd paired mutation or somewhere near it. Crosses and temperature shifts were completed to demonstrate if the RS transgene has fallen under the transcriptional factor of the odd paired.
Data is still being collected, however, preliminary results show the temperature shift at too high of temperatures does not allow the flies to escape the embryonic phase.
Student(s):
Adriana Maroney, Dr. David Marcey
Faculty Mentor:
Dr. David Marcey
Measuring variability in the Spectra of the Magnetic B-type Star HD 23478
~10% of OBA type stars (~3x Msun) exhibit strong, large-scale, stable magnetic fields, which can significantly impact both their internal dynamics and surrounding galactic environments. This study focuses on the magnetic B-type star HD 23478, whose geometry complicates the analysis of its dynamics and magnetic field. 10 archival and 18 new high-precision spectropolarimetric observations of the star were obtained using ESPaDOnS@CFHT and Narval@TBL. These data were analyzed using custom Python scripts that normalized individual elemental absorption lines, calculated their equivalent widths and corresponding errors, and plotted these against the star’s rotational phase. Our analysis focused primarily on Hα, Hβ, HeI 4713 Å, HeI 5876 Å, along with Si 4553 Å, and C 4267 Å.
Student(s):
Dylan Matthiesen-Johnson
Faculty Mentor:
Dr. Mary Oksala
Evidence of Microplastic Contamination in Fishes Off the Coast Near Channel Islands Harbor
Microplastic contamination is a global issue that is harmful to the environment, marine life and even humans. Our study focused on the potential microplastic pollution of fish in the area of Channel Islands Harbor in Ventura County. Our method was to dissect and analyze the tissues of several fishes, Scomber japonicus (Pacific mackerel) and Atherinopsis californiensis (Jacksmelt), using a dissecting microscope and Fourier transform infrared spectroscopy (FTIR) for quantification and identification of the micro- contaminants. The previously frozen fish had been caught in May of 2024 as future lobster bait four months prior to the start of this study. Our research identified a number of microplastics found deeply embedded in the tissue walls of the fish including the presence of polyacrylamides (PAMs) in the oviduct of a female Pacific mackerel and in the G.I tract of another Pacific mackerel. This study suggests that remediation of micropollutants in the marine environment is needed to protect marine species and the humans that consume them.
Student(s):
Shanna Miller, Brian Pena, M.S.
Faculty Mentor:
Dr. Robert Richards
The Effects of Fatigue on Hip, Knee, and Ankle Kinematics During Single Leg Squats
Anterior knee pain and lower extremity injury risk may be influenced by altered movement patterns during fatigue, especially in females. The purpose of this study was to examine the effects of fatigue on hip, knee, and ankle joint kinematics during single leg squats (SLS) for each lead leg. Sixteen recreationally active females (18-25 years old) completed SLS at a cadence and depth until they reached a fatigue threshold. Eight Vicon motion capture cameras collected positional data from 25-markers which Nexus software used to calculate joint angles in the sagittal and frontal planes. Repeated measures ANOVA (p<0.05) determined if fatigue status (pre-fatigue, mid-fatigue, and post-fatigue) affected hip, knee, and ankle joint kinematics (range of motion (ROM), minimum, and maximum values). Hip ROM in the sagittal plane significantly increased for both lead legs as fatigue progressed (p<0.001), due to greater maximum hip flexion. Right knee sagittal plane ROM also increased (p=0.026), as a result of less knee extension. Ankle ROM decreased significantly in both legs (p< 0.001 left; p=0.006 right), due to less maximum dorsiflexion. In the frontal plane, knee ROM increased bilaterally (p=0.005 left; p=0.004 right), while hip and ankle ROM showed no significant changes. These results suggest that fatigue alters lower limb mechanics by increasing movement at the hip and knee while reducing ankle mobility.
