2025 Graduate Poster Exhibition Abstracts

Abbas, Joseph– Program: Ph.D – Prevention Science

Balancing the Plate and GPA: Exploring Food Pantry Usage and Academic Performance in College Students 

Food insecurity poses significant challenges to academic success, disproportionately affecting underserved student populations and threatening equity in higher education. Although food pantry services have become a common institutional response, their impact on academic performance remains underexplored. This study investigates the relationship between food pantry usage and cumulative GPA among undergraduate students at Rutgers University–Camden, offering critical insights into the interplay of food security, student support interventions, and educational outcomes. Drawing on self-determination theory, this research proposes that food pantry services, by addressing immediate physical needs, may indirectly support psychological needs—autonomy, competence, and relatedness—thus fostering intrinsic motivation and academic persistence. This study investigates the relationship between food pantry usage and cumulative GPA among a sample of 4,270 undergraduate students at Rutgers University–Camden, including 384 students who utilized pantry services and 3,886 non-users, with non-users coded as zero for pantry usage.  Using a dataset of sample students with food pantry visit records and GPA data linked by student IDs, the study employed quadratic regression analysis to uncover a statistically significant U-shaped relationship between food pantry visits and GPA. GPA declines as food pantry usage increases between 1-5 visits, consistent with expectations that food insecurity may disrupt academic focus. However, a subset of students with regular food pantry use—between 6 and 10 visits per semester—demonstrated higher GPAs, suggesting that regular utilization of food pantry resources may reflect a resourceful adaptation enabling academic success. Although this association is not causal, it raises the intriguing possibility that encouraging consistent food pantry usage could influence academic outcomes. This nuanced relationship suggests an opportunity for targeted interventions: by encouraging consistent food pantry use, institutions may support students in developing sustainable strategies to address food insecurity. These findings, although not causal, underscore the interdisciplinary implications for psychology, public health, education, and policymaking. They highlight the importance of integrating food pantry services into holistic student support frameworks that promote GPAs while balancing accessibility with engagement strategies to address students’ academic, emotional, and social needs. Future research should explore mediators such as resilience, stress, resource utilization strategies, and longitudinal impacts to better understand the relationship between food security and academic outcomes. This study contributes to the broader discourse on equity and student success in higher education, providing actionable insights for policymakers, administrators, and psychologists. 

Keywords: Food insecurity, academic success, food pantry services, GPA, underserved students, higher education equity, self-determination theory, student support, resilience, stress, public health, education policy, interdisciplinary, holistic frameworks. 

Abi Abdi, Ahmed – Program: MS – Biology

Stemness, the ability of cells to self-renew and differentiate, is a fundamental property of stem cells that has been extensively studied in model organisms. Key regulators of stemness, such as the Yamanaka factors (Oct4, Sox2, Klf4, c-Myc), have been shown to reprogram somatic cells into pluripotent stem cells, highlighting the conserved molecular mechanisms underlying cellular plasticity (Takahashi & Yamanaka, 2006). In early-diverging metazoans like Cnidarians (e.g., Hydra and Nematostella), conserved signaling pathways (e.g., Wnt, BMP) play critical roles in stem cell maintenance and regeneration, suggesting deep evolutionary origins of stemness (Watanabe et al., 2009; Juliano et al., 2014). Similarly, leeches (e.g., Helobdella austeniensis) have emerged as valuable non-model organisms for studying stem cell biology, with shared signaling pathways (e.g., Notch, Hedgehog) implicated in stem cell regulation across species (Srivastava et al., 2014; Özpolat & Bely, 2016).

 Despite these advances, the role of conserved protein motifs in differentially expressed genes associated with stemness remains underexplored in non-model organisms, including H. austeniensis and Cnidarians. This study aims to identify and compare these conserved motifs to uncover evolutionary conservation in stemness mechanisms. By analyzing gene expression profiles and protein sequences, we seek to elucidate how conserved molecular signatures contribute to stem cell function and regeneration across phylogenetically diverse species.

 To visualize the developmental context of stemness in H. austeniensis, we present a schematic cell lineage diagram tracing cellular progression from early developmental stages to stage 4c. This diagram highlights key transitions and potential stem cell populations, providing a framework for understanding the cellular basis of regeneration and differentiation in leeches.

 Our findings will contribute to a deeper understanding of the evolutionary conservation of stemness mechanisms and provide insights into the molecular basis of cellular plasticity across metazoans. This research has implications for regenerative medicine, evolutionary developmental biology, and the study of stem cell biology in non-model organisms.

Aborn, Tembisa – Program: MFA – Creative Writing

This spring, I have been invited by Barrelhouse Magazine to present and read at AWP a piece of creative work I have generated in the course of my studies thus far. For this piece, my purpose is to not only demonstrate the creative skills I have honed in my first two terms at Rutgers-Camden, but also to use the opportunity to connect with writers and professionals in the creative writing space.

Adams, Evan – Program: MS – Data Science

 Alaska’s salmon industry is a vital economic and ecological mainstay for the state, contributing approximately 80% of North America’s wild caught salmon and serving as a major source of employment and revenue. However, the industry faces growing challenges from climate change, economic pressures, and international competition. This study investigates the correlation between environmental factors such as rising ocean temperatures, sea level changes, and ocean acidification and the quantity and quality of salmon caught in Alaska. 

 The null hypothesis (H₀) assumes no significant relationship between these climate variables and salmon populations, while the alternative hypothesis (H₁) suggests that climate change has a measurable impact, affecting salmon migration, spawning behaviors, and overall catch potential. 

 Additionally, this research explores the economic implications of these changes, labor costs, fuel prices, and global market competition. By analyzing historical environmental data, salmon population metrics, and economic trends, this study aims to provide insights into the sustainability of the Alaskan salmon industry amid evolving climate conditions. The findings may inform policy decisions, conservation efforts, and adaptive strategies for industry stakeholders.

Adebajo, Olorunfunmi – Program: MPA –  Public Administration

Prevention as Cure: Leveraging Shared Housing to Disrupt Intergenerational Homelessness in Camden, New Jersey

Camden, NJ, faces a persistent homelessness crisis exacerbated by soaring rental costs and limited affordable housing, rendering costly emergency interventions unsustainable. The 2024 Annual Point-In-Time Count reveals 743 persons experiencing homelessness, a 24% increase in homelessness from 2023. Nearly half of Camden’s homeless population are aged 18 to 44—a vibrant group whose economic potential is undermined by this housing instability. Also, Camden City, home to Rutgers University and Cooper Hospital, contributed over 67% of this population (Monarch Housing Associates, 2024).These statistics underscore the urgent need for proactive, upstream interventions that prevent homelessness before it begins. This study examines the potential of Shared Housing as a “prevention-as-cure” solution to disrupt the cycle of intergenerational homelessness in Camden. In this model, unrelated individuals maintain private living spaces while sharing common areas, thereby pooling resources to substantially reduce per-person housing costs—a critical strategy in a market where median rental prices have surged by $450 against a modest household income increase of $175 (Camden County, NJ – Data USA). Beyond cost efficiency, Shared Housing provides vital psychosocial support, reducing social isolation and fostering community well-being, which are essential for mitigating the destabilizing effects of homelessness (Henwood et al., 2015; Tsemberis, 2010; Johnson et al., 2018; Byrne et al., 2021). The U.S. Department of Housing and Urban Development (2022) estimates that Shared Housing programs could save up to 30% in housing assistance costs compared to individual subsidies. Leveraging promising models implemented by local organizations such as Senior Citizens United Community Services (SCUCS) and Volunteers of America (VOA), this research integrates a strategic Camden County Landlord Engagement Program into the Shared Housing framework. This program aims to transform property owners into proactive partners through targeted education, financial incentives, and robust legal protections. A centralized coordination office will streamline resource allocation and communication, expanding Shared Housing options and advancing Camden County’s goal of achieving functional zero homelessness by 2030. By shifting the focus from reactive emergency responses to preventive, community-based solutions, this research presents actionable policy recommendations that harness Camden’s existing housing stock to secure long-term stability for its most vulnerable populations.

Agbo Eje, Ojobo – Program: MS – Data Science

Explainable Artificial Intelligence Applications in Healthcare: A Systematic Review

Artificial Intelligence (AI) has demonstrated potential in healthcare in recent years, with significant adoption in areas like disease diagnosis and medical imaging analysis, clinical decision support systems (CDSS), and personalized medicine. In most cases, the nature of “black-box” AI models poses significant challenges related to trust and interpretability, hindering their clinical adoption. Explainable Artificial Intelligence (XAI) is a field of research that seeks to close this gap by providing transparency into these AI model decision-making processes, which can improve AI adoption. This study presents a systematic review of previous works on the applications of XAI techniques in healthcare, categorizing methods into model-agnostic (e.g., SHAP (Shapley Additive Explanations), LIME (Local Interpretable Model-Agnostic Explanations), and Grad-CAM (Gradient-weighted Class Activation Mapping)) and inherently interpretable approaches (e.g., decision trees and random forest). These methods are evaluated across multiple healthcare domains, assessing their accuracy, interpretability, and clinician usability. Findings show that while XAI methods improve model interpretability, they are not always intuitive for healthcare professionals. Also, challenges such as lack of standardized XAI evaluation metrics, biases in AI-driven diagnoses, and limited real-world clinical validation of XAI remain key barriers to AI adoption in healthcare. To improve the adoption of XAI applications in healthcare, the study proposes further research roadmap for hybrid AI models, human-centered XAI interfaces, and standardized evaluation frameworks to enhance model transparency and clinical usability. This research will foster trust, accountability, and broader adoption of AI-driven healthcare solutions by advancing explainability in medical AI applications. Keywords: Explainable Artificial Intelligence, Healthcare, Medical Imaging, AI Transparency, Trustworthy AI, Model Interpretability, Regulatory Compliance, Human-Centered AI

Al Juaid, Fawaz – Program: MA – Criminal Justice

The research explores criminal activity in Philadelphia from 2019 to 2023. We used 750,463 records. We discovered some important aspects through the exploratory stage of the data, time series analysis, and crime types analysis. Crimes from 2019 to 2020 decreased by 17% but rose strongly in 2023. Also, the data revealed seasonal patterns with high crime rates in summer and lower in winter. Moreover, from 2021, there was a rise in carjacking, which shows continuous security issues. Drug crimes witnessed a decrease, which is attributed to shifts in decriminalization and policies reshaping. However, vandalism and property-related crimes remain an ongoing problem. We conclude from these trends that more concentrated measures must be taken, for instance, installing cameras in residential neighborhoods, establishing community police programs, and taking preventive measures in affected areas. This analysis may be useful for decision-makers and police to improve safety in the city.

Allen, Jasmaine – Program: Ph.D – Prevention Science

 Undergraduate student retention remains a significant challenge, particularly among first-generation students, who comprise 56% of college enrollees but have lower graduation rates than their continuing-generation peers. Systemic barriers such as financial strain, lack of family guidance, and limited academic preparedness contribute to their attrition. Universities have implemented programs like Summer Bridge Programs (SBPs) and first-year initiatives as interventions to support student persistence. However, research on the effectiveness of condensed SBPs remains limited, particularly concerning first-generation students.

 This study evaluated the student success outcomes of a four-day SBP for freshmen students at Rutgers University-Camden, measuring student sense of belonging, academic self-efficacy, and relationships with success coaches. Using a quantitative approach, the study surveyed students on these measures, comparing participants with non-participants and first-generation with continuing-generation students. Results indicated that Summer Bridge participants reported higher levels of academic self-efficacy and sense of belonging, with first-generation students benefiting the most. While statistical significance was not observed for all measures, early coach introductions were significantly associated with higher student trust. Additionally, participation in SBPs correlated with increased campus involvement. Findings suggest that even short-term interventions can positively influence first-generation student outcomes, though further research with larger samples is needed. This study highlights the potential of condensed SBPs in fostering student success and informs future institutional strategies for improving retention and persistence.

Alonso, Elizabeth – Program: MA – Psychology 

Although survival rates for pediatric hematopoietic stem cell transplant (HSCT) patients have improved since its first utilization in 1957, symptom burden and quality of life concerns are significant during the post-transplant hospitalization period and into survivorship. Sleep quality is also impacted in this population, especially during hospitalization due to the artificial environment and frequent room entries. There is a need for additional research on symptom burden in patients undergoing HSCT and on the relationship of the hospital environment to sleep. There is also the possibility that symptom burden could impact patient engagement with daytime supportive care. This secondary analysis study aimed to describe symptom burden and supportive care use during the peri-transplant period and explore the effect of extending nighttime vital sign checks (EVs) to allow for a 6-hour block of sleep. This study utilized the PROMIS sleep short forms, PedsQL-Stem Cell Transplant Module, and medical chart review to assess symptom burden and supportive care use. Findings indicate that general symptom burden, as well as the severity for most individual symptoms, peaks on Day +5. Additionally, hierarchical linear modeling showed that greater sleep-related impairment predicted decreased supportive care use, and that greater supportive care use predicted reduced sleep disturbance. The EV intervention did not influence general symptom burden nor supportive care use; however, there is preliminary evidence that the intervention did reduce sleep disturbance. These findings have implications for clinical practice. Healthcare teams should monitor patients during times associated with greater symptom burden and encourage supportive care engagement to alleviate or provide distraction from symptom burden. Future studies should continue to explore the connection between symptom burden, sleep, and supportive care use, as well as investigate interventions to mitigate disturbed sleep during hospitalization.

Alvarez Outerino, Elisa – Program: MAT- Teaching Spanish

 Aspectual Development in the Online-Asynchronous L2 Classroom through Concept-Based Instruction

 Framed within a Sociocultural Theory approach to Second Language Development (Lantolf, Poehner & Swain, 2018), this proposal examines the implementation of Concept-Based Instruction (CBI) to the teaching and learning of Spanish preterite and imperfect tenses in a remote-asynchronous second semester university Spanish course. The way this course was previously designed embraced traditional structural models to the teaching and learning of Spanish, which tend to approach preterite and imperfect in a piecemeal and taxonomic fashion by which students are asked to rely on prescriptive rules and rote memorization to support their language use (Negueruela-Azarola, 2008). 

