草榴社区

T lymphocytes play a vital role in immunity to pathogens and cancer. These cells recognize foreign proteins through their antigen receptor (also called as T cell receptor), and mount an immune response for the body鈥檚 defense. While pathogens trigger a robust immune response, tumor cells, which are mostly self-tissues, do not. Aiding helper and cytotoxic T lymphocytes to generate a strong adaptive response against all types of tumors will aid in the cure for cancer. Lymphocytes, after sensing a foreign protein, undergo cell division and expand in numbers to generate billions of clones for immune defense. Many of these lymphocytes travel inside the tumor (tumor-infiltrating lymphocytes - TILs) to fight the tumor. The tumor microenvironment incapacitates these TILs, resulting in their exhaustion and inactivity. Our laboratory is currently studying the role of Ly-6A protein and its contribution to T cell activity/inactivity. We are using a mouse tumor transplantation model, where mouse melanoma cancer cells are transplanted to the skin of genetically defined mouse strains to induce tumors.  TILs from the tumors generated in Ly-6A-deficient and sufficient mice are being analyzed for their capacity to respond to the tumor tissues. These studies will provide insights into the mechanisms underlying TILs to fight and eliminate the solid tumor.

This project examines the educational effects and student perceptions of the Immediate Feedback Assessment Technique (IF-AT) when used for group quizzes in a one-semester General, Organic, and Biochemistry (GOB) course taken by nursing majors. IF-AT scratch cards provide item-level, answer-until-correct feedback that permits calibrated partial credit and immediate learning during assessment. While IF-AT has documented benefits in STEM classrooms (greater discrimination of partial knowledge, error correction, and positive student affect), less is known about its impact specifically for nursing students in a GOB course where foundational knowledge directly supports clinical reasoning and licensure preparation.  
 
During the Spring 2026 semester, IF-AT will be implemented for low-stakes, team-based quizzes distributed across the term. The study will compare student performance on IF-AT group quizzes with matched group-control quizzes (identical group conditions using parallel item forms) and will measure individual retention using a follow-up individual assessment. Additionally, student perception data will be collected via a short, validated survey and brief focus prompts to capture immediacy, perceived learning, anxiety, and fairness. The project鈥檚 mixed-methods approach combines quantitative analysis of scores (item and test level, partial-credit effects, item discrimination) with qualitative analysis of open responses and focus-group comments.  
 
A collaborating first-year Match research assistant will help administer IF-AT quizzes, orient groups to proper use of cards and scoring, manage consenting and de-identification procedures, collect and aggregate quiz and survey data, and assist with data cleaning and analysis. Findings will inform whether IF-AT boosts short-term learning and student engagement in nursing cohorts, provide recommendations about partial-credit schemes, and produce practical guidance for scaling IF-AT in allied-health curricula. Deliverables include a final report summarizing study findings and implementation recommendations.

Scholars in sustainable human-computer interaction (HCI), have begun using a social practice theory approach to explore how people change their routines and behaviors to be more sustainable.  In particular, Soden et al. (2025) have written about what they call 鈥淐limate Data Practices,鈥� emphasizing 鈥渢he situated socio-technical practices that construct and animate鈥� climate data. They argue that there is a need to broaden considerations about what climate data is, who produces it, and how it is used in response to the climate crisis.  

At the same time, there is also a growing interest in the ways that social media (and the massive amounts of data it generates), is shaping culture and practice.  For instance, our previous work examined the #Haul phenomenon on YouTube (Cooney et al. 2024). This project aims to question the way we think about climate data practices on social media.  In particular, does intention matter when promoting sustainable practices?  

The 鈥渢rad wife鈥� has recently become a popular internet archetype, where women (wives) advocate for a return to 鈥渢raditional鈥� marriage and family values.  On the one hand, the politics espoused by trad wife influencers are miles away from the typical understanding of climate change or sustainability advocates. On the other hand, many trad wife influencers also promote 鈥渉omesteading鈥� practices like gardening and canning, mending and cleaning with natural ingredients, many of which have climate benefits.  

In this project, we will use social media data (YouTube, Instagram, TikTok) to examine both the practices and rhetoric of #tradwife influencers on social media, using a climate data practices lens.  The ultimate goal is to understand the role of values and intentionality versus practice in capturing climate data in the influencer age.

The rapid development of generative AI models has resulted in a wide array of programs capable of producing realistic images from input prompts. While many AI detectors are available online, their design and implementation are typically proprietary, and there is a lack of publicly available, comparative evaluations of their performance.

In this project, we aim to evaluate various models used for detecting AI-generated images. Our goal is to assess the current state of the art in AI image detection, identifying the strengths and limitations of different approaches. To build a robust evaluation framework, we will leverage existing benchmark datasets and, if necessary, create or collect new ones.