Student(s):
Sara Mills, Jaime Alvarado, Tyler Lindholm
Faculty Mentor:
Dr. Michele LeBlanc
Reconstructing Late Holocene Environmental Dynamics in the Carpinteria Salt Marsh Over the Last 1,200 Years
The CSM24A core from Carpinteria Salt Marsh offers a detailed paleoenvironmental record spanning approximately 1250 years, focusing on wildfire activity, storm events, and human influences. Two dates are used to construct the chronology of the core, a ¹⁴C date of 1005 CE at 2.68 m and the previously dated 1862 storm event at 0.66 m. The core is divided into five zones based on paleoenvironmental proxies. Zone 5 (775–1005 CE) shows slightly elevated wildfire activity, indicated by elevated charcoal concentrations. Zone 4 (1005-1445 CE) is characterized by increased charcoal abundance, likely related to the combination of persistence drought conditions towards the end of the Medieval Climate Anomaly and increasing population density of local Chumash communities throughout this period. Zone 3 (1455–1862 CE) reveals evidence of the 1861–1862 storm event, with pronounced peaks in magnetic susceptibility, water content, and organics, consistent with a previous study which demonstrated the presence of a high energy deposit on the marsh surface. Zone 2 (1862–1950 CE) displays a significant and protracted decline in charcoal levels, reflecting the widely recognized post-1850 California fire deficit after Spanish colonization. Increased wildfire activity since 1950 (Zone 1) is consistent with the resurgence of wildfire in southern California since the latter half of the 20th century. These findings emphasize the influence of human activities on fire regimes and highlight the importance of sediment core paleoenvironmental analysis in contextualizing the long-term impacts of climate change and human disturbance in coastal California.
Student(s):
Emily Mohammadi, Dr. Robert Dull and Anthony Delgadillo Salas
Faculty Mentor:
Dr. Robert Dull
Method Development for the Determination of Model Phenolic Compounds and Associated Oxidative Metabolites
The aim of this study was to develop an analytical method for subsequent determination of 2-methoxy-4-propylphenol and its associated metabolites. Separation and mass spectral determination of known solutions of 2-methoxy-4-propylphenol were performed using gas chromatography-mass spectroscopy (GC-MS). Adequate separation was achieved which led to acceptable calibration results. The mass-to-charge ratio was utilized for full characterization. Future work will be centered on the determination of unknown concentrations of 2-methoxy-4-propylphenol and subsequent oxidative metabolites as a result of cytochrome P450 catalyzed reactions.
Student(s):
Brenda Morales
Faculty Mentor:
Dr. Grady Hanrahan
Discovery of Microplastics in Root Vegetables
Microplastic contamination of the world’s ecosystem has been documented in many studies overthe past two decades. This study focused on the possibility that microplastics could make theirway into root vegetables that are constantly in contact with the soil. The presence ofmicroplastics in root vegetables was proven using Fourier transform infrared spectroscopy onsamples collected from three, organically grown, root vegetables from various sources. Thisstudy strongly suggests the prevalence of micro-contaminants in a common component of thehuman diet. The spectra were compared with a commercial library of standards from AgilentTechnologies. Peaks were assigned for functional group recognition. These spectra matchedmaterials such as alginic acid, polystyrene, fiberglass, and butadiene rubber. Anthropogeniccontamination of these food groups could signify a major cause of disease now and in the future.
Student(s):
Fatima Nemi Revilla
Faculty Mentor:
Dr. Robert Richards
Creating a Library of Oxidated Microplastics Using Fourier Transform Infrared Spectroscopy
Studying the structure and function of oxidized microplastics is vital to correctly identifying them in their oxidized form in the environment (beach, ocean, etc.). The current limitation of microplastic analysis is that FT IR Spectroscopy reference libraries only show the unoxidized form of these plastics, making it challenging to identify aged and weathered plastics. This study aimed to bridge that gap by creating a library of artificially oxidized polymers. We hypothesized that exposure to UV light would induce oxidative changes in the microplastics, leading to the addition of peaks in their FT IR spectra. First, we established a baseline of the three most commonly found plastics in nature–polypropylene (PP), polyethylene (PE), and polystyrene (PS)–and confirmed their spectra to current reference libraries. The samples were then subjected to three separate oxidation techniques: heat, UV, and an electrophoresis salt solution. We then took weekly scans of each sample and analyzed changes in their spectra. Our results indicate that oxidation introduced new spectra peaks and shifts, specifically adding a 1700 cm-1 carbonyl functional group. We then hypothesized an oxidation mechanism for the three plastic structures. The development of this FT IR reference library for oxidized microplastics will benefit the field-based identification research of microplastics. This has implications for our understanding of identifying which microplastics are in nature and their true origins. This study impacts further understanding of microplastic degradation pathways and improves analytical methods for environmental research.
