Cornell University

NSF Awards · FY 2025 · 2025-09

Marine mass mortality events can drastically alter the structure and function of marine habitats, shifting the balance between organisms and leading to ecosystem degradation. The long-spined sea urchin Diadema antillarum plays a critical role in Caribbean coral reefs by maintaining the balance between corals and algae. Mass mortality of this urchin species has been linked to a pathogenic ciliate from the Diadema scuticociliatosis Philaster clade (DaScPc). Despite identifying the agent responsible, we still lack a fundamental understanding of the disease and its progression. By combining field monitoring, laboratory experiments, and molecular approaches, this project investigates how DaScPc, environmental conditions, and the microbiomes of Diadema interact to shape disease outcomes. These findings will be essential for predicting and potentially mitigating future urchin die-offs, thereby protecting the sensitive coral reef habitats they inhabit. This project includes the training and mentoring of graduate, undergraduate and high school students across four institutions, and the development of a global network of collaborators to facilitate monitoring and early detection of the disease. This project addresses two major questions surrounding mass mortality of the long-spined sea urchin Diadema antillarum. Aim 1 identifies environmental reservoirs of DaScPc and explores the environmental factors influencing its emergence and spread. This team conducts systematic time-series surveys of DaScPc presence in urchins, macroalgae, and corals in the U.S. Virgin Islands while monitoring physicochemical oceanographic conditions, host densities, and indicators of biological productivity. Additionally, field and experimental work assesses the potential reemergence of DaScPc from corals and other sympatric surfaces and its subsequent infection dynamics in Diadema. Aim 2 defines the growth conditions of DaScPc and determines how environmental factors and host microbiomes influence parasite-host interactions. This team assesses the effects of salinity, temperature, and nutrients on DaScPc growth and behavior in the laboratory. Co-culture experiments with bacterial strains isolated from the reef evaluates whether these microbes serve as prey or antagonists. Finally, this project characterizes the microbiomes of healthy urchins from unaffected sites and conducts experiments to determine how microbiome variation alters DaScPc-Diadema relationships. This integrative approach advances our understanding of the interactions among marine pathogens, hosts, and the environment, and provides critical tools for predicting and mitigating future outbreaks of marine diseases. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

With support from the Macromolecular, Supramolecular, and Nanochemistry Program in the Division of Chemistry, Professors Tristan Lambert and Brett Fors of Cornell University will develop a metal-free variant of a powerful polymerization reaction known as ring-opening metathesis polymerization (ROMP). Materials produced by ROMP are critically important for a range of applications; however, current ROMP methods rely exclusively on metal-based catalysts that are often toxic, costly, and limited in availability. As a result, ROMP materials typically contain residual metal impurities, requiring difficult and labor-intensive purification procedures—especially problematic for applications such as medical implants, drug delivery systems, and microelectronics, where metal contamination is unacceptable. The new ROMP methodology developed by Lambert and Fors will employ an organic catalyst composed solely of carbon, nitrogen, and hydrogen, making it inexpensive, non-toxic, and sustainable. The broader impact of this work includes the creation of a metal-free platform for preparing ROMP polymers, thereby enabling their use in sensitive and high-value applications, and training graduate students in the areas of polymer chemistry and catalyst development. More technically, this program aims to develop and apply 1,2-dialkylhydrazines as organocatalysts for the ROMP of a wide range of cyclic olefins. A central goal is to elucidate the structural features of these hydrazine catalysts that govern their reactivity and selectivity, and to use these insights to design more efficient and broadly applicable catalysts. The project will integrate empirical and computational approaches to achieve a detailed fundamental mechanistic understanding of key steps in the catalytic cycle, including cycloaddition, cycloreversion, and chain transfer. Catalysts with enhanced reactivity and control will be exploited to expand the scope of ROMP reactions to a more diverse array of monomers, thereby opening new avenues for the synthesis of functional polymeric materials without metal contamination. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

The world is becoming a hybrid society, where people interact with each other and with artificial intelligence (AI) agents. These human-AI relationships are hard to study. To enhance such investigations, the researchers build new online tools that let thousands of humans and AI “bots” meet and interact. These new tools are integrated into existing platforms for the behavioral sciences (Dallinger and PsyNet). This project helps scientists run large, realistic experiments to see exactly how humans and AI influence each other and learn how to make teamwork with AI more productive for everyone. The goal of this project is to make it easier for scientists across many fields to study how AI and humans can work together through three related aims. First, scientists build new tools to run social network experiments that combine people and AI agents. Second, there are three case studies showing different ways humans and AI can interact in group settings. Third, training and educational materials help researchers use the platform. This research helps understand hybrid societies and helps design better businesses and industries where people and AI collaborate effectively. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-09

PROJECT SUMMARY / ABSTRACT T lymphocytes are adaptive immune cells that protect the host against a diverse array of pathogens. In the last decade, immunologists have started to learn how to harness the function of T cells for immune benefit, which has revolutionized the field of immunology and changed the way we treat a variety of immune-mediated diseases, including cancer. However, recent discoveries in the fields of immunology and hematopoiesis suggest that our current model of T cell development is out-of-date, which may be contributing to inefficient treatment approaches to immune-mediated diseases. Up until now, the dogma in the field has been that all T cells are generated from bone marrow-derived hematopoietic stem cells (HSCs), and that the T cell compartment is comprised of a single lineage of T cells that are created equally. However, stem cell biologists have recently discovered that there are a wide variety of non-HSC progenitors at different anatomical sites in early life that can produce T cells. In separate studies, immunologists have learned that T cells made in early life persist into adulthood and exhibit unique roles during infection and disease. The stem cell biologists have yet to characterize the T cells made by non-HSC progenitors, and the immunologists have yet to map the precise hematopoietic origins of T cells found in adulthood. Essentially, the fields of hematopoiesis and immunology are not in sync, preventing us from updating the current model of T cell development. Our goal is thus to sync the fields, dissecting out the unique lineages of T cells made from different non-HSC and HSC progenitors in early life to determine how the function of T cells in adults is linked to their developmental origins. In Aim 1, we will establish what types of T cells can be made from various embryonic progenitors. In Aim 2, we will focus on a particular type of T cells (CD8+ T cells) and examine how their immune function and programming correspond to their derivation from non-HSC and HSC progenitors in early life. To accomplish these goals, we will employ embryo dissections, artificial thymic organoids, cell transplantations, and single cell sequencing, ultimately propelling us toward a new paradigm for T cell development. Once this paradigm is established, we expect to (i) uncover lineages of T cells ideally suited for immunotherapies, (ii) predict how patients will respond to current treatment strategies, and (iii) learn how to fully regenerate the T cell compartment after immune reconstitution therapies.