 To address this issue and to support online instruction in our elementary Spanish sequence, we attempted to redesign our online language curriculum by avoiding the previously mentioned structural way to teach preterite and imperfect prevailing in our current program, and implemented a concept-based approach to develop aspectual awareness and control over Spanish preterite and imperfect in a way that was more conducive towards L2 learner conceptual development (Negueruela & García, 2016). CBI moves away from memorization and repetition and supports learners in reflecting on their communicative choices by using concepts with the intent of constructing new understandings (García, 2018). 

 Using Vygotsky’s (1986) distinction between scientific and everyday concepts applied to L2 development (Negueruela, 2008), this poster showcases a pedagogical implementation of Concept-Based Instruction to promote L2 learner development in a remote-asynchronous context. We also examine the effectiveness of this model to the teaching and learning of the grammatical concept of aspect in a second-semester Spanish course (n=8 learners) over a two-week asynchronous sequence. Learners’ emerging L2 conceptual development is analyzed through multiple sets of developmental data, including conceptual awareness, morphological and aspectual accuracy through written narrations and cloze activities, verbalizations, personal reflection data, scaffolded activities, and the creation of personal conceptual schemas that encouraged students to manipulate and internalize conceptual distinctions rather than memorize rules. Preliminary results indicate that learners’ conscious manipulation of concepts during activity, promoted the development of a more sophisticated semantic understanding of conceptual categories, mediating in the internalization of L2 grammatical concepts such as aspect. Findings from this study indicate learners’ increased precision in the use of preterite/imperfect both morphologically and conceptually, suggesting that an asynchronous pedagogical model based on sociocultural principles can effectively promote the development of conceptual categories of meaning in novice L2 learners of Spanish.

Amonga, Brian – Program: MS – Data Science

Misinformation and DeepFake Detection on Social Media 

The authenticity of digital information is increasingly becoming a contentious issue in recent years due to the proliferation of AI generated content aimed at misleading the public, creating chaos and confusion, and tarnishing reputations of individuals and organizations. Detection of misinformation and deepfakes calls for a multifaceted approach and frequent upscaling of the existing robustness testing techniques, considering that techniques used to generate the fabricated texts and images are becoming more smarter than the guardrails. This study investigates the reliability of deep learning models in detecting misleading text and fake images by leveraging on different machine learning techniques. For instance, the study utilizes feature-based approach and deep convolutional neural networks (CNN) such as EfficientNet, ResNet-50, and XceptionNet to classify images. In feature-based classification, the classifier extracts statistical attributes such as color and pixel-variations while the CNN learns from hierarchical patterns differentiating real from synthetic media. For text classification, the study introduces smooth labels consisting of a truth score between 0 and 1, which replaces the binary labels improving generalization. 

The study takes a standard approach that starts with data collection and preprocessing.  In the initial training stage, the Liar Dataset for text analysis and Cifake dataset, generated from the popular CIFAR-10 dataset for image-based classification were utilized. Initial training runs play a pivotal role of evaluating the model performance, while explainability helps in understanding the technicalities behind the decision-making processes. To ensure high quality results, hyperparameter tuning and data augmentation techniques are incorporated prior to the final training. Exporting the trained models in ONNX format paves way for other processes such as robustness testing against adversarial perturbations to explore their ability to out of distribution samples. Adversarial attacks and unreliability in handling out of the distribution samples are primary obstacles that hinder the effective performance of the deep CNNs. Generally, the study emphasizes the significance of hybrid strategies with the potential of processing text and images datasets simultaneously to address the proliferation of deepfakes and misinformation. Insights from the study contribute to the existing literature aimed at developing advanced AI filtering systems, boosting efforts in combating digital misinformation globally. 

Annapureddy, Baladileep Reddy – Program: MS- Biology

3’UTR of D. virilis gruken gene is disfunction in D. melanogaster

In Drosophila egg chamber, the TGF-alpha-ligand Gurken (GRK) instructs axes formation through the epidermal growth factor receptor (EGFR) activation. Research from Yakoby Lab demonstrated that grk from D. willistoni rescues D. melanogaster flies null for grk. At the same time, the grk from D. virilis, a more distant species, could not rescue a D. melanogaster null for grk. Preliminary data indicate that while the D. virilis grk mRNA is properly localized around the oocyte nucleus, GRK protein was absent or found at low levels. Hence, we hypothesize that D. virilis grk mRNA inefficiently translates in D. melanogaster. The 3’UTR of grk was shown to play an important role in mRNA translation. Specifically, the oocyte nucleus localized grk gets polyadenylated prior to translation. Using CRISPR/Cas9, we investigate the role of the 3’UTR of D. virilis grk for translation. When the D. virilis grk vector was built, most of the 3’UTR of D. melanogaster remained, which generated transgenic flies containing two 3’UTRs. In addition, the transgenic fly containing the D. melanogaster grk with the 3’UTR of D. virilis also contained the same portion of an additional D. melanogaster 3’UTR. Surprisingly, the latter has eggshells like the wild type eggs. Hence, we hypothesize that D. melanogaster 3’UTR is being used for the poly-A tail in those lines. Using a cDNA, we have shown that in all cased that we could amplify the 3’ end of grk, the D. melanogaster 3’UTR was detected with different portions of the D. virilis 3’UTR. We are generating new transgenic lines containing only the 3’UTR of D. virilis to then circle back to the original hypothesis regarding the efficiency of this 3’UTR in grk translation.

Bhandarkar, Athang Sanjeev – Program: MS – Data Science

 The U.S. airline industry has undergone significant transformations over the past two decades, with low-cost and regional carriers playing a pivotal role in shaping market dynamics. This study examines the evolution of market share and financial resilience of eight key airlines—Frontier, Allegiant, Spirit, Sun Country, Hawaiian, SkyWest, Alaska, and JetBlue—between 2003 and 2023 using quarterly data. Our analysis integrates market share trends, financial recovery patterns, and competitive clustering to provide a holistic view of industry shifts.

 We employ data-driven methodologies, including time series analysis, economic event correlation, and clustering techniques, to evaluate airline growth trajectories, financial robustness, and strategic positioning. The study further explores the role of geographic expansion, identifying key cities and regions that have contributed to airline success. By comparing the resilience of low-cost and regional carriers, we highlight differences in financial recovery speed and profitability trends across economic cycles.

 Through competitor clustering, we classify airlines based on business performance, uncovering hybrid models that blend low-cost and regional strategies. The study provides critical insights for airline executives, policymakers, and industry analysts, offering data-backed recommendations for future strategic planning. As the industry continues to evolve, understanding these historical patterns will be crucial for anticipating future growth and identifying regions poised for airline expansion.

Bonell, Melanie – Program: MS – Data Science

Spatial Analysis and GIS Mapping: Uncovering Crime Patterns in New Jersey Communities 

In the United States specifically, crime rates have been known to have many influences such as socioeconomic conditions, law enforcement strategies, and societal influences. The dataset used for this project is from the State of New Jersey Department of Law and Public Safety for defendant charges using adults aged 18 years and older that overlooks every county in the state of New Jersey. This project aims to answer the question of whether there are geographic patterns of criminal activity across New Jersey’s counties. The goal is to identify crime hotspots and trends of the crimes being committed per county using geographic information system tools. Another question outlined in this project is whether crime density varies between different counties, and if more densely populated counties have higher crime rates than less densely populated counties. If so, what specific types of crimes are committed in those counties that have higher crime rates. The likelihood of reoffending in those counties of higher crime rates will also be analyzed within the dataset versus those of lower crime rates. Early predictions of where crime hotspots are can aid in early intervention to prevent more crimes from happening in these areas of high crime. The study analyzed different demographic groups and age groups as key variables from all counties in New Jersey to see whether or not their background also was a key part of the different crimes being committed, and if they were jailed for those crimes or not. By integrating, the predictive portion of the model using GIS techniques, we can eliminate more crime by counties hotspots.  

Bregaudit, Duessa Red – Program: MS – Chemistry

Spermine-Mediated Assembly of DNA Nanoparticles

Polyamines, such as spermine, have been shown to aid immune response pathways like cGAS-STING by condensing viral DNA for optimal recognition. However, their interactions with DNA nanostructures remain underexplored. This study investigates how spermine influences DNA particle assembly, leveraging its positive charge to reduce electrostatic repulsion within DNA’s negatively charged phosphate backbone, potentially facilitating aggregation. Dynamic Light Scattering (DLS) is used to assess particle formation at varying spermine concentrations and evaluate the stability of these structures after spermine removal. Preliminary results indicate that spermine promotes the aggregation of Tetrahedral DNA Nanostructures (TDNs). At the same time, ongoing studies aim to determine its effects on other DNA nanostructures, including Poly(T) single-stranded DNA (ssDNA), M13 phage DNA, and rectangular DNA origami. By examining how spermine induces assembly across different DNA nanostructures, this work aims to provide insight into the self-assembly of functional DNA nanoparticles for biomedical applications

Bryan, Chrisann – Program: Ph.D – Biology: Computational and Integrative

Chronic stress can increase inflammation and other neuropsychological changes, promoting the development of various mood disorders including anxiety and PTSD. Some individuals are more susceptible to the adverse effects of stress and develop stress-related disorders whereas others are more resilient and do not develop disorders despite experiencing comparable stress. This study aims to carefully examine the expression of pro-inflammatory nuclear factor kappa B (NF-κB) and anti-inflammatory glucocorticoid receptors (GRs) in the dentate gyrus (DG) of the hippocampus using immunohistochemistry. NF-κB is a master transcriptional regulator of inflammatory genes. Its activity in monocytes is increased by noradrenaline. In the brain, noradrenaline increases microglia reactivity and cytokine production, which are important markers of inflammation. However, it is unknown whether noradrenaline drives NF-κB mediated transcription in the brain. The hippocampus is prone to stress-induced inflammation in susceptible mice, but it is unknown if this is due to increased pro-inflammatory NF-κB or reduced anti-inflammatory GRs. Temporary increases in inflammation can be caused by acute stress, but this can be reverted quickly to baseline levels due to the anti-inflammatory effects of GRs. This is because GRs are highly effective in inhibiting NF-κB activation which occurs via protein-protein interaction within the nucleus or cytoplasm. It is well-established that chronic stress impairs the anti-inflammatory effects of glucocorticoids, causing glucocorticoid resistance. This shifts the balance toward pro-inflammatory processes, increasing the impact of stress on the immune system. Here, we show that stress increases NF-κB activity in all stressed males and females with females presenting higher levels compared to males in the absence of stress. Overall, GR levels are increased in resilient males. Together, our findings demonstrate that stress increases nuclear levels of NF-κB, which are particularly high in females, but this might be countered by stress-induced GRs in resilient males.

Co, Phoebe – Program: Ph.D – Childhood Studies

During the American colonial regime in the Philippines, “juvenile delinquency” or youth incarceration policies were proposed and institutionalized. Through analyzing archival material (e.g., government publications and newspaper articles), I intend to historicize the Philippines’ nascent youth incarceration system to illustrate how its emergence was embedded in the larger colonial project of prison imperialism. I also seek to problematize invocations of dominating and normalized constructions of childhood (e.g., childhood innocence, childhood’s apolitical guise, future citizen) that further push the multiply marginalized into subalternity and justify carcerality under the banner of welfare, benevolence, protection, nation-building, and social progress.

Dalmia, Anupam – Program: MS – Biology: Computational and Integrative

 Sphingolipid-derived ceramides are found ubiquitously in eukaryotes from fungi, to plants, to animals. By contrast, these lipids have been reported in only a small number of bacterial taxa . Despite the limited number of sphingolipid (SL)-producing bacteria, there is a great variation in acyl chain length and degree of saturation, acyl chain hydroxylation, and lipid headgroups. This structural diversity is paralleled by a wide range of physiological roles for SLs including modulation of host-microbe interactions, protection from bacteriophage , bacterial life cycle and sporulation , and microbial predation . Mechanistic studies of the physiological roles of bacterial SLs have been limited by a lack of knowledge of their biosynthetic pathway. Given their importance in human health and disease , it is not surprising that the eukaryotic SL biosynthesis pathway has been elucidated in tremendous detail . By contrast, bacteria do not appear to have homologous synthetic enzymes, with the exception of serine palmitoyltransferase (Spt) which performs the initial step of ceramide synthesis. The other hypothesis is how to figure out how the SL transporters transport SL to the outermembrane of the bacterial cell membrane.

Dibia, Meron – Program: Ph.D – Biology: Computational and Integrative

Macroterminitae, a subfamily of termites, live in arid regions with limited water resources. They also cultivate fungi species as a food source and desiccate easily at low relative humidities. Therefore, the procurement and conservation of water is very important for their survival. There is no clear understanding of the mechanisms by which they collect and conserve water. The objective of this research is to identify if the water and thermal dynamics in vapor and liquid form between the mound and its immediate environment. Our premise is that the abiotic structure(mound) not only regulates their climate needs but also facilitates vapor capture and conservation of water for the colony. In order to observe this, sensors positioned within and around a termite mound were used to collect temperature, relative humidity and soil moisture data. The preliminary results show correlation between daily fluctuations in external conditions and the corresponding thermal and moisture profiles within different sections of the termite mound.

Donthi, Shailesh – Program: MS – Data Science

A Sentiment-Driven Game Recommendation Engine for Personalized Discovery 

Game recommender systems typically rely on popularity rankings or tag-based filtering that may not capture a player’s personal preferences. My project demonstrates a sentiment-based hybrid recommender engine, incorporating Steam game reviews into recommendations to enhance recommendation quality. Through emotion-based knowledge extraction from user reviews, this system aims to provide personalized, interpretable, and contextualized recommendations. Ongoing work includes data collection using the Steam API and web scraping techniques, along with preliminary sentiment analysis on review text. 

Du, Xuanzhou – Program: MA – Psychology

Research suggests that individuals’ values, or what people consider important and meaningful in life, vary by socioeconomic status (SES; Cohen & Sherman, 2014). People from lower-status backgrounds tend to be more relationship-focused, whereas higher-status individuals prioritize individual achievement (Stephens et al., 2012; Covarrubias & Fryberg, 2015). However, less is known about how subjective and objective statuses may relate to individual values and qualitative expressions of values assessed by text analysis. Understanding more specific differences with which individual values may differ across SES can inform value-based interventions, such as value affirmation interventions, and help at-risk individuals cope with resource scarcity (Hooker et al., 2018). This study examined how subjective (perceived social status) and objective (educational background) SES may relate to the degree individuals pursue self- vs. other-related values. We also explored whether participants across different education levels expressed different kinds of values using rich text data. 