Student(s):
Dean Olsson, Dr. Robert Richards
Faculty Mentor:
Dr. Robert Richards
A Comparative Analysis of Microplastic Pollution in an Agro-Riparian Environment Versus a Busy Marine Harbor
The presence of microplastics in the environment has become an increasing concern, mainly in regions where coastal and agricultural activities contribute to contamination. This study investigates and compares the impact and extent of microplastic and microfiber pollution in two different locations. We choose to compare the Channel Islands Harbor, a center for recreational and commercial maritime activity and the Santa Clara Riverbed, which passes through farmland right before draining into the Pacific Ocean near the Ventura Harbor. Our goal in examining sand sediment from the two sites is to identify the quantities and type of the micro contaminants found.
Sediment samples were collected at high tides at both sites using Rose Bengal stain to differentiate between synthetic and organic microfibers. Samples were then vacuum filtered to isolate the contaminants and analyzed with a Nikon dissecting microscope and Fourier transform infrared spectroscopy.
Initial results indicate that there is a greater concentration of microplastics in the Santa Clara Riverbed, suggesting that they are likely brought in by irrigation systems, textile waste, and agricultural runoff. In contrast, microfiber contamination was more prevalent in the Channel Islands Harbor, potentially due to boating activity, urban runoff and marine infrastructure. The findings draw attention to the different sources of pollution that impact inland river systems as opposed to coastal ones, highlighting the necessity of more studies to create efficient mitigation plans and legislative actions.
Sediment samples were collected at high tides at both sites using Rose Bengal stain to differentiate between synthetic and organic microfibers. Samples were then vacuum filtered to isolate the contaminants and analyzed with a Nikon dissecting microscope and Fourier transform infrared spectroscopy.
Initial results indicate that there is a greater concentration of microplastics in the Santa Clara Riverbed, suggesting that they are likely brought in by irrigation systems, textile waste, and agricultural runoff. In contrast, microfiber contamination was more prevalent in the Channel Islands Harbor, potentially due to boating activity, urban runoff and marine infrastructure. The findings draw attention to the different sources of pollution that impact inland river systems as opposed to coastal ones, highlighting the necessity of more studies to create efficient mitigation plans and legislative actions.
Student(s):
Jodie Oparanaku, Kiana
Faculty Mentor:
Dr. Robert Richards
Optimization of Microplastic Extraction from Sand Sediments
Microplastic contamination is truly an escalating environmental concern, specifically in sedimentary environments where accumulation raises risks to marine ecosystems and human health. Traditional methods for extracting microplastics, as well as microfibers, from sand sediments include mixing the sand with artificial seawater at a salinity of 35 ppt (3.5% NaCl), and then by decanting the supernatant onto filter paper. The saltwater helps with buoyancy, allowing the microplastics to remain suspended after mixing. However, it is not very convenient and improving efficiency and scalability in microplastic extraction remains a challenge. This study aims to use sand sediment samples extracted from the estuary of the Santa Clara riverbed that empties into the Pacific Ocean. We plan to compare different extraction methods for retrieving microplastics from sand sediments optimizing the microfibers/microplastic extraction by testing three different solutions, deionized (DI) water, 3.5% NaCl, and a 2% sodium dodecyl sulfate (SDS) against two agitation techniques: mechanical stirring and vibrational energy. We hypothesize that SDS, because of its ability to interact with hydrophobic particles, will enhance microplastic recovery compared to NaCl and DI water. Also, we explore vibrational energy as an alternative to stirring by using a standard laboratory vortexer. If vibrational extraction demonstrates comparable or superior efficiency, scaling up sediment processing using vibrational energy is possible. Fourier Transform Infrared (FTIR) spectroscopy will be used for polymer identification after filtering. Our results aim to optimize microplastic recovery techniques, providing a finding for more effective sediment contamination assessments and potential large-scale investigations.