NSF Awards · FY 2025 · 2025-09

Generative AI tools like ChatGPT and Microsoft Copilot are rapidly changing how teams work together by helping people brainstorm ideas, write documents, and make decisions. But while these tools can boost productivity, they may also disrupt collaboration or strain team relationships. This project asks: What does good teamwork look like in the age of AI? And how can we design AI systems that make teamwork not just faster and more productive, but also more sustainable, supportive, and individually rewarding in the long term? Drawing from decades of research on what makes teams effective, the project will investigate how generative AI tools affect team dynamics, relationships, and individual well-being—not just task performance. The goal is to shape the next generation of AI tools so that they help teams thrive over time, both professionally and personally. This project applies a comprehensive framework of team effectiveness to investigate and design generative AI tools that support team interaction across three key dimensions: instrumental performance, team viability, and individual growth. Using a mix of qualitative interviews, fieldwork, and controlled laboratory studies, we will examine how current generative AI systems influence interaction patterns and effectiveness in text-based team collaboration platforms (e.g., Slack, MS Teams). We will then iteratively design new AI tools to improve team viability and individual well-being which are dimensions often overlooked in AI-supported collaboration research. These tools will be evaluated through rigorous empirical studies measuring outcomes across all three dimensions. The work contributes to research in human-computer interaction, computer supported cooperative work (CSCW), and computer mediated communication (CMC) by developing theory, design principles, and real-time intervention techniques for AI-mediated teamwork. It also builds technical foundations for integrating generative AI into everyday collaborative systems in ways that foster effective and sustainable work in teams. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

Ice shelves fringe most of Antarctica’s coastline where the ice sheet detaches from the underlying bed and enters the ocean. These expansive floating bodies of ice provide an important bottleneck on seaward flow of ice from the Antarctic ice sheet and sea level rise through their resistive buttressing forces. The ocean melts ice shelves from below, which removes their mass and influences their buttressing capacity. Local regions of thin ice, referred to as channels and crevasses, form in the base of ice shelves from ocean-driven melting and strain-driven cracking. These features represent potential weak points in the ice that are particularly susceptible to increased melting if ocean temperatures rise, because of their high slope angles. Improving understanding of the processes that determine the rate of melting along ice bases of varying slopes has global societal and environmental implications, as it stands to reduce uncertainties in projected sea level rise. This project will study variable melting along sloping ice using a novel field data set collected in a channel etched into the base of Fimbul Ice Shelf, Antarctica in 2024. Results from this project will provide a baseline for how sloped bases of ice shelves melt, which can be used to improve parameterizations of this process in large-scale models responsible for sea level rise projections. This study focuses on ocean-driven melting along sloping sidewalls of local thin points in ice shelves, such as basal channels or basal crevasses. These features are particularly sensitive to oceanographic forcing, as they exhibit high slopes that can melt rapidly when exposed to warm ocean conditions. If this melting is strong enough, then it can erode features to the point that they become unstable, resulting in full-thickness fractures that promote iceberg calving and ice shelf destabilization. Notably, the upstream effect of ice shelf changes related to oceanographic forcing is a major source of uncertainty in projections of Antarctica’s contribution to future global sea level rise, which could amount to 53 cm by 2100. The primary objective of this project is to make significant improvements to the understanding of how the ocean melts sloped ice shelf basal topography under various forcing. The secondary objective is to understand how ice topography then evolves over time from this ocean forcing. The principal study location is Fimbul Ice Shelf, Antarctica, where in situ data was collected with the Icefin underwater vehicle and other instrumentation in January 2024, as part of an international collaboration with the United Kingdom and Norway. Results from this study will be placed into the larger context of ice shelf melting around Antarctica by comparing with previous data collected with Icefin in various oceanographic settings. The goal of this effort is to better constrain the poorly understood coupled ice-ocean processes that control melting along variable slopes. The Icefin data will be analyzed alongside data from surface-based ice penetrating radar, remote sensing, long-term oceanographic mooring, and output from the Finite-Volume Community Ocean Model. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

Computer networks are critical for distributed software systems, and their failure can lead to major disruptions. Tools for verifying networked behavior can help catch bugs and improve trust, but there is a fundamental challenge that limits the applicability of formal methods in networking: the need for precise models of the devices, topologies, and systems being verified. This project will develop new techniques that enable automatic construction of such models, which will enable broader use of formal verification in networking, ensuring that networks satisfy specified correct properties and are aligned with the intents of human operators. The project will also pursue education and outreach efforts to share results with the broader community and will seek to transition the fundamental research into practice through collaborations with industrial partners. Despite many recent successes of verification in networking, many important properties remain challenging to formally reason about. One challenge is the lack of models that automatically evolve as changes are made to network devices, configuration, and topology. Another challenge is the lack of models that capture essential quantitative properties such as latency and reliability. This project will further the vision of automated inference of models and provide efficient techniques for quantitative model learning that do not require access to source code or handcrafted models of network behavior. At the core of the development will be a unique integration of new closed-box learning algorithms and symbolic techniques crucial to tame the scale of network models. The outcomes will be a new generation of formal verification techniques based on closed-box learning and grounded in applications in networking. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-08