 A community sample of 220 individuals (mean age = 33.75 years (SD = 11.62); 144 females; 96 Black, 86 White, 16 Asian, 9 Hispanic, and 13 Other) ranked self-related (power/status, wealth, fame) and other-related (compassion/kindness, family/friends, spirituality) values, rated their importance to self on a 0-10 scale, and reported their perceived social status and educational background. A subset (n = 80) wrote about their highest-ranked value (Kang et al., 2018). Participants were categorized into three groups based on their education and their parents’ education: no-college-degree (n = 63), first-generation (n = 74), and continuing-generation (n = 83). Participants indicated their perceived social status within their community using the 10-point MacArthur Scale (Adler et al., 2007). We tested differences in self- and other-related values across educational backgrounds by regressing the mean scores of self- and other-related values on perceived social status. Using the text data, we generated word clouds to explore linguistic characteristics across educational backgrounds. 

 We found a significant relationship between education status and other-related values, F(2, 217) = 6.45, p = 0.002. The continuing-generation group placed less importance on other-related values than the no-college and first-generation groups, who did not significantly differ. Education was not related to self-related values. Higher perceived social status was associated with greater importance placed on self-related values (β = 0.25, t = 3.55, p < 0.001). Perceived status did not significantly relate to other-related values. Among 80 participants who described their highest-ranked value, 78 chose other-related values. “Family” was the most common word across all groups but appeared less frequently in the continuing-generation group.

 Findings support that personal values vary by socioeconomic background. Lower-education individuals emphasized other-related values, especially family, while higher-education individuals endorsed a broader range of other-related values. Higher perceived social status was linked to prioritizing self-related values. These findings align with research showing that lower SES individuals emphasize interpersonal values and interdependence (Carey & Markus, 2017; Snibbe & Markus, 2005). Building upon these findings, we are currently testing whether individuals’ daily value alignment – the degree to which their actions reflect the values they cherish – relate to health outcomes, using experience sampling and text data.

Dunkley, Truman – Program: Ph.D – Biology: Computational and Integrative

 Bioenergetic profiles of psychrophiles across domains of life are unusual in that intracellular ATP levels increase with declining temperature. Whole-transcriptome sequencing of the glacier ice worm Mesenchytraeus solifugus revealed a unique C-terminal extension on the ATP6 protein, which forms part of the proton pore of mitochondrial ATP synthase (Complex V). This extension, positioned near the proton exit pore, comprises alternating histidine residues thought to increase proton flux through Complex V leading to elevated ATP synthesis. To test this hypothesis, we fused the M. solifugus C-terminal extension to Escherichia coli AtpB (the ATP6 orthologue) and observed a ~ 5-fold increase in ATP synthesis. This enhancement was unidirectional as we observed no change to ATP hydrolysis rates. These findings offer an avenue for identifying critical factors associated with ice worm adaptation.

Ebrahimi, Alireza – Program: Ph.D – Biology: Computational and Integrative

In recent years, research on catalytic nucleic acids as a unique class of bioinspired catalysts has expanded significantly. Among these, the hemin/G-quadruplex (hGQ) DNAzyme has attracted considerable interest due to its well-documented ability to mimic horseradish peroxidase activity. In this study, we explored the potential of integrating a DNAzyme sequence (hGQ) as the catalytic site with an aptamer as the binding site to create a functional hybrid conjugate with enzyme-like properties. We then evaluated the activity of this hybrid system in comparison to the hemin aptamer under varying pH conditions. Our findings indicate that the hybrid conjugate exhibited enhanced activity in acidic environments, while both structures displayed similar oxidase activity under basic conditions. These results highlight the structural and environmental influences on the catalytic behavior of hemin-binding DNA structures, providing valuable insights into potential applications of hGQ-based systems.

Ennis, Jahmal – Program: Ph.D – Biology: Computational and Integrative

BamA is a beta-barrel outer membrane protein (OMP) that assists in folding other OMPs in Gram-negative bacteria. Because of its key role, high conservation, and location at the cell’s surface, it is a promising target for therapeutic intervention and is worth further investigation. The exact mechanism that modulates the open and closed state of BamA’s lateral gate is currently not well understood. Recently, a novel anionic ceramide, ceramide polyphosphoglycerate (CPG2), was found in the Gram-negative bacterium Caulobacter crescentus. Knocking down BamA or anionic ceramide synthesis yields similar phenotypes, including increased antibiotic susceptibility and induction of the unfolded-protein response, suggesting the two may be linked. Here, we computationally study anionic ceramide interactions with BamA using coarse-grained molecular dynamics. We found several super cial binding sites for anionic ceramide and measured their binding affinity to BamA in ternary mixed POPC, neutral ceramide, and anionic ceramide model membranes using the density threshold a nity method.

Everill, Bronwen – Program: MFA – Creative Writing

 An die freude: A creative nonfiction reading

 There’s a moment in the fourth movement of Beethoven’s ninth symphony. We’ve heard the melody of the Ode to Joy a few times as it’s been passed around from the bass soloist to the male choir, then the soprano and the women. Beethoven has embellished it with the quartet of singers. And it sort of ends. You think, okay, Ode to Joy done, now he’s going to move onto other stuff. He does for a while. There’s a little military-sounding jaunty march version of the song. The singers come back, but then the strings take over and it starts to sound like the Beethoven you know from the Fifth Symphony or the Sixth Symphony, the one in “Fantasia” with the Pegasus and cherubs. At around 6 or so minutes in, though, depending on your recording, or the speed of the conductor, or the orchestra’s enthusiasm, that all changes.

 What is technically happening at 6 and a half minutes into the fourth movement is a shift to B major, then b minor, then D major. The shift from b minor to D major is a shift from a relative minor to a relative major key – they are related, the sad and happy versions of the same scale. Weezer does it in Buddy Holly. Madonna does it in Like a Prayer. Turn that frown upside down.

Was it more important to make art? To help people experience a feeling? Or to analyze it, to help them understand why they felt it?

Fernandez, Evelyn – Program: MAT – Teacher Education

This action research project investigates the impact of multisensory phonemic awareness activities on the early reading skills of preschool students. The study focuses on four students from diverse cultural and linguistic backgrounds in an urban preschool classroom. Recognizing that traditional phonics instruction often falls short for students with learning challenges, the project integrates multisensory activities such as the Sound Scavenger Hunt, Rhyming Songs, and Musical Name Chairs. These activities engage sight, sound, touch, and movement to make learning interactive and accessible. Data collection involved pre-assessments to establish baseline literacy skills, followed by structured observations of student engagement and progress during the activities. Results revealed that most students demonstrated measurable improvements in letter recognition, rhyming awareness, and phonemic identification. Students A, B, and C exhibited significant progress, while Student D showed minimal improvement, indicating the need for additional support. Engagement levels were generally high, with activities rated positively by students, underscoring the effectiveness of incorporating multisensory elements. This research highlights the potential of multisensory methods to enhance foundational literacy skills, particularly for students from diverse backgrounds or with developmental challenges. By tailoring instruction to individual strengths and sensory preferences, educators can foster inclusive, engaging learning environments. These findings advocate for the broader implementation of multisensory strategies in early literacy programs to support diverse learner needs and promote academic success.

Fleser, Tyler – Program: MFA – Creative Writing

Despite Overwhelming Evidence, Maybe it Was Loneliness That Killed the Dinosaurs & Other Stories is a short story collection manuscript following misfits who come from the same small township in West Michigan. These stories interrogate the way we wear our differences from others as a shield, as well as the loneliness that comes from feeling like an outsider in a. Following my reading at AWP, I will read further from excerpts of these stories, offering time for questions afterwards.

Fredricks, Katie – Program: Ph.D – Childhood Studies

Building Solidarity to Break Cycles of Repression

In the current historical moment, political and popular discourses invoke “children” and “childhood” for multitudes of competing ends. While perspectives of childhood have shifted throughout history, in contemporary media children are both praised as the revolutionary future—inflated with expectations that they will make up for centuries of others’ carelessness—and denigrated for their innocence with comparisons to leadership adults perceive as incompetent or petulant.. Although inconsistent, discourses of childhood are inseparable from discourses of politics (James and Prout 1990). In this paper, I examine the intertwining discourses of childhood and politicized issues as presented in the British reality television program Gogglesprogs (2016-2019). The project also considers paratextual material in the form of critical and popular responses to the program. The analysis interrogates these mediated representations of children, specifically those related to their capacities as consumers of media related to highly politicized issues such as gender, climate justice, and democratic elections in both the UK and US.  

Held up across critical reviews and casual tweets as genuine, witty, inspirational, and even restoring some viewers’ “faith in humanity”, the kid stars of public broadcast Channel 4’s Gogglesprogs captured the public’s attention across two seasons and a few holiday specials.  Week after week, home audiences tuned in to watch edited and packaged footage of British children in their homes, watching and reacting to television. Watch children watch TV—that was it, the whole schtick. My analysis employs data from the reality television program, as well as its critical and popular reception, as sites prime for inquiry into contemporary representations and conceptualizations of young viewers and their relation to the social world (as seen on TV.) 

Based on a critical content analysis of all sixteen existent Gogglesprogs episodes, as well as a sampling of news articles, blog posts, and social media exchanges, this examination illuminates the complex ways in which children’s media consumption, something long opined as a social problem, is actually a rare avenue through which children are afforded the opportunity to actively negotiate politicized messages. This paper explores how children watching television are an active audience of meaning-makers. Through paratextual analysis, it also explores how the meaning children make while watching television, when subsequently televised itself, may serve to shape public perspectives on children and childhood more broadly. This research demonstrates how reality programming can be used as a barometer for our social construction of childhood. Ultimately, Gogglesprogs reinscribes contradictory representations of the child as a political consumer. In terms of the mundane activity of viewing as well as in relation to large-scale social and political forces, Gogglesprogs challenges notions of children as passive objects. The wider reception of the program, however, serves to reinforce classic conceptualizations of the child as authentic, “wise beyond their years,” and a force for good in ways that are both flattering and limiting. In an era when media presentations  and representations are frequently questioned, the expected purity and honesty of the child—even when interwoven with mediated politics—preserves  television consumers’ faith in the future of democracy. 

Garg, Gaurangi Sandeepkumar – Program: MS – Computer Science

 Optimizing Disk Power Consumption Through Predictive Modeling of IO Request Interarrival Times

 Modern data centers continuously strive to optimize energy efficiency, with storage systems playing a crucial role in power consumption. In this study, I analyze real-world IO traces (from Google) from a single-disk, single-server setup to maximize disk power savings by turning off disks when they are idle.

 Our research is structured into two key objectives:

 1) Predicting Interarrival Times for Disk Reads – We analyze interarrival times of disk read requests using historical patterns to determine if a disk can transition to a low-power state between consecutive IO requests. This involves building a lightweight predictive model that estimates the duration until the next request while minimizing overestimation errors.

 2) Application-Aware Disk Packing – Once interarrival times are effectively modeled, we explore strategies to co-locate applications on a disk to maximize disk idle periods without violating Service Level Objectives (SLOs), particularly the Time to First Byte (TTFB) threshold and also making sure that the disk is sufficiently full. 

 Our methodology involves:

  1) Extracting and filtering IO requests where disk activity occurs (i.e., cache miss).

  2) Computing disk interarrival times and idle times post-request completion.

  3) Developing a probabilistic model that predicts safe disk sleep durations based on historical patterns and time of day.

  4) Later, I will also look into employing lightweight machine learning models suited for real-time inference to minimize energy overhead.

 This research aims to reduce disk power consumption while maintaining performance reliability. The insights gained could contribute to efficient disk power management strategies for data centers, enabling cost-effective and energy-aware storage solutions.

Gebretensay, Samrawit – Program: MA – Psychology

Maternal healthcare inequalities in the United States are evident in the pressures that Black, Hispanic, Asian, and White women face to undergo labor induction, epidural analgesia, and cesarean section interventions that perpetuate disparities in autonomy, morbidity, and mortality (Centers for Disease Control and Prevention, 2024; Hoffman et al., 2016; Vedam et al., 2019). Although these interventions might be necessary in some cases, systemic racism, cultural bias, and structural barriers disproportionately expose women to nonconsensual or inadequately explained interventions. For instance, among Black women, 36% of births are via cesarean section (Huesch et al., 2015). Comparably, Hispanic women experience pressure for medical interventions, such as labor inductions and cesarean delivery, and face a heightened risk of discrimination due to language barriers (Valdez et al., 2023). Similarly, Asian women experience higher odds of cesarean delivery due to failure to progress in labor compared to White women (Okwandu et al., 2021). White women, though privileged by systemic bias, also report feeling more pressure to induce labor compared to Latina women (Declercq et al., 2020). Although there are differences in birth outcomes by race, more research is needed regarding the specific pressures that women face during labor induction, epidural use, and cesarean section and which racial groups experience heightened pressure for specific interventions. The purpose of this study is to explore racial disparities in perceived pressure for birthing interventions and to highlight how provider practices may disproportionately constrain autonomy among Black, Hispanic, Asian, and White women during childbirth. We hypothesize that Hispanic, Black, and Asian women will report higher perceived pressure for medical intervention during childbirth compared to White women. Specifically, we predict that Black women will experience the highest perceived pressure for cesarean section, Hispanic and Asian women for labor induction, and Hispanic women for epidural use. The findings and implications of the study will be discussed.

Harken, Jillina – Program: MA – Emerging Media

Ringing Out of Time: The Payphone as a Cultural Artefact in Modern Spaces

This research explores the payphone as a cultural artefact and its potential role in shaping public engagement with obsolete technology. While once a critical part of global telecommunications infrastructure, the payphone has all but disappeared, replaced by mobile communication. This study investigates the implications of reintroducing a fully functioning payphone into a modern public space-examining how individuals might interact with, reinterpret, and repurpose defunct technology in a digital age.

Using media archeology (Fickers & van den Over) and medium design (Easterling) as guiding frameworks, this project theories that such an instillation would not only spark curiosity but also reveal deeper insights into the societal impact of technological obsolescence. A hypothetical payphone installation at Rutgers-Camden’s campus quad serves as a conceptual model for exploring questions of nostalgia, infrastructure awareness, and shifts in communications behavior.