Student(s):
Josie Oparanaku
Faculty Mentor:
Dr. Robert Richards
UI and AI for RetroArch Console
This project presents a user interface rework of RetroArch, coupled with the development of an AI chatbot designed to assist users by providing contextual, game-related information. The primary objective is to enhance the visual appeal of the front-end interface while significantly improving usability and overall user experience. The AI component functions as an in-game assistant, offering non-spoiler guidance by summarizing current objectives, recalling previously encountered hints or narrative elements, and responding to user inquiries. It is intended to operate as an optional, seamless companion to the core emulator experience. The overarching goal is to foster sustained user engagement. By providing convenient access to previously acquired in-game knowledge, the system aims to mitigate the loss of progress familiarity that often follows extended breaks. In doing so, it encourages players to revisit and complete long-form games that might otherwise be abandoned.
Student(s):
Maxim Orywal
Faculty Mentor:
Dr. Craig Reinhart
Sex-Specific Variation in Physiology of Juvenile Hawaiian Songbirds
Immunity is plastic and varies with many internal and external factors, including sex. Immunity can be lower in males than females, potentially due to differences in circulating immunomodulatory hormones. However, sex-specific differences in immunity remain poorly understood in juvenile wild birds. Avian malaria has negatively affected many endemic Hawaiian songbirds, but the Hawaii Amakihi is one of the only native species with tolerance to the disease. Increased tolerance may be related to increased immunity because Amakihi populations with greater tolerance of avian malaria have higher levels of some measures of innate immunity. Sex-specific differences in immunity could have important implications for individual susceptibility to infection. However, these relationships between immunity and sex remain wholly unexplored in juvenile Amakihi. The goal of our project was to determine whether immunity varies by sex in juvenile Amakihi. We hypothesized that females would have greater immunity than males due to differences in immunomodulatory hormones. To test this hypothesis, we collected blood samples from free-living juvenile Amakihi for molecular sexing and to measure innate immunity and circulating prolactin. In support of our hypothesis, the number of WBCs was significantly higher in females than males (p = 0.0375), while no sex-specific differences were observed for agglutination (p = 0.629). Prolactin levels tended to be higher in females than males (p = 0.206) and correlated positively with number of WBCs (p = 0.317), but these results were not statistically significant, suggesting that prolactin is unlikely to be the mechanism directly mediating differences in WBCs across sex.
Student(s):
Brittany Perez, Brittany Perez, Faith Imber, Elizabeth M. Schultz, Frédéric Angelier, Charline Parenteau, and Gabrielle R. Names
Faculty Mentor:
Dr. Gabrielle Names
Measuring Rotation Curves of Spiral Galaxies with DESI Year 1
We analyze the rotation curves of spiral galaxies to quantify their dark matter component and better understand the impact it has on spiral galaxy dynamics. Previous observations show that the stellar mass alone cannot account for the gravitational effects seen in rotation curves, indicating the presence of large dark matter halos surrounding these galaxies. We calculate rotational velocities from spectra observed during the first year of the Dark Energy Spectroscopic Instrument (DESI) Peculiar Velocity Survey, and model the rotation curves of those spiral galaxies that pass visual inspection. Parameterizing the rotation curves allows us to estimate the total mass of the galaxies within R26 (the radius of the 26 mag/arcsec2 isophote) We find a strong correlation between this total mass and the galaxy’s luminosity, which is to be expected from the Tully-Fisher relation and confirms the validity of our total mass calculations. Additionally, the relationship between total mass and stellar mass provide insights into the distribution of dark matter within spiral galaxies, allowing us to better constrain the properties of dark matter.