PROJECT SUMMARY/ABSTRACT Epidemiological studies highlight the ability of early-in-life microbial exposure to induce environmentally adaptive, long-term shifts in the immune response. The immune system’s adaptive potential may be explained in part by central immune training, the process by which hematopoietic stem and progenitor cells (HSPCs) undergo lasting epigenetic reprogramming after receiving inflammatory signals. Importantly, the epigenetic inheritance of immune training from HSPCs into mature immune cells results in their enhanced responsiveness to pathogens they have never been exposed to previously; however, the mechanisms underlying the maintenance and transfer of immune training remain to be comprehensively investigated. Here, I will focus on the training of CD8+ T cells, thus describing immune training in an adaptive immune cell type for the first time. By leveraging a mouse model system of immune training, I will complete an interdisciplinary project determining the gene-regulatory impact of microbial exposure on central immune training. Specific Aim 1 entails the joint profiling of gene expression and chromatin accessibility at single-cell resolution to uncover the genes, enhancers, and transcription factors that are responsive to microbial exposure in HSPCs. The results of Aim 1 will provide insight into how, and where, microbial exposure initiates epigenetic reprogramming. Specific Aim 2 employs CRISPR-mediated transcription factor knockouts, ATAC-seq, and functional CD8+ T cell assays within an Artificial Thymic Organoid (ATO) system to examine the requirement of the transcription factors FOSB and KLF6, which preliminary data implicate as mediating central immune training. The results of Aim 2 will determine the requirement of specific TFs in maintaining and transferring immune training from HSPCs to CD8+ T cells. Specific Aim 3 utilizes enzymatic methylation sequencing and chromatin conformation capture approaches to investigate the epigenetic mechanism by which immune training is transferred from HSPCs to CD8+ T cells. The results of Aim 3 will determine the role of DNA methylation in transferring immune training from HSPCs to CD8+ T cells and the role of chromatin architecture in facilitating differential gene expression in trained CD8+ T cells. I will utilize the expertise obtained from my previous research experiences, along with expert-led training, to perform the proposed experiments. Further, by enhancing my skills as a communicator, mentor, and teacher of science, I will become a well-rounded researcher, allowing me to seamlessly transition into a postdoctoral position after completing my thesis work. The Grimson and Rudd labs, my mentoring committee, and collaborators within the Genomics Innovation Hub at Cornell University, will provide invaluable mentorship, support, and training, ensuring my mastery of all research techniques required for the completion of this proposal. Ultimately, this proposal will advance the understanding of the gene-regulatory underpinnings of central immune training and will facilitate significant development in my path to becoming an independent researcher.

NSF Awards · FY 2025 · 2025-08

Robots can make our lives better by helping at home, in hospitals, and on farms. But most robots today can only do tasks that are pre-programmed ahead of time. They cannot handle new situations or learn from people. This project supports research to create robot helpers that learn new skills like humans do. These robots will learn by watching people, trying tasks, and improving from feedback. This work should make robots more helpful and flexible, so they can solve harder problems in the real world. It also intends to help us understand how robots can learn and adapt. The project looks to improve robots for homes, healthcare, and agriculture. It will also get stimulate students in science through hands-on robotic activities. This project seeks to develop a new framework for teaching robots to learn tasks in real time. It uses three key approaches. (1) Learning from Demonstration helps robots gain skills by watching human actions. It addresses differences between what humans show and what robots can do. This allows robots to apply knowledge across different tasks and places. (2) Learning through Interaction helps robots take in feedback from people and their surroundings. It handles situations where information may be unclear or incomplete. (3) Learning through Collaboration helps robots understand human intentions and goals. It makes effective teamwork on shared tasks possible. Researchers will test these approaches in real-world settings such as homes and greenhouses. Applications range from collaborative cooking to organizing clutter and managing crops. The project includes educational programs with interactive robotics activities for K-12 students. It provides accessible online resources to increase participation in STEM and robotics research. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

The astronomical community devotes significant resources to ground-based and space-based astronomical surveys—systematic explorations of vast regions of space and time. The resulting catalogs provide “family portraits” of diverse cosmic populations, including asteroids and trans-Neptunian objects in the solar system, nearby stars and their exoplanets, distant galaxies and active galactic nuclei, and transient objects such as gamma-ray bursts, fast radio bursts, or supernovae. A research group at Cornell University will construct a “Cosmic Demographics Toolkit” (CDT), a widely applicable suite of conventional and modern computational tools accessible and appealing to the broad community of astronomers analyzing survey data. The project includes astrostatistics research producing methods with new capabilities, and development of software tools in an open-source, well-documented Python package. The resulting toolkit will help astronomers extract the best results from astronomical surveys. The researchers and their students will contribute to the training of the next generation of astronomers and computational scientists by participating in the Summer Schools on Statistics for Astronomy and Physics hosted by Penn State’s Center for Astrostatistics (CASt) and by producing videos communicating key aspects of advanced data analysis to the broader public. The CDT will include several demonstration applications, addressing important prototypical demographics problems such as estimating luminosity functions of galaxies, fast radio bursts, and gamma-ray bursts, and size distributions of trans-Neptunian objects. It will also address discovering correlations and scaling laws among the properties of astronomical objects. The software will be written in Python as new components of an open source package called Inference, which supports a variety of statistical inference tasks arising in astronomy. It will include hierarchical Bayesian modeling, nonparametric techniques, censored data methods, and Bayesian survival analysis methods. The project will take advantage of the Rubin Science Platform to accelerate development and tailor many capabilities to analyses using data from the Legacy Survey of Space and Time being undertaken by the Vera C. Rubin Observatory. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