Through behavioral mapping, speculative ethnography, and participant interviews, this research considers the ways people might engage with a forgotten technology. Would they attempt to use the phone? Would it serve as a social hub or a relic to be ignored? How does its presence alter perceptions of the surrounding environment? By looking back at the rise and fall of the payphone, this project invites critical reflection on the rapid cycles of technological advancement and the forgotten infrastructures that shaped our modern world.

This study offers insight into the cultural memory of communication technologies, the experimental learning of obsolete tools, and the shifting dynamics of public space in an era of digital ubiquity.

Heitz, Catherine – Program: Ph.D – Prevention Science

Trauma and stress from life events can be contributors to community violence. However, research suggests that trauma-informed care (TIC) training in professionals decreases violence by increasing trauma understanding, improving service networks, and building community-police trust. This study focuses on the impact of a TIC policing initiative on violent crime rates in Kensington, Philadelphia using an interrupted time series (ITS) design. The Police Trauma Training Pilot Program recruited officers in Kensington Police District 24 for training sessions in Summer 2021 educating officers on trauma-informed policing, adverse childhood experiences, biopsychosocial aspects of trauma, and encouraging community engagement. Seven years of violent crime data from January 1, 2017, to December 31, 2023, was accessed via the Philadelphia Police Department. Results show that the average number of violent crime incidents per month was estimated at 30 incidents and negligibly increased post-intervention (p = .819). All crime incidents, both violent and non-violent, in Kensington significantly decreased by roughly 5 incidents per month from an average of 1,083 (p < .000). While violent crime incidents are relatively unchanged, this analysis suggests that TIC may be more impactful for non-violent crime. Further evaluation of TIC initiatives is needed to understand the potential impact on communities and policies.

 Keywords: trauma-informed care, community violence, policing, crime prevention

Jahan, Nishat – Program: Ph.D- Biology: Computational and Integrative 

Development of Hemin-binding Aptamers for Biosensing 

Hemin/DNA aptamers can catalyze the hydrogen peroxide (H₂O₂)-mediated oxidation of various chemical compounds. For instance, hemin/G-quadruplexes convert Amplex Red into fluorescent product resorufin, enabling fluorescence-based detection. Due to their high catalytic efficiency, hemin-binding DNA aptamers have been widely explored for biosensing applications for the detection of various targets. In this study, we investigated a special non-G-quadruplex hemin-binding aptamer (Hem1), which was reported in the literature to enhance the peroxidase-like activity of hemin. Hem1, featuring highly conserved repeating binding loops, cannot form a G-quadruplex and shows selective binding for hemin. Due to stacking-based binding, G-quadruplex structures may non-specifically bind to different porphyrins and other planar molecules, hindering practical applications. Non-G4 aptamers like Hem1, however, allow discrimination of different porphyrins, making more robust hemin/aptamer complexes and providing biosensor versatility and specificity. For this study, we selected the hemin-binding aptamer mutant, Hem1-2T (K_d = 43 nM for hemin), to investigate the catalytic reaction with H₂O₂ using Amplex Red as the dye substrate for the purpose of biosensor development. By splitting the aptamer, Hem1-2T, into two fragments and optimizing their sequences, we aimed to develop split aptamer pairs that will bind to hemin in the presence of target DNA (COVID-19 DNA) with high affinity, exhibiting enhanced catalytic activity towards H₂O₂. Different split aptamers were designed and evaluated under various conditions (pH, temperature, and magnesium ion concentrations) to improve sensitivity and specificity and thus optimize their performance. Hybridization predictions were conducted using NUPACK analysis, and fluorescence spectroscopy was used for experimental characterization. Our findings demonstrate the feasibility of using split aptamers for sensitive and selective target DNA detection, highlighting their potential for biosensing platforms with promising applications in diagnostics and molecular detection.

John, Assa – Program: MS- Chemistry

  Forensic DNA collection traditionally relies on cotton swabs, yet their material composition limits DNA recovery and single-cell analysis. Cotton fibers often retain cells, reducing DNA yield and complicating forensic profiling, particularly in cases with multiple contributors. This study introduces a novel swab featuring a dissolvable alginate coating on cotton fibers to enhance detection sensitivity and single-cell analysis. The swab was developed by reacting sodium alginate with calcium chloride to create the hydrogel coating. Partial dissolution of the hydrogel at 37°C with orbital shaking was confirmed using FTIR, TGA, and SEM-EDS characterization techniques, leaving the cotton fibers intact. Performance comparisons with conventional cotton swabs assessed DNA yield, cell integrity, and forensic applicability. Visualization of cells on a hemocytometer using a Nepagene video microscope at 52× magnification demonstrated increased cellular yield with individual alginate-coated fibers. This non-destructive approach can improve forensic DNA collection, enhancing crime scene investigations and forensic methodologies.

Johnson, Chareina – Program: MA- Psychology

Existential isolation (EI) can be described as the feeling or perception that an individual is alone within their life experience as others do not share similar thoughts, feelings, or actions. First-generation college students (FGCS) whose parents did not attend college may experience higher levels of existential isolation. Research suggests that, compared to continuing-generation college students (i.e., students whose parents graduated college), FGCS have less knowledge about navigating college and are often afraid of asking for help for fear of being viewed as imposters. In this cross-sectional study, we investigated differences in EI and help-seeking intentions in 79 FGCS and 75 continuing-generation college students. The results show that first-generation students report higher levels of EI which is, in turn, associated with more negative attitudes toward academic help-seeking. These results suggest that EI plays a role in the difficulties FGCS experience while navigating college and that EI may have implications for student success.

Jonnalagadda, Sai Anusha – Program: MS – Biology

Sphingosine-1-phosphate receptor 3 (S1PR3) plays a pivotal role in regulating social behavior by modulating neuronal activity in circuits involved in emotional regulation, notably the prefrontal cortex (PFC) and the basolateral amygdala (BLA). Recent evidence suggests that S1PR3 activation enhances sociability by increasing neuronal activity in the infralimbic cortex (IL), which in turn suppresses BLA function. However, the molecular and cellular mechanisms underlying this interaction remain unclear. In particular, while S1PR3 activation excites both glutamatergic and GABAergic neurons, its ability to inhibit BLA activity poses a complex challenge for understanding how these divergent effects contribute to sociability. This research aims to explore how S1PR3 activation regulates neuronal signaling in the IL-BLA circuit, promoting sociability, and to elucidate the molecular and cellular processes involved. 

To address this, we will quantify key neuronal activity markers, including c-Fos, GAD, and NeuN, in the PFC, central amygdala (CeA), and BLA. These markers provide insights into neuronal activation and inhibition, helping to unravel the impact of S1PR3 modulation on these regions. A comparative analysis of treatment effects will be conducted by comparing expression levels in mice treated with CYM5541 (S1PR3 agonist), CAY10444 (S1PR3 antagonist), and vehicle-treated controls. This approach will allow us to assess the differential effects of S1PR3 activation and inhibition on neuronal activity in the PFC, CeA, and BLA, providing crucial insights into the neurobiological mechanisms underlying sociability and emotional regulation. 

The methods employed in this study include systemic injections of CAY and CYM to modulate S1PR3, followed by social interaction and tail suspension tests to evaluate sociability and stress behaviors in these mice. Immunohistochemistry will be utilized to detect specific neuronal markers in brain tissue, and confocal imaging will provide high-resolution, three-dimensional images to visualize neuronal activity and receptor localization. Data from these techniques will be analyzed using ImageJ for quantification and Excel for data organization and interpretation. 

The potential impact of this research extends beyond basic neuroscience, with implications for understanding neuropsychiatric disorders such as anxiety and depression, which are often characterized by dysregulated activity in circuits involving the PFC and BLA. By investigating how S1PR3 modulates neuronal activity in these regions, this study may offer new therapeutic targets for restoring balance in stress-related circuits. Ultimately, S1PR3 could emerge as a novel target for the treatment of neuropsychiatric disorders, offering promising avenues for managing conditions related to impaired sociability and emotional dysregulation. 

In conclusion, this research aims to provide a deeper understanding of the cellular and molecular mechanisms through which S1PR3 activation influences neuronal signaling in the IL-BLA circuit, with potential therapeutic applications for enhancing sociability and treating neuropsychiatric disorders. 

Kasturi, Akhila – Program: MS – Biology

Sociability is a vital aspect of both mental and physical well-being, and impairments in social behavior are a hallmark of numerous psychiatric disorders. Identifying novel mechanisms that regulate sociability is crucial for developing safer and more effective therapeutic strategies. Recent research highlights the role of sphingosine-1-phosphate receptor 3 (S1PR3) activation in enhancing sociability in mice. However, the underlying brain regions activated by S1PR3 that underlie this pro-social effect are not well understood. We propose that S1PR3 activates the infralimbic cortex (IL), a brain region that inhibits the basolateral amygdala (BLA), which regulates fear and social avoidance. The IL inhibits the BLA by activating inhibitory GABAergic neurons adjacent to the BLA. By quantifying the neuronal activity marker cFos, we found that the S1PR3 agonist CYM5541 increases neuronal activity in the IL following a 2-hour social interaction. Interestingly, S1PR3 also increased levels of GAD67, the rate limiting enzyme for the inhibitory neurotransmitter GABA, in the BLA. Future work will determine whether GABAergic intercalated nuclei are active in these mice. We hypothesize that by modulating these neural pathways, S1PR3 activation promotes social interactions. Our findings provide the groundwork for the mechanisms by which S1PR3 promotes sociability, which might reflect reduced social anxiety-like behavior.

Keir, Anelise – Program: MA – Psychology

Introduction: We examine how residents of New Jersey feel about the future development of Offshore Wind (OSW) in the Atlantic coast. 

 Objectives: We develop novel scales assessing concerns people have regarding OSW and other carbon-free energy sources.

 Materials and Methods: In a sample of over 400 residents, these new scales exhibit good psychometric properties and suggest there are multiple dimensions governing resistance to OSW, including 3 primary factors: effects on the biosphere including fish, birds, and mammals, impacts on the visual ecology of the coast, and economic implications. 

 Results: Support for OSW was positively associated with support for other carbon neutral energy sources except for nuclear power. Support for OSW was negatively correlated with the amount of time people spend at the coast, and positively associated with distance people reside from the coast. Political affiliation was a significant factor affecting support, with conservative participants exhibiting greater opposition than liberal participants. Age was not a significant predictor of support, yet gender was with greater support among women than men. We found overall optimism for the future of OSW, but identified causes for concern based on many participants’ concerns

Conclusion: In total, this project provides key insight into thought processes underlying how people think about climate change in relation to infrastructure projects that aim to reduce the existential threat that climate change poses to society, while providing insight upon strategies for better communicating the benefits of OSW for mitigating these threats.

Mahdinia, Fatemeh – Program: Ph.D- Prevention Science  

Sense of Agency (SoA) is a core concept related to our experience as intentional agents in our environment. Explicit and implicit measures have been used to study SoA. Recent findings suggest that the most common implicit measure, namely Temporal Binding (TB), may reflect memory processes rather than SoA. Here, we implemented two TB measures and an explicit measure in a novel goal-directed extended action task to better understand SoA measures. Participants either observed or produced dot movements to a target of choice. They then estimated one of two possible durations; for Temporal Binding version 1 (TB1), they estimated the duration between the end of the dot movement and a tone that played either 300, 500, or 700 ms later (akin to traditional TB studies). For TB2, they estimated the duration between the start and end of the dot movements. After every 10 trials, they also provided explicit SoA ratings. The results indicated that participants reported stronger explicit SoA during active than passive movements. Results from neither TB version aligned with predictions based on TB-accounts as a reflection of SoA. We discuss memory-based and scaling accounts as alternative interpretations for our data. 

Maienza, Cathryn – Program: Ph.D- Biology: Computational and Integrative 

Melatonin, a molecule with diverse biological functions, is ubiquitously present in living organisms. There is significant interest in understanding melatonin signal transduction pathways in humans, particularly due to its critical role in regulating the sleep-wake cycle. However, a knowledge gap remains in fully elucidating the mechanisms by which melatonin influences circadian regulation. To bridge this gap, there is a growing need for a model system to study the role of melatonin in circadian clocks, with Neurospora crassa being a promising candidate. As a first step in this investigation, we focused on identifying melatonin receptors in N. crassa. Given the lack of sequence similarity between potential receptors in this fungus and known human melatonin receptors, we utilized structural similarity analysis through AlphaFold2. This approach led to the identification of a strong candidate gene, gpr-3, which shares structural similarities with human melatonin receptors. Experimental validation confirmed that the removal of GPR-3 from cells results in the absence of melatonin signaling. This proof-of-concept study underscores the potential of N. crassa as a model organism for circadian research and demonstrates the broader applicability of using AlphaFold2, especially when sequence similarity does not lead to candidate genes, for identifying novel receptors across different species.

Markellos, Nicholas – Program: Ph.D – Childhood Studies

Toni Morrison’s controversial story The Bluest Eye has featured heavily on the ALA Banned Book list over the last twenty years. The novel has been called into question for its views on Diversity, Equality, and Inclusion, as well as the graphic portrayal of rape and incest. This work, however, is important because of its approach and message to the community. The narrative structure not only crosses the boundaries of intersectionality but also speaks to the higher level of theoretical interpretation, which warrants its noteworthiness and acclaim. By examining the narrative structure through the intersection of class, gender, and race in The Bluest Eye, it will be shown how various challenges to this work are not only superficial but also unwarranted. Further, this analysis will attempt to reveal how teaching this work can allow high school students to explore the history of racial division in this country.

Marrie, Touba – Program: M.B.S.-Business and Science

Financial Mathematics and Risk Modelling (With a Statistical Case Study of The Gambia’s Economy and Financial Challenges)

This research explores critical subtopics within the domain of financial mathematics, including performance measurement, behavioral finance, data analytics, and cost-of-living analysis, with a focus on applying these concepts to address financial mismanagement and systemic challenges in The Gambia. Using The Gambia as a case study, the study investigates the country’s financial mismanagement through a detailed statistical analysis of its public accounts, proposing solutions for effective fund management, risk mitigation, and sustainable economic growth. Key findings from the 2023 economic assessment reveal significant structural challenges, including a high prevalence of informal employment (63%), gender inequality in access to formal jobs, youth unemployment (74% of the underutilized population aged 15–35), and regional disparities in employment opportunities. To address these issues, the study recommends formalizing the economy, promoting gender equality, enhancing youth employment programs, and investing in rural development.