Student(s):
JJ Pimentel
Faculty Mentor:
Dr. Mary Oksala
Isolating Microplastics from Santa Clara Riverbed Sediment Using a Novel Electrophoresis Method
Microplastics are pervasive environmental pollutants that often go unnoticed due to their small size. Traditional extraction methods from sand or soil sediment involve creating a slurry in a 3.5% salt solution, allowing sediments to settle while keeping microplastics suspended. The supernatant is then decanted onto a filter to collect contaminants for analysis.
This study presents a novel extraction method using electrolysis in a solution of sand, water, and 0.04% sodium dodecyl sulfate (SDS). SDS facilitates microplastic binding and migration through an electric field. Sand sediment samples were collected from the Santa Clara Riverbed. A slurry of sand, water and SDS solution was placed in the well associated with the negative electrode of an electrophoresis box. The box was then filled with buffer solution, and electrolysis was conducted at 100 volts for 30 minutes. Afterward, the aqueous solution near the positive electrode was removed and filtered.
Results were compared to the traditional decanting method using sand from the same location. The electrolysis method extracted significantly more microplastics than the conventional approach.
This study presents a novel extraction method using electrolysis in a solution of sand, water, and 0.04% sodium dodecyl sulfate (SDS). SDS facilitates microplastic binding and migration through an electric field. Sand sediment samples were collected from the Santa Clara Riverbed. A slurry of sand, water and SDS solution was placed in the well associated with the negative electrode of an electrophoresis box. The box was then filled with buffer solution, and electrolysis was conducted at 100 volts for 30 minutes. Afterward, the aqueous solution near the positive electrode was removed and filtered.
Results were compared to the traditional decanting method using sand from the same location. The electrolysis method extracted significantly more microplastics than the conventional approach.
Student(s):
Ana Rodriguez
Faculty Mentor:
Dr. Robert Richards
Creating a 2D Card-Based Game in Unity.
The field of game design has evolved greatly over the last half-century, and now it is possible for a single person to make a video game on their own. This project takes the reader through the development of a 2D video game using state-of-the-art techniques with the Unity game development engine. The game is single-player and has card-based gameplay. In the game, the player will play cards from their hand to deal damage to an enemy. The damage the player receives comes in the form of discarding cards from their hand. The player can only heal themselves by playing cards that allow them to draw more cards. In between fights, the player will be able to buy cards to add to their deck in order to increase their odds in future battles. The player is able to buy these cards and upgrades with the in-game currency gained after defeating each enemy. After completing four battles against progressively stronger enemies, the player wins the game.
Student(s):
Colin Russell
Faculty Mentor:
Dr. Craig Reinhart
Discovery and Characterization of the FtsH RNA Thermometer
RNA thermometers are specialized regulatory elements that are used to allow adjustment of gene expression in response to changes in temperature. These RNA molecules are highly structured and located in the 5’-untranslated region (5’-UTR) of a gene. These molecular thermo-sensors are usually associated with heat-shock, cold-shock, and virulence genes (Chowdhury et al., 2006). Furthermore, these structured RNAs sequester the ribosome binding site (Shine Dalgarno) at colder temperatures. Increases in temperature result in an alternation to the structure, which allows ribosome-binding access and the initiation of translation (Righetti & Narberhaus, 2014). FtsH is a bacterial AAA+ protease that was found, in past studies, to be associated with many cellular activities such as playing a role in the regulation of the E. coli heat shock response. (Langklotz et al., 2012). Previous research conducted in Dr. Abdelsayed’s lab found FtsH to be a candidate for an RNA thermometer; however, further testing must be done to ensure the RNA thermometer is a real RNA thermometer that works. The aim of this research is to further characterize FtsH and to biochemically validate that it is an RNA thermometer.