We will employ open-source image classification models, including pre-trained ones. These will encompass deep learning-based models such as Convolutional Neural Networks (CNNs) and Vision Transformers, as well as feature-based models and traditional machine learning techniques like Support Vector Machines (SVMs) and shallow neural networks.

By the end of the project, we will present comparative performance results for each classification model in detecting AI-generated images. Additionally, we will analyze which types of generated images are more challenging to detect.

Moneyball is the name of the game in modern professional baseball. Clubs are interested in identifying statistical patterns in a player's game that could predict major league success. 

Every minor league player's dream is to get to "The Show", professional baseball, MLB. Back in the day, promising players would be observed by professional scouts. The scouts would make qualitative judgments ("he has a good swing") and combine those judgments into a prediction of major league success. Today, things are a bit more quantitative ("he has an average spin rate of 2500RPM"). One reason for this shift is the introduction of expensive camera and sensor setups in major league parks that take these measurements. But, as a computer vision researcher, I am interested in figuring out if some of these measurements can be made from basic videos of a player. 

This project is focused on one element of those measurements: given a video of a single pitch, can we identify the frame in which the pitcher releases the ball?    

The Weston Lab works on questions around the resilience of tidal wetland ecosystems in the face of climate change and land use change. Tidal wetlands provide many critical ecosystem services, but many wetland systems are threatened by sea-level rise, changing inputs from the watershed, and other impacts of human activities. The First Year Match students will work with the Weston lab on research in support of several externally-funded projects that examine carbon and nitrogen cycling in coastal wetland ecosystems and the response of coastal wetlands to sea-level rise and changing land use. The laboratory conducts fieldwork in the Spring, Summer, and Fall in Plum Island Sound, MA and more locally in the Delaware River estuary. Samples collected in the field are brought to the laboratory for processing.

Over the last decade, landfalling U.S. tropical cyclones (e.g., tropical storms and hurricanes) have caused over $820 billion in losses. In 2017 alone, landfalling tropical cyclones were responsible for nearly $340 billion in losses and 3,167 fatalities, making it the costliest and deadliest year in the last four decades. While tropical cyclone impacts and fatalities are most commonly linked to the larger parent storm, tropical cyclones produce a variety of hazards, such as inland precipitation that can lead to freshwater and/or flash flooding, coastal storm surge, damaging winds, rip currents, dangerous offshore marine conditions, and tropical cyclone tornadoes. Most notably, storm surge associated with Hurricane Katrina (2005) devastated New Orleans, Louisiana, causing $200 billion in losses and killing over 1,500 people. Before Hurricane Katrina, tropical cyclones were primarily thought of as being isolated wind threats to coastal populations. This belief was reinforced by Hurricane Andrew in 1992 when its winds destroyed much of southern Florida. However, Katrina and several other tropical cyclones since Andrew (e.g., Florence, Harvey, Ida, Helene, Sandy) have illustrated that water hazards such as inland flooding and storm surge can wreak havoc on inland communities far removed from the Atlantic and Gulf Coasts. 

This project seeks to determine the fatality causes by hazard type for all landfalling U.S. tropical cyclones from 1995 to 2025. The student working on this project will compile information and statistics on past tropical cyclone deaths so that temporal and statistical analyses can be conducted. Results from this work will provide a more holistic understanding of the relationships between tropical cyclone hazard type and fatalities. Outcomes from this work will be used to inform emergency managers, policymakers, and stakeholders about the tropical cyclone hazards that pose the greatest threat to vulnerable populations.

The trade of reptiles and amphibians for exotic pets is an expansive industry that threatens biodiversity. While much of this trade is both legal and sustainable, it can simultaneously contribute to species declines by incentivizing overexploitation of wild populations, facilitating the spread of invasive species, and acting as a vector for the transmission of zoonotic pathogens. In the United States, reptiles and amphibians are frequently sold at trade shows, exotic pet stores, and online. This project seeks to characterize the species diversity and potential ecological risks of exotic reptile and amphibian trade in the northeastern United States as part of a larger monitoring effort. Information on species composition, selling prices, the geographic distribution of actors in the trade, and ecological risk factors will be collected from trade shows, online retailers, and pet stores will be compared to existing large datasets on exotic reptile and amphibian trade.

Each year, the world produces over 400 million tons of plastic products. Through management, much of these plastics are delivered to waterways via stormwater runoff. Upon entering a riverine environment, plastics can physically and chemically degrade into smaller particles, which can then be ingested by aquatic environmental and/or birds. Additionally, these plastic particles can leach harmful metals and organic compounds both during transport and once ingested by organisms. Understanding the types of plastic materials that make their way to waterways and their potential environmental impact can better inform conservation and regulatory practices in upstream areas. 