The research also analyzes the 2020 public financial accounts of The Gambia, as recently reported by the National Audit Office, revealing substantial financial mismanagement. Key findings include significant discrepancies in financial statements, such as understated cash and cash equivalents (D88,197,691.58), understated deposits (D135,458,189.76), overstated receivables (D1,285,461,104.25), and unidentified account line items (D131,067,541.59). Additionally, misclassification of loan payments (D108,452,653.61) and missing documentation for expenditures (D1,517,329,882.42) were identified, highlighting systemic weaknesses in financial controls.

To quantify the financial risks, the study employs advanced financial mathematics models, including Value at Risk (VaR) and Monte Carlo Simulation. The VaR analysis estimates a maximum potential loss of D246,750,000 at a 95% confidence level over one year due to mismanagement. The Monte Carlo Simulation, based on 10,000 iterations, indicates a 10% probability of losses exceeding D2 billion and a 90% probability of losses remaining below D1.8 billion. These findings underscore the urgent need for improved transparency, stronger internal controls, and regulatory compliance to safeguard public funds and ensure accountability. The research concludes with actionable recommendations to address these systemic flaws and promote sustainable economic development in The Gambia.

Marszalek, Elizabeth – Program: MA – English

Tunnels, Boats, and Bedrooms: Trauma As Space in the BBC’s Peaky Blinders

Heavily drawing from Robert Tally’s Topophrenia, as well as several prominent trauma scholars, this paper discusses how Peaky Blinders operates within established definitions of place, space, and trauma. Through a close reading of three scenes throughout the show, I will argue that trauma is first processed as an oscillation between space and place by protagonist Tommy Shelby. Tommy, as if in an attempt to ‘map’ the space through place, uses the familiar images and environments of a previous trauma in order to experience and comprehend the new one, the former trauma generally being his experiences during World War One. This act falls in line with van der Kolk and van der Hart’s idea that sufferers of trauma are likely to experience new traumas as the return of an old trauma, rather than a new one. Given Tommy’s habit of mapping each new trauma as an old one, it stands to reason that each trauma, now colored by the same lived experience, would eventually compound into a singular traumatic space. Further, as each new traumatic experience is processed through the lens of a familiar trauma, the logic of time and place eventually becomes impossible, creating a ‘thirdspace’: a space uninhibited by specifics of time and place and instead accommodates everything about every trauma. Through continuous mapping, this thirdspace thus becomes unmappable, and by attempting to still map this unmappable space, Tommy soon becomes lost within his own trauma, succumbing to spatial anxiety in a state similar to madness.

Keywords: geocriticism, spatial studies, trauma theory, pop culture, world war one

Mason, Megan – Program: Ph.D – Prevention Science

Individuals’ Experiences of Care Partner Involvement in Heart Disease Management

Heart disease is the leading cause of death in the U.S., affecting both individuals diagnosed and their families. Managing heart disease often requires lifestyle changes, such as adopting a healthier diet and increasing physical activity, which can impact family dynamics and place additional demands on “care partners.” However, most research on care partner involvement focuses on men and married individuals, leaving gaps in understanding how care partners contribute to disease management among women and those of different marital statuses. Given gender differences in heart disease experiences and caregiving roles, as well as high widowhood rates among older women, understanding these dynamics is critical. 

This pilot study examines differences between individuals with heart disease who have a care partner versus those who do not. Additionally, we explore gender differences in care partner involvement in disease management. 

Mcmurray, Allison – Program: MA – Psychology

Romantic Partners’ Comments on Unhealthy Eating: Role of Gender and Sexual Orientation

 An appropriate diet is essential for preventing and managing chronic health conditions (CDC, 2024). Romantic partners are one of the strongest influences on one’s diet by influencing what their partners eat, including engaging in diet-related social control: influencing partners to eat less unhealthily, e.g., eating certain foods less frequently, eating smaller portions (Craddock, 2015). Past research on opposite-gender couples suggests that women are more likely to exert social control, whereas men are more likely to be the targets of control (August & Sorkin, 2010). Diet-related social control has been studied less frequently in sexual minority populations; findings on sexual orientation and gender differences in the frequency of social control within same-gender couples are mixed (August et al., 2016; Markey et al., 2016; Umberson et al., 2018). Social control meaning and frequency differ in the context of chronic health conditions (Rook et al., 2010).

 The present study aimed to understand how frequently romantic partners receive diet-related social control, what specific aspects of unhealthy eating their partners comment on, and how different chronic health conditions requiring changes to diet were related to social control frequency. In addition, we examined how diet-related social control experiences varied by gender and sexual orientation.

 Cross-sectional data from two Prolific studies of individuals in long-term romantic relationships were analyzed (Total N = 833; Study 1, individuals in opposite-gender relationships, n = 444; Study 2, individuals in same-gender relationships, n = 389).

Mohan Kumar, Neha – Program: MS- Computer Science

Optimizing GOSDT-Guesses: A Faster, Memory-Efficient Python Implementation with LightGBM-Based Threshold Guessing

We are interested in studying machine learning frameworks for classification that are easily explainable. One such model, decision trees (and other related tree models like Random Forests) has been the basis of recent work done by Dr. Cynthia Rudin’s research group at Duke University on Optimal Sparse Decision Trees (OSDT) for datasets with continuous variables. In a series of three papers starting with OSDT, they proposed an algorithm to construct Generalized and Scalable Optimal Sparse Decision Trees (GOSDT) and GOSDT-Guesses, where the last model significantly improves performance over the base model, while ensuring that the resulting trees are provably optimal with respect to model complexity and classification accuracy.

However the GOSDT models remain computationally expensive for large datasets and have limited accessibility being implemented in C++. In our work, we propose an optimized version of GOSDT-Guesses that achieves reduction in time complexity and memory requirements while preserving accuracy. This is accomplished by introducing two changes to the process of binarizing continuous variables, which is a precursor to the construction of OSDTs. First, we use the LightGBM framework instead of a traditional GBM (Gradient Boosting Machine) to guess thresholds. LightGBM uses techniques like Gradient-based One-Side Sampling (GOSS), histogram-based learning, exclusive feature bundling (EFB), and leaf-wise growth strategy, which optimize memory usage and reduce computation time. Secondly, we refine the threshold selection process by introducing adaptive threshold grouping, which clusters similar values and retains only representative ones, reducing search space without compromising accuracy. Our work also includes the complete reimplementation of GOSDT-Guesses in Python, eliminating the dependency on C++ bindings. This makes the algorithm more accessible to Python developers and easier to modify for further improvements. Our implementation appears to be significantly more memory-efficient, and enables the GOSDT approach to scale to larger datasets with continuous features. With reduced memory constraints, we are able to increase the number of estimators in LightGBM, further boosting classification accuracy.

Nelson, Elizabeth – Program: Ph.D – Childhood Studies

This poster presents an exploration of two theories of listening that help to reconceptualize communication and listening with babies and non-verbal children. A focus on listening moves away from the speech-centric focus of communication, and the emphasis on the spoken word in discussions of voice and agency in the field of Childhood Studies. Limiting conceptualizations of communication and voice are ableist, unjustly dismissing non-verbal children and babies. Building upon the work of Lisbeth Lipari (2010, 2014) and Bronwyn Davies (2014), I argue that babies can be invited in as capable listeners and communicators equal to older children and adults. To show this, I focus on two aspects of their theories: a de-centering of individuated selves and a turn towards the unknown. First, both authors’ dialogical, holistic theories of listening explicitly dismiss an individualized, bounded notion of a self, and moreover claim that as conceptualized, this false self is unable to engage in listening. This nonexistent self belongs to a neoliberal developmentalist framework that posits children as lesser beings, not fully human until possessing adult capabilities and reason. It conceptualizes the self as separate, with firm and static boundaries between “I” and others. Instead, Davies offers the idea of a “subject as intra-active becoming”, who is a relational, multiple, and alive self, capable of engaging in what Davies terms “emergent listening,” a transgressive space of differentiation or intra-active becoming. Likewise, Lipari points to Buber and Levinas, who present humans as relational intersubjectivities, and each self as “called into being by the relational encounter with the other” to build her concepts of “listening being” and “interlistening.” Lipari’s interlistening is a third space between two communicators, where communication is not two selves “coordinating” with one another, as monological theories of listening claim, but is a space where “each of us is sensitive to a continually changing ‘it‘ between us“. By presenting listening not as a passive activity of a bounded adult self, but as occurring among two or more fluid, relational, intersubjective beings, these theories of listening can welcome in babies as active, agentic listeners, equal to adults. Second, Davies’ and Lipari’s theories of listening demand the listener reject the status quo and cultural norms, partially acquired through and reinforced by an understanding of language as a social reality. Davies’ emergent listening requires turning towards the resulting void to engage in creative evolution. Lipari’s interlistening is an attempt to disrupt the social reality that language has created and imposed upon everyday life. Though babies do not dismiss their culture in the same way as older humans, they do in fact engage in a turning from what they have already experienced or known, towards the unknown. I demonstrate this through discussing babies’ sensory experience beginning in the womb and expanding the definition of awareness beyond adult-centric, self-reflexive notions. Because of the intersubjective reality of all human beings inherent in their theories, more expansive descriptions of turning towards the unknown and awareness allow Davies’ emergent listening and Lipari’s listening being and interlistening to include babies. 

Nwede, Chiamaka Stellamaris – Program: Ph.D – Public Affairs

Puerto Rico School Closures: Saving Our Education to Save Our Community – A Case Study of San Juan

This case study explores the impact of public school closures in Puerto Rico, with a specific focus on San Juan as a case study. Following the economic crisis and natural disasters, Puerto Rico faced significant educational disruptions, leading to hundreds of public schools being closed. This research investigates the social, economic, and community implications of these closures, highlighting the challenges displaced students, educators, and families face. It also examines non- profit organization’s role in making a social impact on community development. Through qualitative and quantitative data collection, including unstructured interviews with community leaders and focus group discussions with students and teachers, The study underscores education’s critical role in community resilience and development. The findings reveal that families, schools, places of worship, government, law enforcement agents, courts, lawmakers—every one of us— have a crucial role to play in this script of saving education to save our community. This research advocates for a holistic approach to education reform that safeguards the future of Puerto Rico’s youth and strengthens its communities.

O’Donoghue, Dylan – Program: Ph.D – Public Affairs

Researchers find that migrant workers in the tech industry experience exploitation as a result of algorithmic determinism and subcontracting. While there is evidence that tech companies advocate for specific labor structures, researchers have yet to explore how tech companies influence the regulations that lead to worker precarity. I address this issue using an Abolitionist Feminist framework to examine tech companies lobbying around migrant labor. I use process tracing to gather data from the U.S. Senate Lobbying Disclosure and congressional testimonies to track tech sector lobbying related to immigration and subcontracting employment structures. I find that tech companies play a role in shaping their labor pool in two ways. First, through immigration lobbying, where companies simultaneously push for decreased barriers for specialty visa holders and an increase in border security and surveillance. Second, tech companies lobby to ensure that gig workers remain subcontractors who lack employee benefits and bargaining rights. My findings demonstrate that tech companies use complex strategies to reconstruct who can and cannot work and what labor protections are available. These activities grow the tech labor pool but also increase worker precarity. This study has important implications for reducing migrant labor exploitation and the future of labor in tech.

Patrick, Labenyimoh – Program: MS – Biology: Computational and Integrative

  Mitochondrial diseases affect an estimated 1,000–4,000 children born annually in the United States, with a significant proportion resulting from mutations in ATP synthase (Complex V) subunits. Among these, mutations in the mitochondrially-encoded MT-ATP6 gene are the most frequently implicated, leading to disorders such as Leigh syndrome and NARP (neuropathy, ataxia, and retinitis pigmentosa) syndrome. Advances in mitochondrial genome editing hold promise for treating these conditions, but current approaches remain in early developmental stages and face considerable technical challenges.

 This study explores an alternative therapeutic strategy by introducing a nuclear-encoded, synthetically engineered ATP6 protein designed to enhance ATP production. Our approach is based on the hypothesis that a unique C-terminal extension found in the ATP6 protein of the glacier ice worm (Mesenchytraeus solifugus) significantly boosts ATP synthase activity. Our Work aims to incorporate this extension into human ATP6 can restore mitochondrial function in a cellular model of Leigh syndrome, providing a foundation for future therapeutic applications.

Peck, Lindsay – Program: Ph.D – Prevention Science

A Bridge Over Troubled Waters: The Association Between Participation in a Pre-College Program and Psychological Adjustment and Flourishing in College

The Student Academic Success (SAS) division at Rutgers University–Camden offers a pre-college experience for admitted students in the days leading up to the start of the first semester, called Summer Bridge. The purpose of this experience is to give students knowledge and strategies for success during their first semester in college and beyond, hosting activities that build competencies aligned with the SAS division’s values: a sense of belonging, academic responsibility, identity consciousness, and self-efficacy. Typically, the evaluation of the impact of this type of program focuses on standard metrics for success, such as retention and first-year GPA. In this pilot study, we focused on psychological adjustment and flourishing indicators instead. At the end of their first Fall term, we surveyed Summer Bridge participants on how engaged they were in the Program and how much they felt like they belonged in the Summer Bridge program. We also assessed the following indicators of psychological adjustment and flourishing in the survey: college meaning, hope, existential isolation, and academic self-efficacy. Preliminary findings suggest that engagement and belonging in the Summer Bridge program were associated with finding meaning/purpose in college, stronger hope, reduced feelings that others do not share your subjective experience (i.e., existential isolation), and greater belief in one’s ability to do what it takes to succeed in college (i.e., academic self-efficacy). Finally, we asked students to write about what they enjoyed about the Summer Bridge Program and how they benefited from it. In addition to naming new connections a key outcome, students reported campus navigation and wayfinding as memorable program components contributing to place-based physical belonging. It is essential to understand the psychological benefits of student success programs, in addition to their impact on student retention and achievement, so that we may more wholly capture the experiences of our students and further set them up for success in college and beyond.