This project will investigate natural irregularities in the earth’s ionosphere, the layer of the upper atmosphere that contains charged constituents, at altitudes of about 100 km, near the edge of space. The irregularities affect radio wave propagation and constitute a form of space weather that affects several ground- and space-based operational systems such as radar, global navigation systems, and satellite imagery, especially during summer evenings when irregularities occur most often. The irregularities occur in so-called “sporadic E” layers that are among the first phenomena that were detected in space about 100 years ago. The specific causes of the irregularities remain unknown, however. This project seeks to find and understand causes rooted in the lower and middle atmosphere associated with regular weather phenomena. It also seeks to understand if space-weather phenomena higher in the ionosphere are causally related to the irregularities that exist near 100 km altitude. The study will use radars located at Cornell and Clemson University along with other regional radars, GNSS receivers, and other instruments. Observations will be interpreted using a combination of theory and computer modeling and simulation. The experimental and computational tools to be developed for this problem will be incisive and groundbreaking. The proposal will address the influence of two-way atmosphere-ionosphere coupling on mesosphere/lower thermosphere neutral dynamics, plasma processes responsible for meter-scale irregularities in patchy sporadic E layers and their effect on F-region. The project will moreover be an important vehicle for early-career professional development. Professional and graduate students will take part in the deployment and operation of radar systems at two universities, affording them radiofrequency, hardware-based training opportunities rare in the present day. Students will participate in all aspects of this work including data analysis, modeling, and simulation, and the hardware and software developed for this project will contribute to course curricula. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

Human activities and global changes alter the movement of elements from the atmosphere to land and ultimately into streams, particularly in northern latitude regions. However, there is limited understanding of how altered environmental conditions impact trace element movement. Even though some trace elements form essential nutrients, others are toxic to humans and wildlife. Therefore, determining long-term patterns in these elements is important to understand their human and ecosystem health impacts. This project will examine long-term patterns of trace elements in streamwater, quantify trends in their concentration, identify the associated drivers, and predict how the concentrations will vary in the future. The results will inform management and mitigation of trace elements in boreal streams. The project will develop curriculum modules for high school students and incorporate them into a high school summer program, focusing on hands-on experience and connecting basic science to real-world problem-solving. Project personnel will participate in outreach activities involving stakeholders and students, train undergraduate and graduate students, and disseminate key outcomes through articles in the popular press. The overarching goal of the project is to understand how shifting environmental processes impact the fate and transport of contaminants now and in the future. This will be achieved by examining the primary drivers and patterns of sixty trace and ultratrace elements within boreal streamwater across temporal (i.e., decadal) and spatial (i.e., nested catchment) scales. This project will quantify long-term streamwater patterns for a suite of trace elements over several decades (1985-2023). The results will identify primary geochemical and hydrological drivers of spatiotemporal variability in streamwater trace element concentrations and flux. The outcomes will make it possible to predict changes in streamwater trace element concentrations and flux under varying landcover and altered precipitation regime scenarios. This research will combine fieldwork, laboratory work, and computational modeling to improve understanding of hydrological and geochemical drivers of streamwater trace and ultratrace element concentrations. This project is funded by the National Science Foundation's Office of International Science and Engineering. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

This project lays the foundation for AI4Ag, an AI-ready living lab for agriculture, where scientists, students, and technology developers will collaborate to create artificial intelligence (AI) systems that support food production and improve farming efficiency. As the global agri-food system increasingly relies on data-driven technologies, stakeholders, from farmers and industry to consumers (and even crops and livestock) stand to benefit. AI driven innovations have the potential to address major challenges including labor shortages, food safety, disease and pest management, and weather variability. However, many farms struggle to adopt advanced technologies due to high costs, limited infrastructure, and the complexity of agricultural systems. Hosted at Cornell University, AI4Ag aims to overcome these barriers by creating a shared, accessible space equipped with tools, data, and expertise to support AI development and testing in real-world settings. The project will also foster a community of researchers and students, helping to train the next generation of innovators in agriculture. By making it easier to test, refine, and share new ideas, AI4Ag will unlock the potential of AI to transform agriculture and contribute to a more sustainable and resilient US food system. This project establishes the foundation for AI4Ag, an AI-ready living lab embedded within Cornell Agricultural Systems Testbed (CAST), designed to accelerate the development and deployment of AI technologies in agriculture. CAST, operated under the Cornell Institute for Digital Agriculture (CIDA), provides a robust platform with sensing infrastructure, live and archived data streams, and interdisciplinary expertise across animal science, crop science, engineering, and computing. AI4Ag will lower barriers to entry for researchers, industry, and practitioners by enabling real-world testing of AI solutions in complex agricultural environments. The project will develop a strategic plan and governance structure, including an executive management team, testbed management team, research steering committee, and advisory board, to guide AI innovation. To drive this effort, the first objective is to build a multidisciplinary user community. Second, is to assess and enhance the CAST infrastructure. Third, is to demonstrate AI-readiness through data integration. Fourth, is to establish operational frameworks for external access. Expected AI innovations include a suite of state-of-the-art AI tools and methodologies, including large language models and agentic AI frameworks addressing agriculture-specific tasks. AI4Ag will serve as a replicable model for AI-enabled living labs, fostering participation and training the next generation of AI and agri-food researchers. By aligning infrastructure with user needs and demonstrating feasibility, this project will catalyze sustainable, scalable AI innovation across the agri-food system. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

For the last 30 years, string theorists and mathematicians have worked together to “count” curves in certain spaces. These “curve-counting numbers’’ are of intrinsic interest in both physics and mathematics. As the two groups simultaneously study these numbers, insights from the string theorists often lead to developments in mathematics, as well as vice versa. The PI will further develop a variety of mathematical techniques for studying these curve-counting numbers. Some of these developments are directly motivated by open questions in string theory, while others are more closely tied to recent advances in algebraic geometry, especially moduli theory. The PI will use these projects to train graduate students and postdocs in conducting mathematical research and in communicating their results. In particular, the PI will make lecture notes on important foundational topics available to both math and physics graduate students. The PI will also participate in activities to improve communication of research within the mathematics community. In a little more detail, the PI has three research objectives. First, the PI will investigate new modular compactifications of the moduli space of stacky curves and their applications to enumerative geometry. These compactifications have direct implications for the study of non-stacky curves (via Hurwitz theory) as well as potential applications to higher-genus, log, and orbifold Gromov-Witten theory. Second, The PI will study the geometry of Geometric Invariant Theory (GIT) quotients of representations of complex reductive groups. GIT quotients of representations are a well-used testing ground for new algebro-geometric theories, and the PI’s research will make it easier to find and use examples. Lastly, the PI will investigate deformed virtual classes in sheaf cohomology, aiming toward a (0, 2) analog of Gromov-Witten theory. The definition of such has been a mystery in string theory for the last decade, but recent advances in derived algebraic geometry make such a definition mathematically feasible. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