For this study, we will both characterize macroplastics (>5mm diameter; e.g., LDPE, HDPE, PE, etc.) found in streams as well as their ability to leach heavy metals, such as cadmium, copper, lead, nickel, and zinc. Characterization of plastics will be determined using Fourier transform infrared spectroscopy. Total metal content of macroplastics will be determined using X-ray fluorescence spectrometry, while the leachable fraction will be determined using acid digestion and inductively coupled plasma mass spectrometry. Finally, relative differences in the types of plastics found in streams (e.g., water bottles, eating utensils, grocery bags, etc.) will be linked to landcover practices in the upstream areas.

Mathematical modeling is a useful technique for describing dynamics and testing hypotheses that may be difficult or impossible to explore in reality. Because real-world systems are complex, models must simplify reality by focusing on key factors while excluding all other factors. Even with these simplifying assumptions, mathematical models can deepen scientific understanding, test the effects of change, and inform decision-making. However, different modeling approaches can lead to varying predictions, making it imperative to compare models and ensure that results are not overly dependent on their structure. 

This project investigates the transmission of Clostridioides difficile, one of the most frequently identified healthcare-acquired infections in United States hospitals. Both symptomatic and asymptomatic patients shed C. difficile endospores that survive for long periods on surfaces and resist many commonly used disinfectants. Transmission often occurs through contact with contaminated surfaces or healthcare workers (HCWs) carrying spores between patients. 

We developed a baseline model to quantify these pathways and found that 63% of the colonizations came from an interaction with HCWs, supporting the hypothesis that poor HCW hand hygiene is a major factor in C. difficile transmission. The next phase of this project will investigate whether these results are influenced by the current model structure by comparing our baseline approach (which assumes average contamination levels across all HCWs) with an alternate structure that distinguishes between 鈥渃ontaminated鈥� and 鈥渦ncontaminated鈥� HCWs. 

Light and matter interact all around us, but these interactions are usually weak. However, when molecules are placed inside specially designed photonic structures called optical cavities, the interaction between light and molecules can be enhanced. This can lead to new and interesting behaviors, especially when the system is pushed out of equilibrium, for example, by changing the light intensity or the molecular configurations.

In this project, we will investigate the dynamics of molecules strongly coupled to light, using classical simulations, with the aim to uncover how molecular motion, energy exchange, and reaction pathways are influenced by the presence of confined light fields in optical cavities. We will model molecules as classical harmonic and anharmonic oscillators, and simulate their interactions with electromagnetic fields using tools like Wolfram Mathematica or Python. Some of the key questions we will investigate include:

1. How does the presence of a cavity photon field affect molecular vibrations?
2. Can light steer molecules toward specific configurations or reactions?
3. What kinds of patterns emerge when many molecules interact with the same light field?

This research is part of a growing field called polaritonic chemistry, which looks at how light can be used to control chemical and physical processes. The project provides a great opportunity for a first-year student to build skills in computational physics and chemistry, learn about optical cavities and light-matter interactions, and contribute to an exciting area of science with potential applications in materials design, chemical control, and energy technologies.

No prior experience with light-matter interactions is required. Strong motivation, curiosity, a willingness to learn, and some comfort with math and programming. The results of this work could lead to new insights in physics and chemistry and may even contribute to a future publication.

Atomic Force Microscopy (AFM) has emerged as an indispensable tool in the study of low-dimensional materials of both hard and soft matter. In this study, we use AFM to study human ferritin heteropolymers overexpressed in E. coli bacteria by genetic engineering techniques. Ferritin is the iron storage protein in living systems where iron, an essential element for life, is sequestered within a spherical protein shell of 8nm inner and 12nm outer diameter, composed of two types of amino-acid chains, H and L. These two types of chains play complementary roles in iron reduction, nucleation and nanoparticle growth within the shell. H-rich ferritins are associated with heart and brain tissue, where fast iron trafficking occurs, while L-rich ferritins are associated with liver and lung tissue, for long-term iron storage. Ferritin malfunction is associated with a host of iron-related disorders, including neurological disorders. Thus, ferritin is of great interest to the physiology of iron homeostasis in health and disease. In addition to ferrihydrite, other metal and metal oxide nanoparticles can be grown within the protein shell, which presents itself as a robust nanotemplate to produce monodispersed nanoparticles, of interest to nanoscience and nanoengineering (nanobiomedicine, targeted drug delivery, nano-architectural designs for device applications, etc.). 