Pellegrini, Leah – Program: MFA – Creative Writing

The Four Bears

The Four Bears is an in-progress memoir-in-essays, centered on stories of disobedient desires that disrupt the American fairytale. Scaffolded by a series of recounted dreams that share a ravenous bear motif, the essays explore confrontations with food, home, chronic illness, queerness, and other domains where wild wanting gets (un)tamed. These pieces intersperse memoir with research on everything from the origins of Goldilocks, to rage rooms, to historical masks and beyond.

Pitman, Connor – Program: Ph.D- Biology: Computational and Integrative

  Though protein structure is the conventional link between protein sequence and function, structure is not directly accessible from the genome for all proteins. Predicting or interpreting the effects of mutations often requires identifying the local sequence context. Since protein structures are modular, the “local sequence” is frequently determined by the local secondary structure element. Some proteins, such as intrinsically disordered proteins (IDPs), can only be analyzed using their sequences. In cases like these, detecting innate modularity independent of secondary structure elements using only residue-level properties such as hydrophobicity and charge is extremely valuable. We have previously used contiguous hydrophobicity (“blobulation”) to detect local sequence context for analysis of disease-associated mutations[1]. Here, we detail the blobulation algorithm and demonstrate its ability to detect subsequences associated with hydrophobic environments (core of a globular protein, transmembrane helices, etc.) We also present results from a coevolution analysis showing that these hydrophobic subsequences (“blobs”) have evolutionary significance. We find that pairs of similarly hydrophobic blobs are enriched for coevolving residues. Additionally, we find that within these blobs, the types of coevolving amino acid pairs change depending on the blob containing them, suggesting that blobulation provides a meaningful framework for defining the context around coevolving pairs and could be useful in further coevolution studies.

Poff, Adam – Program: Ph.D- Biology: Computational and Integrative

  Animal development is regulated by the organismal genes. These genes are broadly divided into protein coding and non-coding DNA. The latter in comprised of different functional groups, here we focus on investigating the role of transfer RNA (tRNA) in posttranscriptional gene regulation. Some tRNAs can function as tRNA fragments (tRFs), regulating mRNAs posttranscriptionally in manner similar to microRNAs. The tRNA:Val-CAC-1-1 (ValT) (CR31572) was computationally predicted as tRF in Drosophila melanogaster, targeting 51 genes. In order to discern between the tRNA and tRF activities of ValT, the sevral genomic modification were done using genome engineering, CRISPR/Cas9. These include changing the predicted tRF domain, anticodon domains, deleting the gene (delta ValT), and replacing it with its native form as a control. All substituted ValTs were homozygous viable. To obtain the mRNA profile changes of these backgrounds, all transgenic lines were analyzed by RNAseq. We found groups of genes that were transcribed at different levels that were consistent across mutations in the tRF and others that were consistent across anticodon substitutions. These genes indicate potential targets for future studies into the function of ValT as a tRNA and a tRF.

Rabgie, Mumin – Program: MA – Psychology

  Background: Mental health challenges are prevalent across racial and ethnic groups, but systemic factors and cultural contexts shape the experiences of Black and Latino individuals differently. Black adults face heightened psychological distress due to systemic racism, yet often experience lower rates of formal diagnoses and treatment compared to White adults (White, 2019; Watkins & Johnson, 2018). Similarly, Latino individuals encounter high lifetime prevalence of psychiatric disorders but are hindered by stigma and cultural norms that discourage help-seeking (Alegría et al., 2007; National Alliance on Mental Illness., 2024). Black and Latinos are disproportionately represented among individuals of low socioeconomic status, who also have a heightened risk of poor mental health (Belle & Doucet, 2003). Interpersonal experiences including social support, interpersonal violence, and loneliness significantly influence mental health. Social support can serve as a protective factor for depressive and anxious symptoms, while negative experiences, such as interpersonal violence, and loneliness, may put individuals at risk for these symptoms (Acoba, 2024; Resnick et al., 1997; Richardson et al., 2017; Umberson et al., 2020; Watkins & Johnson, 2018). Loneliness may be a particularly important predictor of mental health symptoms. Despite existing research on these factors, comparative studies examining how these interpersonal experiences differ between Black and Latino populations are scarce. This study aims to address this gap by identifying the associations between these experiences and mental health outcomes among low-income Black and Latinos throughout adulthood.

 Methods: Participants included 452 Black (n = 343) and Latino (n = 109) adults recruited from Camden, NJ. Key variables included social support, interpersonal violence, loneliness, depressive symptoms, and anxious symptoms. Data were analyzed using independent samples t-tests and hierarchical regression models, adjusting for income. Moderation analyses were conducted to examine whether race/ethnicity influenced the associations between interpersonal experiences and mental health outcomes.

 Results: Findings revealed significant associations between interpersonal experiences and mental health outcomes. Loneliness was the strongest predictor of both depressive (B = 0.52, p < .001) and anxious symptoms (B = 0.62, p = .001). Interpersonal violence was significantly associated with depressive symptoms (B = 0.72, p = .01) but not anxiety. Social support did not significantly predict either outcome. Moderation analyses showed that race/ethnicity significantly influenced the relationship between interpersonal violence and anxiety symptoms (B = 0.19, p = .007), with interpersonal violence being a stronger predictor for Black participants. Simple slope analyses confirmed that interpersonal violence significantly predicted anxiety for Black participants (B = 0.10, p = .001), but not for Latinos (B = -0.09, p = .14). No other significant moderating effects of race/ethnicity were observed.

 Conclusion: These findings contribute to an understanding of the role of interpersonal experiences in mental health outcomes among Black and Latino individuals. The results suggest that loneliness and interpersonal violence are particularly relevant interpersonal experiences in understanding mental health. These findings could inform the development of culturally tailored interventions aimed at addressing loneliness and interpersonal violence to reduce mental health disparities. Future research should further explore the mechanisms through which these interpersonal experiences may affect mental health and investigate interventions that can mitigate these challenges

Rager, Sydnie – Program: MA – Psychology

  Signal detection theory (SDT) has been an important theoretical tool in psychology and related disciplines since at least the 1960s. The basic tenet of this approach is to examine how people detect a signal (such as recognizing one’s own name) from background noise (chatter at a party). One of the basic assumptions of SDT is that people accurately learn the nature of a signal from inductive experience (such as a lifetime of hearing one’s own name.) But this assumption has not been adequately tested as most studies of SDT assume people learn such information in an accurate and unbiased way. In the present study we use lines that differ in length as a mechanism for establishing whether or not people learn distributional aspects of a signal to discriminate from noise in a way that provides maximum accuracy. Across two experiments we find that people do not demonstrate the kind of inductive learning that most SDT models propose. The results of this study raise questions on how psychologists have typically analyzed data in SDT experiments and propose better statistical methods for understanding the cognitive mechanisms underlying stimulus discrimination, memory, and judgment.

Richeson, Alexis – Program: Ph.D – Prevention Science

  There has been extensive discourse about new weight loss medications, glucagon-like peptide 1 agonists (GLP-1s), and their potential to induce significant weight loss among users (Doheny, 2023; Vosoughi et al.2022). Past research reveals strong associations between body size and body image (Gruszka, 2022), which may lead some users of GLP-1s to expect an increase in body satisfaction. To date, no research has explored the relationship between body image and interest in using GLP-1s, which is the focus of the current study. Two hundred and twenty-five participants from a university in the Northeastern U.S. (Mage = 20.03; 71.2% women) completed assessments of their awareness and interest regarding GLP-1s using questions developed specifically for this study. Participants also completed measures of body appreciation, body neutrality, self-objectification (body shame and surveillance), weight bias, concerns about weight, and attempts at weight loss; they self-reported their height and weight. Results indicated that participants who were most interested in trying GLP-1s and willing to tolerate their common side effects reported greater body shame, body surveillance, weight concerns, anti-fat bias, disordered eating behaviors, and higher BMIs, as well as lower body appreciation and body neutrality. Simultaneous regression analyses revealed that BMI and recent attempts at weight loss uniquely accounted for variance in interest in GLP-1 use and willingness to tolerate side effects of GLP-1s. Notably, body appreciation moderated some of these findings, suggesting that a higher appreciation for one’s body may serve as a protective factor against interest in using GLP-1s for weight loss.

Riggs, Lindsey – Program: Ph.D- Biology: Computational and Integrative

  A Val66Met mutation in the disease-associated prodomain of brain derived neurotropic factor (BDNF) is known to alter the conformational ensemble of the protein. It was unclear how a small, charge-neutral mutation would affect the conformation or function of the BDNF prodomain, a long intrinsically disordered protein (IDP). Our lab aimed to determine the mechanism behind the conformational shift in the mutated protein (M66). We discovered that this effect is due to the special nature of Met-Met interactions within the protein. This raised the question whether other charge-neutral mutations at site 66 affect the conformation of the BDNF prodomain, or if our results were specific to M66. Here, we present the analysis of 2 μs all-atomistic, explicit solvent, temperature replica-exchange molecular dynamic simulations of the wild-type BDNF prodomain (V66) and mutated BDNF prodomains that contain a charge-neutral, hydrophobic mutation at position 66 of the amino acid sequence (A66, I66, L66, F66, Y66, and M66). We found that, in addition to M66, F66 also increases the radius of gyration of the entire protein, while A66, I66, L66, and Y66 decrease it compared to V66. We also applied our Blobulation method to the BDNF prodomain and found unique interactions between contiguous regions of hydrophobicity. This method was key to identifying the mechanism behind the conformational effects due to M66. These findings demonstrate that the conformational changes in the BDNF prodomain due to the mutations are not specific to M66 and extends to other charge-neutral amino acid mutations. Our results suggest that even IDPs are sensitive to specific interactions between non-aliphatic, hydrophobic residues, which can detectably shift the ensemble toward a more globular conformation.

Rivera Galarza, Jeniska – Program: MA – Psychology

  Background: Adults experiencing homelessness are disproportionately affected by mental health problems, adverse childhood experiences (ACEs), and intimate partner violence (IPV). Little is known about whether there are sex/gender differences in the associations between ACEs or IPV with mental health problems in this population. This study examined whether sex/gender moderated the associations between ten widely studied ACEs or IPV with mental health problems among adults who had been homeless. Methods: Data were used from participants in wave II of the NESARC who identified as experiencing homelessness (n=1,026). Logistic regression models assessed the main effects of ten ACEs individually or IPV on lifetime mental health problems (i.e., mood disorder, anxiety disorder, posttraumatic stress disorder, alcohol use disorder [AUD], drug use disorder [DUD], and nicotine dependence) and attempted suicide while controlling for age and education. Results: Sex/gender moderated 13 0f 77 (17%) associations among adults experiencing homelessness. For most (i.e., 10) of the associations that were moderated by sex/gender, ACEs were associated with greater odds of mental health problems among men. For example, witnessing maternal IPV was associated with at least twice greater odds of alcohol use disorder, drug use disorder, and attempted suicide among men.. Sex/gender did not moderate any associations between ACEs or IPV with mental health problems. Discussion: Findings document sex/gender differences among adults experiencing homelessness that need to be further investigated.

Robin, Taniya – Program: MS – Forensic Science

  Catalytic Hairpin Assembly (CHA) is an enzyme-free, isothermal DNA amplification method that utilizes complementary DNA hairpins to achieve high-efficiency signal amplification. CHA is a promising alternative to traditional polymerase chain reaction (PCR). PCR, the gold standard for forensic DNA profiling, amplifies short tandem repeat (STR) markers by enzymatically replicating specific DNA regions through cyclic denaturation, annealing, and extension steps. Still, challenges remain in improving PCR kinetics, sensitivity, and resistance to nuclease degradation, which are critical factors for forensic DNA fingerprinting. Despite its reliability, PCR requires high-quality DNA samples, and degraded or low-quantity DNA often hinders its efficiency. This study investigates the potential of cationic copolymers, specifically chitosan, to enhance CHA-based amplification for forensic applications. Chitosan, a positively charged polymer, stabilizes DNA interactions, accelerates CHA reaction kinetics, and protects DNA components from enzymatic degradation. The research aims to accelerate CHA process kinetics, improve the specificity and sensitivity of initiator recognition, and enhance the stability of CHA components against nucleases. By integrating cationic copolymers into CHA-based amplification, this study seeks to improve forensic DNA profiling, offering a viable alternative to traditional PCR-based methods for analyzing degraded and trace DNA samples.

Rufai, Basirat – Program: Ph.D- Biology: Computational and Integrative

The research objective was to evaluate the enzymatic activity of ADAM10 on liver cancer exosomes and compare these activities to their pure enzymes. The study utilized nanoparticle tracking analysis to characterize the concentration and size of isolated exosomes from Hep G2 cells and ELISA to determine the concentration of this protease on exosomes. The Enzyme-to-Exo ratio indicates the enrichment of the protease biomarker on Hep G2 exosomes. Assays were done to profile the enzymatic activities and the effect of an inhibitor on the enzymes in exosomes relative to the pure enzymes at the same concentration levels. The study revealed that the protease, when present on the surface of exosomes, has significantly higher activities and is less inhibited than in its free-floating states. The study provides insights into the usefulness of ADAM10, particularly on exosomal surfaces protected by a lipid bilayer, as a diagnostic and prognostic marker for cancer conditions.

Salzer, Regina – Program: Ph.D- Biology: Computational and Integrative

Synthetic polymeric vesicles, known as polymersomes, hold great promise as carrier vesicles for the encapsulation and delivery of cargo. However, controllably releasing cargo on-demand remains a challenge when implementing these vesicles in applications such as drug delivery. The self-assembly of amphiphilic diblock copolymers into vesicles results in the formation of two compartments: an aqueous lumen which allows for the incorporation of hydrophilic cargo, and a bilayer membrane with a hydrophobic region which can encapsulate non-polar cargo. Gold nanoparticles (AuNPs) have been shown to act as photosensitizers in response to pulsed laser irradiation, allowing for light-responsive membrane disruption and subsequent cargo delivery. In this work, two pathways of AuNP incorporation into nano-polymersomes are investigated. In one method, dodecanethiol-functionalized AuNPs are shown to partition into the hydrophobic region of the bilayer membrane of polybutadiene-b-polyethylene oxide nano-polymersomes during self-assembly. It is shown that the concentration of AuNPs present during vesicle self-assembly influences a variety of characteristics such as vesicle size, percent membrane fill, and resistance to surfactants. Furthermore, the concentration of AuNPs correlates with membrane disruption as a function of laser pulse energy, suggesting that greater AuNP concentration increases membrane disruption upon irradiation. The second AuNP incorporation method requires the insertion of a cholesterol membrane anchor to tether desired structures such as AuNPs to the exterior of the polymersome via DNA hybridization. Cholesterol is shown to successfully insert into the polymer membrane both before and after vesicle self-assembly and the amount of cholesterol accepted into the membrane is shown to scale with concentration. The dual purpose of the polymer membrane as both a compartment and an anchoring region may provide a pathway to greater control over both the size and location of AuNPs in these promising photosensitive polymersome cargo delivery systems.