Student(s):
Julianne Sampang
Faculty Mentor:
Dr. Robert Richards
2D Fighting Game Development Using the Unity Engine
In game development, it is common for developers to use an existing game engine to create their game, helping to speed up the development process and allow developers to focus on creating systems specific to their game. The goal of this project is to research, design, and create a two player fighting game within the Unity development environment. The primary focus is to design and develop the objects and logic that comprise the gameplay loop and mechanics of a fighting game, allowing players to compete against each other to win two out of three rounds. To achieve this goal, game objects are designed to combine components from the Unity engine with scripts written to perform actions like jumping or hit detection for attacks. The game also requires scripts for taking player inputs and determining which actions can be performed based on a character’s current state. In order to update the characters visually, the scripts also update the variables that transition from one animation to another inside Unity’s animation state machine. Lastly, one of the considerations that affected the outcome of this project is the desire to design these systems so that new functionality, features, or playable characters that could not be implemented during the course of this project could be added in the future.
Student(s):
Ian Smith
Faculty Mentor:
Dr. Craig Reinhart
Client Data Management System
This project develops a web-based client data management system that automates the collection and organization of company and employee information into exportable spreadsheets. The application solves critical inefficiencies in manual data entry by providing businesses with an intuitive platform where new clients can input structured details about their organization and workforce, which the system automatically formats and exports as CSV or Excel files. Built with a full-stack architecture, the solution features a responsive frontend (HTML/CSS, JavaScript) for seamless user interaction, a backend (Node.js) to process data and manage business logic, and a database (MySQL) for secure storage. Key functionalities include dynamic form validation to ensure data accuracy, customizable spreadsheet templates for varied use cases, and optional user authentication to protect sensitive information. Designed for small businesses and HR departments, the tool streamlines client onboarding and record-keeping while reducing human error. The project demonstrates proficiency in system design, database management, and automation, with potential for future expansion through CRM integrations or analytics dashboards. By bridging the gap between user input and structured output, this application highlights how targeted web development can optimize administrative workflows. The technical implementation prioritizes scalability, usability, and interoperability, offering a foundation for broader enterprise solutions in data management.
Student(s):
Sebastian Smith
Faculty Mentor:
Dr. Craig Reinhart
Is Male Seasonal Fattening in Squirrel Monkeys (Saimiri collinsi) a Product of Female Choice?
Female choice, a form of sexual selection, can drive the evolution of male ornaments or behaviors. Squirrel monkeys (genus Saimiri) are small, neotropical primates that live in large groups, characterized by seasonal breeding, polygamy, and female dominance. Adult males undergo hormonally induced seasonal fattening, gaining up to 20% of their body weight during the 2-month mating season. We examined the possible relationship between male fattening and female choice in Saimiri collinsi. We hypothesized that female squirrel monkeys prefer to mate and associate with more robust (“fatter”) males. We collected behavioral data on two squirrel monkey groups, in Eastern Amazonia, Brazil during two mating seasons (N= 272 contact hours). Groups averaged 33 individuals, with 9 adult males per group. Each adult male was classified visually between 1-3 (1= least fattened; 3=fattest). We collected adult focal data by documenting nearest neighbors (within 3m) and recorded male-female sociosexual interactions, noting female receptivity. Females copulated mostly with grade 3 males (90% of 20 copulations) and were more frequently observed near these males (χ2= 55.05; p<0.0001). Grade 3 males also spent more time near adult females (H=7.61, df= 2; p=0.02) and associated with more individual females than grade 1 and 2 males. Grade 3 males initiated 73% of 881 male sexual solicitations and had the highest female receptivity (78% of observations). These results indicate that fatter males have higher mating success, and that female choice likely plays a role in the evolution of male fattening in squirrel monkeys. NSF funded this research (Award no. BCS-2140666).
Student(s):
Landon Stouch
Faculty Mentor:
Dr. Anita Stone
Uncovering the Transcriptional Regulon of EmrR
EmrR is a transcriptional regulator found in Escherichia coli. EmrR is known to repress transcription of the emrRAB operon, which makes up two components of the EmrAB-TolC efflux pump. EmrAB-TolC allows E. coli to resist stresses caused by toxic molecules, including antibiotics, by extruding them from their interior. Although EmrR is known for its role in regulating the efflux response by controlling the expression of emrA and emrB, there is a possibility that EmrR could regulate additional genes related to other pathways and metabolic responses. We performed a transcriptomic study to discover potential target genes of the EmrR regulator. RNA extracted from E. coli BW25113 wild type and ΔemrR mutant cells was sequenced and analyzed to identify changes in gene expression between the two strains. Upregulation of the arginine-dependent acid resistance system genes adiA and adiC in the ΔemrR mutant indicates potential involvement of EmrR in regulating the acid resistance response. We also observed increased transcription of many genes involved in efflux and transport. This study suggests that EmrR may play a role as a master transcriptional regulator for multiple transport systems.