The nanomechanical properties of heteropolymeric shells will be studied by investigating the elasticity, deformation and Young鈥檚 modulus of H-rich and L-rich heteropolymers in the absence and presence of a ferrihydrite core. The student will be trained in the preparation of biological samples for AFM investigations and the operation of this diverse, multimode microscope to obtain topographical maps of ferritins and measure the elastic properties of various ferritin samples. This study is of relevance to biology, bioengineering and nanotechnology.

A supernova is the dramatic explosion of a star at the end of its life, briefly outshining an entire galaxy. Type Ia supernovae occur when a white dwarf star in a binary system undergoes runaway nuclear fusion, producing a very bright and predictable explosion. Because these supernovae all reach nearly the same peak brightness, astronomers can use them as 鈥渟tandard candles鈥� to measure distances across the universe. When combined with redshift measurements鈥攖he stretching of light to longer wavelengths as galaxies move away from us鈥攖hese distances provide direct evidence that the universe is expanding. This discovery, first made in the early 20th century, is the foundation for modern cosmology.

X-ray binaries (XRBs) are systems where a normal star orbits a compact object such as a black hole or neutron star. Matter pulled from the star forms a hot accretion disk around the compact object, producing intense X-ray emission. These systems are highly dynamic鈥攖he disk, corona, and even relativistic jets change over time鈥攍eading to complex variability in their brightness. Astrophysicists study these variations to understand the physical processes in the accretion environment.

In this project, the student will study how the X-ray emission from XRBs emulates the damped and driven 鈥淒uffing鈥� oscillator. The Duffing oscillator can exhibit what we call 鈥渃haotic鈥� behavior: motion that appears unpredictable but follows deterministic equations of motion. Previous studies have found that the Duffing oscillator can be used as a mathematical model to describe how the X-ray emission coming from XRBs changes over time. We will compare various solutions to the Duffing equations to X-ray data from various XRBs in our galaxy. This will involve coding in Python and working with data from X-ray telescopes.

Histone modifications are epigenetic changes that can impact gene expression in the absence of changes to the DNA sequence. Several distinct chemical groups, including methyl groups and acetyl groups, have long been known to become attached to histone proteins and to influence chromatin structure and gene expression. However, recent evidence suggests that serotonin, which is most commonly known as a neurotransmitter, can also be covalently attached to histones and that histone serotonylation is a permissive epigenetic modification. This project seeks to determine the extent to which inhibition of serotonin synthesis and reuptake, either alone or in combination, impacts the levels of histone serotonylation in the mouse brain. To inhibit brain serotonin synthesis, the lab uses a genetically modified mouse line that expresses a partial loss-of-function mutation in the tryptophan hydroxylase 2 gene, which is the gene that encodes the rate-limiting enzyme for brain serotonin synthesis. The results of this study could provide new insights into the molecular mechanisms through which genetic alterations in serotonin levels influence gene expression (and ultimately, behavior) as well as the mechanisms through which selective serotonin reuptake inhibitor antidepressants (like fluoxetine/Prozac) exert their effects.

Speech is a fast, adaptive process that allows humans to comprehend and produce language despite tremendous variability in how speech sounds are realized. Classical models proposed that listeners map acoustic input onto discrete, categorical representations. However, research increasingly shows that listeners remain sensitive to fine-grained acoustic detail, providing strong evidence that speech representations are gradient rather than strictly categorical. Gradience supports comprehension by increasing tolerance to variability and allowing listeners to update their perceptual categories in real time (perceptual retuning).

Crucially, evidence suggests that adaptation in perception may carry over to production, a process known as phonetic drift. Phonetic drift is thought to arise when updated perceptual targets influence articulatory planning, thereby introducing subtle shifts in a speaker鈥檚 own speech. Together, perceptual retuning and phonetic drift may form a feedback loop that not only facilitates individual adaptation but may be a core driver in the propagation of sound change within speech communities.

Despite its theoretical importance for laboratory phonology and psycholinguistics, the perception-production link has rarely been directly tested. The proposed project addresses this gap by combining behavioral and neurophysiological methods to examine how perceptual adaptation translates into production in a controlled laboratory setting, representing two common types of sound change: splits and mergers. Using electroencephalography (EEG), we will focus on two well-established event-related potentials (ERPs): the mismatch negativity (MMN), which indexes early, automatic detection of acoustic change, and the P3 component, which reflects category stability and decision-related processing. By relating individual differences in MMN and P3 responses to patterns of perceptual retuning and phonetic drift, this work will provide new insight into the neural and cognitive mechanisms that bridge speech perception and production鈥攁nd, ultimately, into how languages evolve over time.