Understanding and mitigating the rapid spread of infectious diseases requires innovative mathematical tools that capture the complexity of real-world human interactions. This research project develops novel stochastic mathematical models that combine queueing theory and advanced probabilistic methods to shed a new light on how diseases propagate within service systems and broader social networks. By quantifying personalized infection risk in crowded environments such as hospitals and transportation systems, these models reveal how quickly a single infected individual can trigger widespread outbreaks in service systems. Furthermore, analyzing infectious disease spread in queueing systems offers valuable insights into side-channel attacks in cybersecurity. In particular, cryptographic operations can be modeled as tasks in a queue, with processing times influenced by factors such as data or device characteristics. Attackers exploit these time variations to extract sensitive information, such as cryptographic keys. By treating security systems as queues, the mathematical models in this proposal can help reveal potential information leakage, thereby contributing to the design of more robust cybersecurity measures. The resulting insights from this work will empower public health officials to make data-driven and model-driven decisions, which will ultimately reduce spread and optimize resource allocation during subsequent health crises, and also provide a framework for understanding and mitigating vulnerabilities in cybersecurity systems. Undergraduate and graduate students will participate in these research activities, contributing to STEM workforce training. This project develops new stochastic models to capture how infectious disease spread in service systems, with a parallel application to understanding cybersecurity vulnerabilities. The primary goal of this work is to provide deeper insights for service system managers and public health officials to assess the infection risk associated with waiting in lines at airports, hospitals, and transportation hubs during public health crises. Simultaneously, these models offer valuable insights into side-channel attacks in cybersecurity, where cryptographic operations can be conceptualized as tasks in a queue, and processing time variations can be exploited to extract sensitive information like cryptographic keys. By leveraging tools from queueing theory and stochastic processes, the research will quantify the individual infection risk of a susceptible individual and the community-based risk that an infected individual has on society, as well as reveal potential information leakage in security systems. The theoretical results will be validated through stochastic simulation and compared with real-world data from various service system environments and relevant cybersecurity contexts. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

Microplastics are tiny plastic particles that accumulate in the environment. This project will study how microplastics move through soil and water and interact with potentially toxic chemicals, such as chlorinated ethenes and per- and polyfluoroalkyl substances (PFAS). The research will explore the fate of microplastics in the environment by conducting a series of laboratory experiments of increasing scale and complexity under conditions that mimic those of contaminated sites. Results of the project will advance scientific knowledge that can be used to protect public health, remediate contaminated sites, and protect water supplies. Although the interactions between microplastics and halogenated organic compounds (HOCs) are recognized, fate and transport and risk models do not capture the critical roles and dynamics of these interactions. This CAREER project aims to address these critical gaps by exploring the fate and transport of microplastics and their interactions with the associated adsorbed HOCs in soil and groundwater environments. The specific objectives are to: (1) evaluate the impacts of aged and weathered microplastics on HOC adsorption and kinetics; (2) assess the impacts of weathering and aging of HOC-adsorber microplastics on HOCs biotransformation, and resolve how microbial community composition differs at the microplastic and soil interface; (3) evaluate the effects of flow (e.g., simulated rainfall events) on the fate and transport of microplastics and adsorbed HOCs in the unsaturated zone and at the water table; and (4) create programs to disseminate knowledge about the impact of HOCs and microplastics on water quality. The project will generate critical data for improving mathematical transport models and conceptual site models, leading to more representative risk and exposure predictions. Additionally, the project will provide valuable insights into the potential for soil and subsurface microbial communities to affect the fate of microplastics and HOCs. Overall, this research will contribute to the protection of soil health and drinking water supplies, ultimately promoting environmental sustainability and public health. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-08

PROJECT SUMMARY Formaldehyde is a well-recognized human genotoxin found commonly in the environment but also produced endogenously in mammalian cells. While environmental exposures to formaldehyde are strictly minimized, recent work has shown that mammalian cells are exposed to quantities of endogenous formaldehyde sufficient to cause cancer and pre-mature aging in children with genetic deficiency in formaldehyde-detoxification enzymes or DNA repair pathways. However, beyond these genetic syndromes, we know very little about the life factors that determine the level of endogenous formaldehyde and its impact on health within the general population. Although several biochemical reactions in the cell can generate formaldehyde, such as by demethylase enzymes, and through spontaneous decomposition of folate, the contribution by these reactions to formaldehyde burden and toxicity in vivo is not characterized. Therefore, the overall objectives of this R35 proposal are to better understand the fundamental biology of endogenous formaldehyde production and toxicity in mammalian cells by 1) identifying the physio-pathological sources of formaldehyde in vivo and define how these pathways can be dysregulated to elevate formaldehyde burden, and 2) dissecting the different mechanisms by which endogenous formaldehyde disrupts cell function. To address these objectives, we have developed an innovative and ultra- sensitive workflow that combines a formaldehyde biosensor with mass spectrometry to detect formaldehyde production in mammalians cell lines and mouse models under basal homeostasis and stressed metabolic states. We will leverage this workflow with metabolic and genetic approaches to identify the metabolic conditions and factors that impact formaldehyde production. In parallel, we have utilized a genome-wide CRISPR screen to highlight novel formaldehyde-sensitizing genes that could reveal novel mechanisms of mammalian formaldehyde-toxicity. Successful outcomes from our research will help to build a framework to understand the key cellular factors that determine endogenous formaldehyde level and the subsequent cellular toxicities that arise. The knowledge gained and novel research tools generated will form the foundation for future studies into how endogenous formaldehyde contributes to human diseases and designing interventions that target formaldehyde to improve health.