Student(s):
Elisha Tong
Faculty Mentor:
Dr. Dana Harmon
Effect of Load on Lower Extremity Joint Torques in Single Leg Squats
Joint torques provide information regarding the net contribution of each joint during a movement, which is useful in training and rehabilitation settings. This study investigated the effect of load on joint torques during right lead leg single-leg squats (SLS). Fourteen injury-free males, ages 18-25, who regularly performed lower body resistance training and were experienced in SLS, completed sets of SLS with various loads. A one-repetition maximum (1RM) test was conducted in a previous session to determine each subject’s 1RM for a right leg single-leg squat. Each participant stepped on a force plate with their right leg, with their left leg supported behind them, and completed five repetitions with no load (0% 1RM) and 40% 1RM and three repetitions with 80% 1RM. An eight-camera Vicon motion capture system (200 Hz) was used to collect data from 25 reflective markers, while a force plate (1000 Hz) recorded the net ground reaction force on the lead leg. Joint torques were calculated using Nexus software, averaged over repetitions for the concentric phase, and then were normalized for both body mass and total mass (body + load). The effect of load on joint torque values in the sagittal and frontal planes for the right hip, knee, and ankle was determined with a repeated measures ANOVA through SPSS software (p < 0.05).
Student(s):
Daniel Trounday, Connor Dominici
Faculty Mentor:
Dr. Michele Leblanc
Machine Learning in Tetris
Reinforcement learning, specifically DQN models, have been created to play games from the 1970s and 1980s, such as Space Invaders (Mnih et al., 2013). This study attempts to create a DQN model to play the game of Tetris and reach a score of at least 1 million points. In this model, the Q-learning algorithm was used in tandem with a deep neural network to maximize the rewards, in this case points, obtained from interacting with the Tetris environment. Every time a new piece was chosen by a randomizer, the agent chose an action to take by ranking all possible moves based on the rewards gained from the resulting state observation. This was done by passing all state observations through a neural network with two hidden layers each with 42 artificial neurons and comparing the ranks from the single node output layer. Each state observation consists of five features which describe the maximum height reached on the board, the minimum height on the board, the number of holes between placed pieces, the number of filled rows, and the difference in adjacent column heights. After successive training sessions and iterative adjustments to gamma and to other learning constants, training stopped at a loss of 185 and a high score of 5.265 billion points, played in real time with the use of the hold move, where a secondary piece is evaluated alongside the main piece. Results suggest that the agent gained a fundamental understanding of the complexities of Tetris, mirroring other studies’ results.
Student(s):
Eric Tuesta
Faculty Mentor:
Dr. Chang-Shyh Peng
Investigation and Analysis of Quantity of Microfibers Found in Crassostrea virginica (Atlantic Oyster)
Microfiber pollution in marine environments is a significant concern due to its ecological and health implications. Synthetic microfibers are a major source of marine contamination, with an estimated 13 million tons of synthetic fabric waste entering oceans annually (Mishra et al.). Crassostrea virginica (Eastern oyster), a commercially and ecologically important bivalve native to the Atlantic coast, is particularly susceptible to microfiber contamination due to its filter-feeding behavior (Kennedy et al., 1996).
This study analyzed 200 oysters to assess microfiber contamination in relation to shell length. Oysters were processed using a standardized microfiber extraction protocol, including tissue grounding, vacuum filtration, and manual quantification under a microscope. The average shell length was 79.17 mm, with an average of 8.66 microfibers per oyster. Oysters between 7.43 mm and 9.25 mm had the highest microfiber counts, while smaller (<6.82 mm) and larger (>11.68 mm) oysters contained fewer. However, microfiber contamination showed no significant correlation with shell length (r = -0.0032, p = 0.9641), suggesting external environmental factors influence accumulation. The majority of microfibers were black (64.3%) and blue (26.5%), likely originating from textiles or fishing gear.