Samba, Bindu Manaswitha – Program: MS – Data Science  

Enhancing iProm-Sigma54:Leveraging RNA LLM and Advanced Machine Learning Models for Improved σ54  Promoter Prediction

The control of bacterial genes is mostly dependent on σ54-dependent promoters, which have an impact on a number of biological activities, including pathogenicity and nitrogen fixation. Drug discovery, biotechnology, and genetic research all depend on the precise identification of these promoters. However, inadequate sequence representation, false positives, and the incapacity to detect long-range relationships in DNA sequences are common problems with current computational approaches. In order to overcome these obstacles, we provide the advanced iProm-Sigma54, a machine learning framework that combines ensemble learning methods, RNA Large Language Models (LLMs), Convolutional Neural Networks (CNNs), and Transformers to produce a highly accurate σ54 promoter prediction. In order to over come these challenges, we came up with the advanced iProm-sigma54, a novel machine learning framework that integrates RNA Large Language Models (LLMs), Convolutional Neural Networks (CNNs), Transformers, and ensemble learning techniques for highly accurate σ54 promoter prediction.

Sandberg, Jesse – Program: Ph.D- Biology: Computational and Integrative

It is now well established that lipids bind specifically to membrane protein transmembrane domains (TMDs) and may allosterically modulate protein function, but measuring a lipid binding affinity experimentally remains a challenge. Coarse-grained molecular dynamics (CG-MD) simulations have been used extensively to study lipid-protein interactions due to the enhanced lipid diffusion and the longer accessible time-scales afforded by a CG model. Nonetheless, a number of conceptual challenges arise when MD trajectories are analyzed, including how to define the ‘bound’ state, and how to differentiate between a bulk lipid versus a specifically bound lipid. Most solutions to these problems have relied on measuring residence times to calculate off-rates, but these quantities are difficult to compare across force fields as well as to experimental data. We previously introduced1,2 the Density-Threshold Affinity (DTA), a method for determining the binding affinity of a lipid for a defined binding site by measuring thermodynamic quantities in unbiased CG-MD. The DTA quantifies the excess density of a defined lipid species in a binding site, compared against a bulk membrane patch of equal area. In the present work, we show how the DTA can be used to quantitatively rank binding sites as well as to determine which lipid species will out-compete the other membrane components for a particular binding site.

Santiago-Mcrae, Ezry – Program: Ph.D- Biology: Computational and Integrative

 Identification of EM-resolved Lipid Fragments using Streamlined Alchemical Free Energy Perturbation (SAFEP)

Over the past ten years, the number of cryo-EM structures with partially resolved lipids has increased dramatically. Lipid binding sites may be important both for native protein function and as potentially drugable sites. Although cryo-EM data can indicate the presence of ordered lipids, those lipids are rarely identifiable from the density alone. We use our Streamlined Alchemical Free Energy Perturbation protocol (SAFEP) for computationally estimating free energies of phospholipid binding (Gbind) to identify lipids and compare affinities across sites and protein conformations. There are several challenges posed by the binding of phospholipids to membrane proteins: 1) lipid flexibility makes traditional restraint schemes less effective, 2) slow relaxation of lipids slows convergence of the free energy estimates, and 3) lipid-water phase separation necessitates careful interpretation of any lipid Gbind. Addressing the first two issues required simple, but highly designed restraints combined with thorough sampling. The phase separation of the system was addressed theoretically by framing the results in terms of binding probabilities – leaving any “standard” binding free energy as an intermediate result. Convergence of all calculations was monitored by several metrics as any one metric was been found to be fallible. We use lipid binding by Erwinia Ligand-Gated Ion Channel (ELIC) as a model system. ELIC is a GABA-activated, prokaryotic member of the pentameric ligand-gated ion channel (pLGIC) protein family. Like other pLGICs, ELIC is known to be modulated by its lipid environment through unknown mechanism(s). By a combination of relative and absolute binding free energy calculations we were able to identify a partially resolved lipid as POPG, determine that POPC is a non-binder to the site, POPE can compete for the site at higher mole fractions, and POPG binds with greater affinity to the open conformation of ELIC.

Schwartz, Ravi – Program: Ph.D- Biology: Computational and Integrative

Controlling water transport across surfaces is essential for all living organisms. Spider egg sacs are multifunctional membranes that protect eggs and spiderlings from the external environment, but past research gives conflicting results about whether or not these mats of silk fibers can reduce evaporation of water in part because the diffusive resistance of any membrane cannot easily be measured independently of the system in which it is studied. We develop a model to describe water vapor transport across porous surfaces that reveals that the diffusive resistance across the gap space underneath the membrane and the boundary layer on the outside of the membrane can play much greater roles in controlling water vapor flux compared to diffusion across the membrane itself. The model accurately predicts diffusive resistance of a variety of synthetic surfaces from empirical studies, as well as the egg sacs of Latrodectus hesperus and Argiope aurantia. We show that “typical” spider egg sac membranes offer surprisingly low diffusive resistance to water because they are highly porous at microscopic scales. However, silk egg sacs still play key roles in controlling water loss by separating the diffusive resistance of the interior of the egg sacs from the outside boundary layer. Our model provides both a tool to explore diverse egg sac geometries and to facilitate comparison of diffusive resistance across membranes studied in diverse systems.

Scott, Mariah – Program: Ph.D – Prevention Science

Students Helping Individuals Facilitate Transition (SHIFT): A Sickle Cell Disease (SCD) Intervention

Purpose: In the United States, physicians and residents report inadequate training in managing adolescents and young adults (AYA) during the transition from pediatric to adult care, in particular, AYAs with chronic illnesses such as sickle cell disease (SCD). Medical students can serve as similar-aged “peers” who can offer informational support and developmentally appropriate social support to AYA patients during the transition period. This report describes the design and preliminary medical student outcomes of a medical student mentor intervention to improve transition education for AYAs with SCD. 

Methods: We developed a medical student mentor intervention that pairs medical students with an AYA with SCD preparing to transition from pediatric to adult care. After training, medical students conducted six-monthly video chat calls with mentees to address transition and disease self-management. Students completed baseline and follow-up surveys measuring attitudes towards chronic illness, SCD knowledge, etc., and qualitative feedback interviews upon completion of the intervention. Feasibility and acceptability were measured by enrollment and retention rates, satisfaction with the intervention, and medical student Attitudes Towards Chronic Illness.  

Results: Nine medical students were paired with a total of 24 patients. Student retention was 100%. Students reported increased knowledge about managing a chronic illness and transition, improved understanding of the patient’s experience navigating the healthcare system and barriers to healthcare, and enhanced patient communication skills. Medical students reported high satisfaction with the intervention. 

Conclusion: A medical student mentor intervention was feasible and acceptable to medical students and may provide an opportunity for value-added medical education. Further research is needed to evaluate the efficacy of this type of student mentoring intervention on both student and patient outcomes. 

Sharma, Naimish and Shetty, Gaurav – Program: MS – Data Science

AI-Powered Custom Loss Function Generation and Beyond for Deep Learning Models Training deep learning models requires careful selection and design of loss functions, activation functions, optimizers, and network architectures, often demanding significant expertise. Our capstone  project aims to automate the generation of custom loss functions using Llama7B, a large language  model, combined with advanced prompt engineering techniques. The proposed system allows users to upload a code snippet and a query describing their optimization objective. Based on this input, the AI driven assistant will generate a custom loss function tailored to the task, along with a unit test to validate its correctness. The system ensures that the generated function adheres to PyTorch’s interface and performs robustly under testing conditions. Beyond loss functions, our approach and architecture can be extended to other key aspects of machine learning, such as activation functions, optimizer functions, and model architecture decisions (e.g., selecting the number of layers in a neural network). By integrating LLMs for automated code generation, our solution reduces the complexity of deep learning development, making it more accessible to researchers and practitioners. Additionally, this framework aligns with broader AI-assisted model training innovations, such as meta-learning and automated model architecture selection. Through this novel approach, our project streamlines deep learning workflows, enhances efficiency, and minimizes reliance on manual coding expertise, paving the way for more intelligent and adaptive AI  driven model development.

Siddhabathula, Aryan – Program: MS – Data Science

Real-Time Casual Ads Optimization: An Integrated Web-Based Data Collection and Automated Analysis Pipeline

Part 1: Data Collection and Ad Display Web Application

This phase of the project focuses on designing and implementing a web application that not only displays advertisements to users but also captures real-time interaction data (impressions, clicks, contextual details). Developed with Java and Spring Boot (or a similar framework), the application logs each ad event alongside critical metadata- such as session information, device type, and time of day- and stores the information in a structured database. By ensuring high-quality, real-time data capture, Part 1 provides the foundation for the automated causal analysis in Part 2. This streamlined data pipeline enables the immediate availability of fresh insights, supporting a continuous feedback loop that informs ad targeting and budget allocation decisions as soon as new data is collected.

Part 2: Real-Time Casual Analysis and Automated Ads Optimization

Building on the data collection pipeline from Part 1, this phase integrates both internal ad interaction logs and external datasets (such as the Criteo Attribution Dataset) to conduct real-time, automated causal analysis for ads optimization. By employing advanced causal inference and machine learning techniques, we accurately estimate advertisements’ incremental impact (uplift)on key metrics like conversions and user engagement – rather than relying on correlation-based approaches. The automated pipeline continuously ingests new data, recalibrates casual models, and provides actionable recommendations in near real-time. This enables marketers to dynamically refine targeting strategies, allocate budgets more effectively, and maximize return on investment through data-driven decision making.

Siner, Aleece – Program: MS – Biology

Organisms adapt to seasonal changes caused by Earth’s rotation around the sun, with photoperiods influencing their biological cycles. While photoperiodism in plants is well-documented, less is understood about fungi. This research focuses on Neurospora crassa, a model organism for photoperiodic studies, to explore how fungi adapt their reproductive cycles to seasonal changes. Through experimental manipulation of light and dark phases, we hypothesize that fungi utilize a unique mechanism, distinct from plants, to measure photoperiods. Understanding this mechanism can shed light on broader circadian and circannual rhythms across eukaryotic organisms, with implications for human health.

Soomiyol, Mnguashima Valentina – Program: Ph.D – Public Affairs

The Intersection Between Gender, Programs, and Economic Development. A Case Study of Puerto Rico

Women’s empowerment has long been recognized as a cornerstone of sustainable economic development. However, the journey toward gender equality continues to be challenged, particularly within specific cultural and regional contexts. Puerto Rico offers a compelling lens to explore the intersection of gender, gender-sensitive programs, and economic development. Puerto Rican women have shown remarkable resilience despite the structural challenges they encounter. While they make up 58% of university graduates, they continually face systemic inequalities, earning $0.82 for every dollar that men earn in comparable roles and occupying only 20% of executive positions. Moreso, domestic violence remains a significant barrier, with one in three women in Puerto Rico experiencing physical or sexual violence in their lifetime. This study aimed to fill these gaps by exploring the intersection of the glass ceiling effect, women’s under-representation, the inadequacies of gender-sensitive programs, and the systemic challenges women face in contributing to Puerto Rico’s economic development.  

Utilizing a case study methodology, appreciative inquiry framework, and asset mapping, the study highlights how women-led initiatives, such as the LEAP Academy and Taller Comunidad La Goyco, serve as exemplary models for fostering resilience, empowerment, and sustainable community development. The study inquiry is grounded in the Capability Approach and Community Capital Framework. The study’s Findings demonstrate that LEAP Academy embodies the principles of the Capability Approach and the Community Capital Framework by actively discovering and leveraging community assets to promote education, empowerment, and economic development in Puerto Rico. Additionally, initiatives like Taller Comunidad in La Goyco have focused on cultural preservation and environmental sustainability, providing resources for recycling education and tackling environmental justice issues.   

More importantly, LEAP social enterprise exemplifies a holistic approach to education, functioning as a beacon of hope and improving the capabilities and well-being of women and families in Puerto Rico. By empowering victims of domestic violence, LEAP provides economic opportunities and restores agency and freedom. These findings highlight the need for gender-sensitive programs and scaling holistic models like the LEAP Academy to address the interconnected challenges of gender inequality and economic development. Though often undervalued, these contributions highlight women as critical agents of economic transformation in Puerto Rico. 

Springate, Violett – Program: Ph.D – Prevention Science

This study aimed to explore the relationships between pornography exposure, anti-fat attitudes, and gender beliefs. Specifically, we examined whether early exposure to pornography is associated with more frequent pornography use, negative perceptions of women, and increases in weight stigma. Participants were asked to answer a series of questions about their exposure to sexual media, beliefs about gender, and attitudes about their own weight and the weight of others. We found a difference in pornography frequency scores by gender. Females were significantly less likely to report pornography consumption. Additionally, we found a positive association with porn viewing and anti-fat attitudes and a negative association with benevolent sexism.

Stolte, Matilda – Program: MFA – Creative Writing

Lux et Lux

My novel Lux et Lex follows a cast of characters in Tremont, a fictional liberal arts college in Lancaster where two bodies have been found buried beneath a historical-house-turned-fraternity. It centers around Grace who is struggling with her newfound open relationship with her girlfriend Quinn. Grace is a recent graduate working with the archivist in the college’s library. Her main project is archiving photos and documents from Miss Tremont, a beauty pageant held on Tremont’s campus from 1920s to the early 1990s where young girls in need competed for a scholarship to the college. When the bodies are found and protest erupts on campus, Grace is forced to move her focus to the house and the controversial man who lived there.