NSF Awards · FY 2025 · 2025-08

This project aims to support undergraduate engineering programs to easily build effective partnerships with local wastewater treatment plants (WWTPs), an important but often overlooked part of communities. These partnerships can give students significant, real-world, hands-on learning experiences about water treatment that make their education more meaningful and practical. Smaller colleges, especially in rural areas, often have a hard time connecting with industry partners. By studying successful partnerships and sharing what works and what challenges were overcome, the project intends to help more schools and industries work together. A planned result of this project is a workshop that gives colleges tools and strategies to start or improve partnerships that support better learning outcomes for students and stronger ties between schools and local industries. This project is also designed to enhance the knowledge, skills, and practices in STEM education research for the postdoctoral researcher. With over 16,000 wastewater treatment plants (WWTP) nationwide, the wastewater industry is an underutilized and accessible learning resource for colleges and universities. Collaborating with industry partners in engineering education can enhance students' professional skills, benefit industry partners, and provide authentic learning environments. Despite these benefits, many faculty and departments lack the knowledge and tools to form and leverage these college-industry partnerships. Through a multi-case study of established WWTP partnerships, this research seeks to answer these research questions: 1) What successes, challenges, and necessary contributions from engineering programs and WWTPs help form and sustain mutually beneficial partnerships? 2) What are the unifying features and differences in these partnerships? 3) How do these partnerships impact student motivation and their beliefs about the authenticity of their engineering education? This study employs semi-structured interviews with faculty, wastewater treatment professionals, and students, and analysis of partnership agreements and course materials developed. The goal of this research, grounded in Activity Theory and Stakeholder Theory, is to obtain a deeper understanding of how universities and WWTPs can form mutually beneficial and sustainable partnerships and how these experiential learning environments impact student learning beliefs and motivations. This project is funded by the STEM Education Postdoctoral Research Fellowship (STEM Ed PRF) program that aims to enhance the research knowledge, skills, and practices of recent doctorates in STEM, STEM education, education, and related disciplines to advance their preparation to engage in fundamental and applied research that advances knowledge within the field. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-08

Project Summary/Abstract Cellular aging is a complex physiological process defined by progressive loss of function over time, and constitutes a major risk factor for age-associated diseases. The budding yeast, Saccharomyces cerevisiae, is a foundational model for our understanding of cellular aging. We have established a new genetic system, the “daughter extinction program” (DEP), scalable for system-wide profiling of replicative aging. Enabled by the DEP, we have generated a first genome- wide quantitative aging map of a eukaryotic cell. Autophagy is a universal modifier of aging at the center of cellular aging and a promising target for therapeutic intervention in aging and age-associated disease. However, autophagy involves complex combinations of non-selective (bulk) and selective forms of autophagy for the regulated degradation of an unparalleled scope of substrates. The goal of this project is to systematically dissect the individual and combinatorial contributions of the different genes, forms and functions of autophagy to cellular aging. We have identified not only differential functions of autophagy for cellular aging, but also two new paradigms of extending replicative lifespan (RLS) by modifying specific aspects of autophagy. Our goals are to gain mechanistic understanding of how tuning autophagy towards selective forms or inhibiting a specific form of selective autophagy of mitochondria (mitophagy), which we show regulates mitochondrial size, significantly prolongs RLS. To achieve our research goals, we propose the following three aims: (1) Deconvolve the differential functions of autophagy for replicative aging. We will systematically dissect how autophagy governs cellular aging at single and combinatorial gene resolution, study how limiting or enhancing bulk and/or selective forms of autophagy determine RLS at basal level and in different genetic and chemical models of elevated autophagy, and characterize key regulators of autophagy and all selective autophagy receptors (SARs) for their potential as bone fide aging modifiers. (2) Determine the genetic landscape of Atg32-dependent replicative aging. The conserved mitophagy-receptor Atg32/BCL2L13 as a novel bone fide aging modifier. We have generated a genome-wide aging map of an Atg32-deficient cell to define the genetic requirements for Atg32-dependent RLS regulation. We will characterize the functional interactions of Atg32 with the autophagy machinery and focus on understanding how two different mitophagy receptors cause opposing effects on cellular aging. (3) Define the molecular mechanisms of Atg32/BCL2L13-mediated mitophagic scaling in replicative aging. To interrogate the function of Atg32 as an aging modifier, we aim at understanding how molecular mechanisms of mitophagic scaling interface with cellular aging utilizing multidisciplinary approaches. In sum, these studies will lead to a comprehensive model for the regulation of cellular aging by autophagy, foundational to the promise of rationale tuning of autophagy for improved human aging and age-associated disease.

NSF Awards · FY 2025 · 2025-08

This project introduces and examines softmax mixture ensemble models to address contemporary questions related to the evaluation and interpretation of data generated by trained large language models (LLMs). These statistical models will help summarize varied document corpora by identifying their semantically meaningful latent topics. Current LLMs contain billions of parameters, which makes them difficult to interpret and use directly in subsequent analyses. There is an urgent need for corpus summaries that strike a balance between the complexity of the generating models and a user-friendly representation. This project aims to develop new metrics based on the interpretable corpus summaries to provide critical insights into the similarities and differences between human-generated and AI-generated text. This research develops computationally efficient inferential methods, with sharp mathematical guarantees, for learning and analyzing softmax mixture parameters from data consisting of a collection of samples, each modeled as mixtures with common mixture components and sample-specific coefficients. The softmax mixture ensemble model will be shown to be a crucial building block in a more complex mixture-of-experts model. The project will provide experimental evidence for the benefits of this framework in analyzing LLM data. Solving the statistical questions of this project requires bringing to bear tools from optimization theory, probability and high-dimensional statistics, while addressing the application questions will require tools from computer science, specifically from the areas of natural language processing (NLP) and, more generally, AI. The ultimate goal is to develop procedures which yield accurate, robust, and interpretable results, readily applicable to scientific applications. The central goals of this research are parameter identifiability in softmax ensemble models, polynomial-time algorithms for parameter estimation in high-dimensional softmax ensemble models with dense and sparse parametrization, finite sample (minimax) guarantees, asymptotic inference for parameter estimates in softmax ensemble models, as well as the development of necessary tools for evaluating LLM output in open-ended text generation. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-08