These findings highlight the persistent exposure of C. virginica to microfiber pollution and emphasize the need for further research on its ecological and commercial implications. Understanding microfiber contamination in marine bivalves is critical for assessing the broader environmental impact of synthetic pollutants in aquatic ecosystems.
This study analyzed 200 oysters to assess microfiber contamination in relation to shell length. Oysters were processed using a standardized microfiber extraction protocol, including tissue grounding, vacuum filtration, and manual quantification under a microscope. The average shell length was 79.17 mm, with an average of 8.66 microfibers per oyster. Oysters between 7.43 mm and 9.25 mm had the highest microfiber counts, while smaller (<6.82 mm) and larger (>11.68 mm) oysters contained fewer. However, microfiber contamination showed no significant correlation with shell length (r = -0.0032, p = 0.9641), suggesting external environmental factors influence accumulation. The majority of microfibers were black (64.3%) and blue (26.5%), likely originating from textiles or fishing gear.
These findings highlight the persistent exposure of C. virginica to microfiber pollution and emphasize the need for further research on its ecological and commercial implications. Understanding microfiber contamination in marine bivalves is critical for assessing the broader environmental impact of synthetic pollutants in aquatic ecosystems.
Student(s):
Julianna Valderas
Faculty Mentor:
Dr. Andrea Huvard
Search for High DOS Features in Materials Databases
Many high-temperature superconducting materials are characterized by high density of states (DOS) near the Fermi level and strong magnetic fluctuations. In this project we developed Python code to fetch tight-binding parameters from the Jarvis materials database (https://jarvis.nist.gov/) and then calculate the DOS and bare magnetic susceptibility. Using these calculations, we hope to identify potential candidate materials, either in their pure or doped forms, having desired properties such as superconductivity or magnetism. The standardized tight-binding format used in the Jarvis database, which is derived from the Wannier-90 project (https://wannier.org), enables rapid searches through the Jarvis database and provides a demonstration for accelerating research in materials physics through standards-based modeling and simulation.
Student(s):
Hayden Williams, Dr. John Deisz
Faculty Mentor:
Dr. John Deisz
Moneymind Financial Organizer
This project focuses on the development of a financial management application that enables users to track their income and expenses efficiently. Built as an iOS application, it connects to a MySQL database through a Flask API, allowing users to input and monitor their financial data in real time. The application calculates current balances and projects future financial standings based on spending and income trends. By integrating backend database management with a seamless user interface, the system ensures accurate financial tracking and accessibility. A key aspect of this project is the interaction between the Swift application, the Flask API, and the MySQL database, which facilitates secure data retrieval and updates. The API, hosted on an Apache server, processes requests to ensure synchronization between user input and stored financial data. The project also explores challenges related to data consistency, performance optimization, and real-time updates, addressing them through efficient database queries and structured API interactions. Ultimately, this application provides users with a streamlined tool for financial organization, demonstrating the integration of mobile and backend technologies in a practical, user-centered system.
Student(s):
Micah Wisniewski
Faculty Mentor:
Dr. Craig Reinhart
Pitch Corrector
This project focuses on the development of a pitch corrector app used to automatically adjust the frequencies of a given sound to their nearest standard musical note frequency. The user is able to upload and playback a file. The wavelength of the file is displayed for visual representation. When the pitch corrector is applied the user can adjust the degree of correction and playback the processed file. The user also has the option to select a scale or choose to omit certain notes from being shifted to. The shifted wavelength is displayed in comparison to the original and the processed file is made available to be downloaded.The development of this app is used as a means to study audio software development and accompanying concepts such as Fast Fourier Transform (FFT) and digital signal processing (DSP). It is designed to offer an intuitive way to enhance a recording.
Student(s):
Emily Woo
Faculty Mentor:
Dr. Craig Reinhart