Stott, Helen – Program: Ph.D- Biology: Computational and Integrative

Animal development is controlled by a tightly regulated network of gene expression where genes are often reused across multiple life stages and tissues (Cardoso-Moreira et al., 2020). The signaling pathways and genes making up this network show deep conservation across the animal family tree (Carroll, 2008). Our current understanding of the underlying mechanisms regulating the same gene in multiple tissues is limited to a few genes. Recent studies have also demonstrated that paralogous genes with co- expression can share cis-regulatory modules (CRMs) (Baudouin- Gonzalez et al., 2017; Bourbon et al., 2022, Lan and Pritchard 2016;

 Levo et al. 2022; Loker and Mann, 2022; Stevens et al. 2022). The T-box (Tbx) transcription factors are an ancient family of genes in animals (Sebé-Pedrós et al., 2013). The Drosophila Tbx20 homologs midline (mid) and H15 demonstrate conserved functions across arthropods, with roles in the determination of posterior fate during oogenesis, axis formation in the developing limbs, specification of neurons, and formation of the heart tube in developing embryos (Buescher et al., 2004; Buescher et al., 2006; Miskolczi-McCallum et al., 2005; Prpic et al., 2005; Pyrowolakis et al., 2017; Svendsen et al., 2009). Additionally, these two genes sit within an isolated topologically associating domain (TAD) that contains no other genes, making mid/H15 an excellent system for expanding our understanding of developmental gene regulation. We explored the role of CRM proximity and CRM duplication on mid/H15 patterning during oogenesis. Using CRISPR-Cas9 gene editing, moved the G04 CRM from its position in front of mid to an analogous position in front of H15 in both wild type genetic backgrounds and G04 deletion backgrounds. Impacts from the duplication and relocation of G04 expanded H15 expression during oogenesis while also reducing mid expression. Flys also exhibited ovulation defects.

Sunnolia, Eryn – Program: MFA – Creative Writing

The Small Bird in My Fist

I will share from my memoir-in-progress The Small Bird in My Fist. The book is a mid-twenties queer coming-of-age story that follows my polyamorous experimentation, unfolding queerness, and subsequent expulsion from my evangelical Christian family.

At seven years old, I was baptized in a lake in southern Maryland. Almost twenty years later, my boyfriend and I tried polyamory so I could explore my sexuality. Enter Quinn: polyamorous, non-binary, and living the queer life I never imagined possible for myself. As I unboxed my queerness from the basement of my body, my relationship with my boyfriend dissolved. What resulted was three different love stories, each following their own trajectory and impacting each other in unexpected ways: my relationships with Quinn, my mom, and my dad.

As I fell more deeply in love with Quinn, I was forced to reckon with the combustible mixture of fear and love playing out in my life, my mom’s life, and my dad’s life. In the space my parents left behind, a tether between us remained: the desire each of us have to control the uncontrollable. My mom wanted me to stop being queer in order to save me from hell. My dad, hurt by his own childhood abandonment, saw my queerness as another abandonment and wanted his power back. In the wake of my expulsion from the family I clung to Quinn as a lifeboat, so terrified they would leave me too that I was squeezing the life out of our relationship. I had to choose: I could love Quinn the way my parents loved me, which was about trying to control me, change me. Or I could try to love Quinn in a way that was about their freedom and mine.

Trieu, Ethan – Program: MA- Psychology

In the quest for personal growth and well-being, the decision to seek help can be transformative, yet many individuals hesitate to take this crucial step. Seeking help can be difficult for anyone and understanding ways in which we can overcome these obstacles can allow us to get the help that we need. Self-efficacy has been seen to impact help seeking attitudes as efficacy gives people the belief and inspires people to act, which may include reaching out and asking for assistance. Understanding this avenue to help seeking, the proposed research aims to test the impact of a meaning-centered intervention based on college education and the relationship it has on students’ inclination to seek help through self-efficacy. By investigating the relationship between a sense of meaning and attitudes toward help-seeking behaviors, this study seeks to uncover strategies to encourage students to overcome barriers and embrace seeking assistance, thereby promoting their academic success and goal achievement.

Vega, Samantha – Program: MS – Prevention Science

 Three Mile Island Nuclear Aftermath: Elevated Depression Rates and Future Implications

The 1979 Three Mile Island (TMI) nuclear accident left a lasting imprint on public perceptions of nuclear safety and the psychological well-being of affected communities. This study examines the long-term impact of the TMI disaster, focusing on spatial variability in depression prevalence within 5-, 10-, and 50-mile radii surrounding the nuclear site. Using GIS analysis and regression modeling with data from the CDC PLACES dataset, we assess whether proximity to TMI is significantly correlated with higher depression rates. Our findings indicate that depression prevalence is highest within 5 miles of TMI, decreases at 10 miles, and plateaus beyond 30 miles, suggesting a proximity effect where psychological distress is exacerbated by perceived exposure risks and historical trauma.

 Additionally, this study contextualizes these findings within broader environmental justice concerns, highlighting the intersection of mental health disparities and Superfund site contamination. We also explore the contemporary implications of repurposing nuclear sites, such as Microsoft’s initiative to power artificial intelligence operations with nuclear energy, and the ethical considerations tied to public trust, transparency, and community engagement. By integrating historical data with modern energy trends, this research underscores the need for interdisciplinary approaches that bridge prevention science, environmental health, and technological innovation. These findings advocate for policies that prioritize mental health resources, equitable energy transitions, and community-driven decision-making to mitigate the compounded risks faced by vulnerable populations.

Warrington, Charles – Program: MS – Chemistry

 Living systems have evolved self-healing to protect the organized structures they rely on from disruption and restore damage when it occurs. As biotechnology and medical development evolves there is a growing need for a bioinspired healing design and materials approach for applications in therapeutics, synthetic biology, and tissue engineering. 

 Biomaterials are not only frequently benign to life but with an increase in implantable medical technology the fields of chemistry, biology, and engineering find themselves facing similar problems life evolved to overcome, namely a lack of durability and requirement for repair. 

 These devices can be made more resilient by understanding the chemical properties and molecular mechanisms of these self-healing biomaterials. This understanding aids in the development of new synthetic and non-cellular approaches to produce devices made of self-healing biomaterials. Each of which can safeguard the functions of these increasingly vital technologies.

Weightman, Ryan – Program: Ph.D- Biology: Computational and Integrative

Epidemiological models are essential for understanding, predicting, and mitigating the spread of infectious diseases. The COVID-19 pandemic underscored the importance of robust modeling approaches that account for urbanization, population demographics, travel behavior, and migration patterns. Commonly used models include Agent-Based Modeling (ABM), which captures individual-level heterogeneity and complex interactions, and Equation-Based Modeling (EBM), which efficiently handles large-scale aggregate data. However, ABMs can be computationally intensive, and EBMs may oversimplify nuanced interactions.

 To address these challenges and improve city resilience to epidemics, we propose the development of a hybrid city planning tool that hierarchically integrates ABM and EBM methodologies. This tool uses two primary data sources: city sociodemographic data to simulate agent-based urban mobility patterns, and virus-specific parameters (e.g., infection rates, recovery rates, severity factors) to model epidemic propagation. Additionally, probabilistic methods are incorporated to manage data uncertainties, such as missing data points or irregularly occurring travel patterns, ensuring robust and reliable predictions.

Westley Henson, Kiersten – Program: Ph.D – Prevention Science

Cancer preventative screenings are essential for early detection and improved patient outcomes, yet disparities persist in Southern New Jersey, particularly in rural and underserved communities. This poster examines two projects aimed at increasing lung and colorectal cancer (CRC) screening rates in Southern New Jersey leveraging provider education, patient navigation, and process improvements to address screening barriers and increase knowledge.

 The pilot project included a lung training program that was implemented to enhance lung cancer screening rates by educating ordering physicians and nurse navigators. Over the course of the study, 203 individuals (102 male, 101 female) underwent lung cancer screenings, while 55 individuals declined participation. Following the training, referrals to the Quit Center increased dramatically from 45 to 165, with 59 patients completing assessments. A lung navigator facilitated scheduling improvements, demographic-targeted outreach, and direct provider appointment access. The introduction of a Shared-Decision Making (SDM) DotPhrase in the electronic medical record (EMR) system streamlined documentation, ensuring compliance with screening guidelines. These interventions resulted in a significant increase in lung cancer screenings, demonstrating the effectiveness of workflow optimizations and provider education in encouraging preventative screenings. However, the population was not diverse or representative of the Southern New Jersey population. 

 Building upon these insights, there is an ongoing colorectal cancer screening project focuses on high-risk, underserved populations, including Black, Hispanic/Latino, Nanticoke-Lenni Lenape, undocumented, uninsured, and low-income individuals. This multi-level study aims to improve CRC screening rates through patient navigation and provider education at clinical sites in Southern New Jersey. The current CRC study employs surveys and focus groups to assess barriers to screening and the impact of provider training on screening completion rates. The project’s primary objectives include surveying current providers on CRC eligibility criteria and screening tools, surveying current medical students at Rowan School of Medicine pre and post education, surveying and conducting focus groups with community members on their experiences, perspectives, barriers and knowledge surrounding CRC screening. 

 This work highlights the importance of structured provider education, patient navigation, and enhanced outreach strategies in increasing screening rates and facilitating early cancer detection. However, more research and sustained efforts are crucial to address persistent disparities. While progress has been made, some of this work is ongoing and must continue to evolve to meet the needs of the South Jersey community. There is an urgent need for increased funding, policy advocacy, and community engagement to ensure preventative cancer screenings reach all eligible individuals, particularly in historically underserved areas. Future efforts must scale provider training, expand patient navigation programs, enhance culturally competent community outreach, and advocate for policy changes that support screening accessibility. Without continued research and investment in preventative screenings, high-risk populations in Southern New Jersey will remain disproportionately affected by late-stage cancer diagnoses and poorer health outcomes.

Yue, Chengcheng – Program: Ph.D – Public Affairs

  The relationship between health expenditure and health outcomes is highly context-dependent, shaped by the unique political, social, and governance structures within democratic systems. While various studies have explored this issue, research specifically focused on OECD European countries remains limited. This study examines how the quality of democracy moderates the impact of health expenditure on health outcomes using a fixed effects model. A panel dataset covering 27 OECD European countries from 2000 to 2015 is constructed, drawing data from sources including the Economist Group, Global Welfare, World Bank, International Monetary Fund, World Health Organization, and the OECD. Democracy quality is measured using the Democracy Index developed by the Economist Group, which consists of five key components: electoral process and pluralism, functioning of government, political participation, political culture, and civil liberties. Health expenditure, the main independent variable, is assessed as domestic general government healthcare expenditure as a percentage of GDP, while health outcomes—the dependent variables—include life expectancy at birth, number of hospitals per million population, number of physicians per 1,000 population, pharmaceutical sales, and perceived health status. By analyzing both the overall Democracy Index and its individual components, this study provides a nuanced understanding of how democracy quality moderates the relationship between health expenditure and health outcomes, offering insights into the role of governance in shaping public health effectiveness.

Zhong, Nelson – Program: Ph.D – Public Affairs

 Unearthing Perspectives: A Mixed-Method Exploration of Forging the Right to a City in Distress

Persistent-poverty and distressed communities are in many ways the epitome of left-behind places. Seemingly detached from the nation’s broader economic growth, the question becomes, how can residents establish or gain a “right to the city” in distress? Using a community-engaged framework, this study aims to explore this phenomenon by investigating community well-being and quality of life in Camden, New Jersey – a city that has been insolvent since 1985. Using quality of urban life indicators, we seek to answer the research question: “How do objective, subjective, and behavioral quality of life indicators influence community well-being in a distressed city?” To achieve this, we apply a mixed-method approach that merges GIS with both objective and subjective spatial data, comparative economic well-being data, and residents’ subjective assessments gathered from focus groups and surveys to pinpoint areas of growth, strength, and potential improvement. We find that community well-being and certain key quality-of-life indicators are necessary to realize the “”right to the city”” because they encompass the multidimensional aspects of residents’ lives that make a city livable, equitable, and democratic. We conclude by providing strategies for community building and policy interventions that help make the right to the city possible in even the most distressing of contexts.”

Zmijewski, Sebastian – Program: Ph.D- Biology: Computational and Integrative

Polymersomes hold significant potential for drug delivery applications due to their ability to effectively encapsulate both hydrophobic and hydrophilic molecules. The capability to control cargo release from these vesicles with high spatiotemporal precision would be transformative for biological systems. One method to achieve this is through a light-responsive nano-polymersome system, where pulsed laser irradiation can be utilized to induce membrane disruption and subsequent cargo release. Light-responsive nano-polymersomes can be formed by incorporating dodecanethiol-functionalized gold nanoparticles. Gold nanoparticles serve as optimal photosensitizers due to their strong interactions with light, stemming from their surface plasmon resonance (SPR). The interaction between light and free electrons results in the thermomechanical release of energy, which causes membrane disruption. Previous research has demonstrated the capacity of photosensitized micron-polymersomes and nano-polymersomes to enable controlled cargo release in response to pulsed laser irradiations. Nano-polymersome uptake and toxicity were investigated in NIH 3T3 cells, along with bulk picosecond irradiations of cells containing endocytosed polymersomes. The uptake of nano-polymersomes was analyzed using confocal microscopy with fluorescence and confocal laser scanning microscopy. Polymersomes were incubated with cells that were subsequently stained with either NucBlue or Cellmask and imaged with a z-stack to confirm the endocytosis of polymersomes. The toxicity of polymersomes was assessed by incubating NIH-3T3 cells with a concentrated polymersome solution and utilizing an LDH assay post-incubation to evaluate potential toxicity. Methods were developed for cell irradiations using a 532 nm picosecond laser, and cytotoxicity was studied as a function of irradiation energy. Additionally, irradiations of cells with endocytosed doxorubicin-loaded polymersomes doxorubicin were performed. These studies suggest that polymersomes are readily uptaken into NIG 3T3 cells with minimal toxicity. Additionally, it was demonstrated that picosecond laser irradiation of cells in the absence of vesicles also results in minimal toxicity. Finally, cytotoxic doxorubicin was successfully released upon irradiation from uptaken vesicles as demonstrated by cell death.