Breast cancer is the most common cancer in women and frequently metastasizes to the skeleton, leading to poor prognosis. As existing therapies are relatively ineffective, a promising strategy could be to eradicate dormant tumor cells before they become activated. Indeed, tumor cells disseminate to the skeleton early during disease where they can remain quiescent for years or decades. Better understanding the mechanisms that regulate the fate of disseminated tumor cells prior to lesion formation promises to inform therapeutic options to prevent metastasis. Upon arrival in bone, tumor cells colonize osteogenic niches where they interact with mineralizing bone matrix and are protected from immune attack by Natural Killer (NK) and T cells. Although decreased bone mineral density is a known risk factor for bone metastasis, it remains elusive how varied bone mineral density affects early stages of the metastatic cascade including immune evasion. Defining these links is critical as bone matrix is a dynamic biomaterial whose physicochemical properties regulate the behavior of many cell types and, thus, is likely to also influence disseminated tumor cells and their interactions with the immunological microenvironment. Our preliminary data suggest that physiological bone mineral content inhibits tumor cell growth while increasing latency due to metabolic and mechanosignaling reprogramming. In addition, we recently identified that bone mineral content-dependent metabolic programs direct tumor-cell synthesis of immunomodulatory “glyco-codes” that favor immune evasion. Based on these promising preliminary data, we have assembled a multidisciplinary team of experts in tumor engineering and bone metastasis, mechano- and glycobiology, and transcriptomic approaches to study the overall hypothesis that the mineral content of bone matrix influences early-stage metastasis by enriching for less proliferative, stem-like tumor cells that are immuno-privileged and can drive metastatic outgrowth when bone mineral content is perturbed. To address this hypothesis, we will combine innovative in vitro, in vivo, and bioinformatic approaches with advanced imaging to test how variations in bone mineral content regulate tumor cell mechanosignaling and metabolism and how these changes impact the proliferative and dormant phenotype of tumor cells in the skeleton. In parallel, we will elucidate how changes in bone mineral content instruct tumor-cell synthesis of an immunomodulatory glycocalyx using our unique tools for glycocalyx materials research which include genetic methods to engineer the structure of the glycocalyx, computational models to understand the physical behaviors of the glycocalyx at biointerfaces, and optical tools to characterize the glycocalyx structure. This powerful combination of innovative conceptual advances and technologies will enable us to define how changes in bone mineral content regulate skeletal metastasis and link these changes to targetable cellular mechanisms.

NIH Research Projects · FY 2025 · 2025-08

PROJECT SUMMARY / ABSTRACT Understanding the fundamental processes that determine an optimal immune memory response has been the bedrock of vaccine therapy and cell-based immunotherapy. Historically, therapeutic strategies have focused on strengthening the adaptive immune arm, the primary agents that effectuate immunological memory. More recently, research has expanded to include an innate immune arm that can retain a durable enhanced capability after initial activation, a phenomenon more commonly known as trained immunity. Natural killer (NK) cells have been pioneers in traversing the borders of innate and adaptive immunity, as best demonstrated in studies on mouse and human cytomegalovirus infection, whereby NK cells can acquire long-lived memory in an antigen-specific manner. At the same time, NK cells can acquire features of trained immunity, whereby activation by proinflammatory cytokines can nonspecifically enhance longevity and functional potential that is maintained weeks after primary exposure. Together, this malleability makes NK cells a promising and versatile tool for cell- based immunotherapy. One of the fundamental mechanisms by which these NK cells and all immune memory cells retain potential is via epigenetic reprogramming. This R01 proposal thus focuses on understanding and dissecting out epigenetic features that establish an immune memory program, and how these programs fundamentally differ between antigen-specific and trained immune memory. We study two related but non- mutually exclusive readouts of epigenetic regulation: chromatin accessibility and DNA methylation. Based on our previous work comparing the in vivo open chromatin landscapes of antigen-specific NK cells and CD8+ T cells, we have identified AP-1 factor JunB as playing a role at least during early antigen-specific responses, and have additional evidence that suggests it may also play a role in trained immunity. Based on performing similar analyses that instead focus on the DNA methylome, we further explore the role of genome-wide active DNA hypomethylation in establishing antigen-specific and trained NK cell memory responses. As a result, Aim 1 focuses on perturbing the open chromatin landscape via JunB-depletion, while Aim 2 disrupts the DNA methylome via Tet2-depletion. By doing so, we will identify key candidate regulatory elements and genes that link epigenetic activity to cellular and functional phenotypes, thereby providing a launchpad for therapeutic strategies that co-opt NK cell memory traits to improve NK-cell based therapies.

NIH Research Projects · FY 2025 · 2025-08

PROJECT SUMMARY/ABSTRACT Interpretation of clinical pathogenesis of somatic mutations is crucial for the advancement of precision cancer medicine, especially given the thousands of cancer genome/exome sequencing data available today and many more to come in the near future. Although traditional linear genomic sequence focused, often “one at a time” approaches have led to seminal discoveries, they now must be complemented by 3D protein structure based methods that take into account rewiring of sub-cellular systems and the molecular “interactome” network during tumor initiation, progression and maintenance. However, 237,787 interactions (94.2% of the current human interactome) do not have any structural information, most of which are not amenable to current structural modeling methods (including AlphaFold-based methods), Here, we propose to develop a big-data-driven deep- learning-based pipeline, named PIONEER (Protein-protein InteractiON IntErfacE pRediction; Aim 1) to generate a comprehensive 3D human interactome with significant improvement in quality and coverage. Importantly, we will integrate PIONEER with the powerful structure-alignment-based PrePPI pipeline to generate atomic- resolution 3D models for the entire human interactome for the first time, addressing a key unmet need in precision oncology and critical for NCI missions. Take advantage of nearly full structural coverage for all individual proteins (by AlphaFold) and protein- protein interactions (PPIs; by PIONEER), we will develop an end-to-end 3D-structurally-informed anisotropic network propagation framework to identify 3D spatial clusters of cancer mutations, especially at PPI interfaces (named “oncoPPIs”), and likely dysregulated network modules/pathways (Aim 2). In Aim 3, we will validate our interface predictions and 3D models using existing large-scale mutagenesis and interactome perturbation data, and through cross-linking mass spectrometry experiments. We will further functionally validate our results using existing cancer proteogenomic and clinical datasets. We have extensive preliminary results confirming that our 3D-structrually-based identification of oncoPPIs and dysregulated modules/pathways significantly correlated with patient survival and treatment across diverse cancer types, providing excellent candidates for developing personalized medicine and treatment strategies, as well as better understanding of molecular mechanisms underlying specific cancer types. Furthermore, we will perform functional validations in cancer cell models and patient-derived tumor organoids through collaborations. We will deploy an interactive web portal to disseminate our results and all of our tools for on-demand 3D model building, network analysis, and cancer multi-omics studies. Our comprehensive multiscale 3D human interactome and the accompanying knowledge portal will greatly reduce the barrier-to-entry for performing systematic structural analysis on a large number of proteins and their interactions, and open the flood gates for such analyses in cancer genetic and genomic studies.

NSF Awards · FY 2025 · 2025-08

Campi Flegrei is one of the most dangerous volcanoes in the world. It is located near Naples, Italy, close to the famous sites of Pompeii and Mount Vesuvius. More than 1.3 million people live in the area, and millions of tourists visit each year. The ground at Campi Flegrei has risen by more than 4 meters since the 1950s, and the area has experienced many earthquakes. These warning signs have become stronger in recent years, raising concern about a possible eruption. This project will build a computer model to help scientists understand whether the volcano is getting closer to erupting. The study will focus on modern times, when satellite and ground data are available, and the centuries before the last eruption in 1538 using historical and archaeological records. The model could also help forecast activity at similar volcanoes in the United States, like Yellowstone and Long Valley in California. This research will provide training for an early-career scientist and a student intern and will strengthen collaboration between scientists in the United States and Italy. This grant supports scientific progress and helps protect lives and communities from natural disasters. This study aims to improve our understanding of the volcanic and seismic hazard associated with the ongoing unrest at Campi Flegrei, one of the most densely populated volcanoes of the world. The goal of the project is to conduct a series of numerical data assimilation experiments using the Ensemble Kalman Filter (EnKF) with high-fidelity, multiphysics 3D finite element method (FEM) models to evaluate Campi Flegrei’s deformation data from 1946 to present and its historical unrest from 1251 to 1538. The EnKF-FEM approach will allow calculation of variations in the stress field from the magmatic system while also considering the effects of topography, rheology, and pre-existing weakness due to caldera faults. The proposed investigation will be the first study of its kind at Campi Flegrei and the first investigation to use the EnKF-FEM technique to calculate stress evolution of a magma system over such a long-time series. An aim is to demonstrate that the EnKF can significantly improve our ability to track the stress field over time at long-term deforming volcanoes. Additionally, the project will address critical questions regarding the Campi Flegrei magma system: Q1. How have repeated unrest episodes at Campi Flegrei impacted the evolution of stress? Q2. What is the estimated state of stress presently? Q3. How might the ongoing unrest impact the state of stress in the future? Q4. What was the state of stress prior to the 1538 eruption? The expected results will provide critical insights into ongoing unrest at Campi Flegrei. This work will provide a tool to evaluate unrest episodes in large felsic calderas worldwide (e.g., Toba, Long Valley, Yellowstone). This project will support the career development of a postdoctoral researcher and fund a summer intern who will participate with a cohort of students to explore opportunities for graduate study and careers in the geosciences. Finally, an educational module on Campi Flegrei will be developed for 100-level Natural Disasters courses and made available via GETSI (GEodesy Tools for Societal Issues) hosted by SERC (Science Education Resource Center). This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-08

Graphs, representing complex sensing and other societal systems like disease networks, social networks, and communication networks, are essential in understanding interactions within these systems. By accurately modeling relationships and structures within data via graphs, today machine learning over graphs (LoGs) plays a vital role in various applications. However, LoG introduces additional hyperparameters such as graph topologies and nodal embeddings into the already complicated neural network training processes. Traditionally, LoG approaches relied on user-defined heuristics to extract features encoding structural information about a graph. However, this process becomes prohibitively expensive in large models and high-dimensional data regimes, and the performance of LoGs highly depends on the choice of these hyperparameters. To address these challenges, the project puts forth a unified bi-level optimization-based training framework for LoGs with automatic selection of hyperparameters. The project also supports the education and diversity goals of the NSF by integrating LoGs research advances into machine learning courses taught in University of California at Irvine and Rensselaer Polytechnic Institute, making cutting-edge LoGs techniques more accessible to a wider range of researchers and students, fostering innovation and inclusivity in the scientific community. Towards this goal, this project aims to develop a bi-level optimization (BLO) framework for trustworthy and efficient LoG, called BLoG. In addition to the basic algorithm and optimization theory development for BLoG, the project will build a tri-level BLoG problem for robust and adversarial graph neural network training tasks, tailoring gradient-based BLO algorithms to these problems. The project will also develop a BLoG framework with multiple lower-level problems for multiple LoG tasks, named Fast-BLoG. Fast-BLoG will tackle fast and efficient semi-supervised graph neural network training. The project will highlight the advantages and new technical challenges of using the BLoG framework for handling machine learning tasks over graphs. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.