University Of California-Irvine

NSF Awards · FY 2025 · 2025-10

Uncrewed aerial vehicles (UAVs) are equipped with high-resolution cameras and a variety of sensors that can be used in different applications such as agriculture, environmental monitoring, disaster response, or infrastructure inspection. UAVs offer an excellent level of precision and adaptability, can be deployed easily, and provide access to remote regions. UAV-based sensing is a crucial asset in challenged settings, for example in remote regions with limited connectivity, during extreme events, and other similar situations. An optimized UAV trajectory ensures that data is collected in the most efficient and timely manner, from the most relevant locations, and at the most opportune times. This allows for real-time monitoring and quick response to dynamic environments or events of interest. This project develops the general mathematical framework that is needed to optimize UAV trajectories while incorporating practical operational constraints. To address the challenging problem of optimal UAV trajectory and deployment, various research communities have identified different objectives. For example, the communication community has focused on optimizing UAV trajectories to improve network capacity or data collection efficiency, often overlooking the UAV dynamics or the domain constraints. On the other hand, work from robotics and control communities tends to emphasize domain and maneuverability constraints with a focus on system dynamics. This project aims to develop a general framework, addressing the limitations of existing approaches, and incorporating the most critical challenges faced by UAVs. The project tackles several challenges related to trajectory optimization of a group of UAVs, including UAV dynamics, energy efficiency, heterogeneity, 3D sensing, collision and inaccessible region avoidance, and communication network connectivity. These are time-varying challenges, requiring dynamic and distributed solutions. To address these challenges, the project develops a mathematical framework with two related components. The first component develops and uses a heterogeneous quantization theory framework to formulate and solve the UAV trajectory optimization problem to find the desired trajectory. Then, the second component defines quadratic programming optimization problems and solves them to find the closest feasible trajectory to the desired trajectory, found in the first component, that satisfies all required constraints. 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 · 2025-10

The science of networks has emerged as a major catalyst for understanding the behavior of complex interconnected entities, which can be described using graphs. For example, cyber-physical systems including the Internet of Things involve interactions among devices, and social networks can be modeled as graphs capturing various relationships among people or groups. Other complex networks emerge in diverse engineering fields, such as power grids and transportation systems. Graph-based machine learning (ML) and signal processing algorithms exhibit well-documented performance in learning over graphs (LoG). Despite their success, the impact of these algorithms in real-world systems depends heavily on how “socially responsible” they are. While graph-based ML models effectively integrate the nodal attributes with the topological information encoded by the graph, they also inherit and may even amplify potential unfairness. Using such models may subsequently result in unfair outcomes in decision- and policy-making in the related applications. While fairness issues have attracted increasing attention in general ML tasks, they are largely underexplored in the graph domain, especially in terms of theoretical analysis and fundamental understanding. To bridge this gap, the proposed research program will develop a systematic understanding of unfairness in LoG, leading to the design of efficient and principled algorithms for fair LoG. This project will further integrate an educational plan with the research goals, The proposed research will provide novel algorithmic as well as theoretical frameworks for fairness-aware Learning over graphs(LoG). The intellectual merit of this research entails transformative advances at the crossroads of machine learning, optimization, and network science to provide a principled means of mitigating potential bias in learning tasks over graphs. From a theoretical perspective, the proposed research provides new ways to systematically analyze the topological and attributive characteristics that result in unfairness, which are missing in existing studies. From an algorithmic perspective, the project offers principled designs of debiasing methods for mitigating unfairness in LoGs, which fill in the gap in existing works. These theoretical and algorithmic innovations have intrinsic intellectual value and create a “virtuous cycle.” The following research questions will be addressed: (RQ1) How to provide a theoretical understanding of the factors that lead to unfairness in LoG? (RQ2) How should one augment, compress, or generate fair graph data? (RQ3) How to design principled and fair LoG models? 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 · 2025-10

Carbon cycles between the earth, ocean, and atmosphere on geological timescales. The relative amount of carbon stored in each of these respective locations is one important control on Earth’s average temperature. During the last ice age 18,000 years ago, carbon moved out of the atmosphere causing decreased global temperatures. However, it remains unknown where that carbon was stored. This project uses radiocarbon dating to test whether the deep Indian Ocean stored substantial amounts of carbon across the last ice age. The researchers will study microscopic fossils in four sediment cores from across the Indian Ocean basin. These tiny fossil shells made by single-celled organisms record the radiocarbon content of the seawater they lived in at the time of their growth. By comparing the radiocarbon age recorded in these microfossils to the age of the sediment, researchers will determine if deep Indian Ocean carbon was abnormally old during the last ice age. The size of the sediment grains will reveal how quickly currents at the bottom of the ocean flowed and how quickly this old, stored carbon could be transported out of the deep Indian Ocean and released back to the atmosphere. Combined, these data will tell researchers whether changes in deep Indian Ocean carbon were—or were not—important for past changes in Earth’s temperature. The work will support early career scientists and provide research opportunities for graduate students. Seawater radiocarbon content is a powerful tracer of air-sea carbon dioxide exchange and ocean carbon storage. Existing observations suggest that Indian Ocean Bottom Water radiocarbon was significantly lower (i.e., having a much older radiocarbon ventilation age) than all other ocean basins during the Last Glacial Maximum, which could reflect a much slower overturning of these waters and enhanced carbon storage via the biological carbon pump. If these existing measurements accurately reflect the entire Indian Ocean, they suggest that the carbon sequestration capacity of the glacial Indian Ocean has been greatly underestimated. In this project, researchers will create four new glacial- interglacial records of Indian Ocean Bottom Water radiocarbon to answer the following question: Do the existing (very old) Indian Ocean Bottom Water records accurately represent the region? In addition to answering this primary question, collaborators will analyze sortable silt content to estimate ocean current speeds and build this data into an inverse model to improve understanding of the processes driving the observed changes. This project will fund a graduate student and will continue the “First Gen BEES” (Becoming an Earth & Environmental Scientist) program initially designed and produced by first generation college students. 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 · 2025-10

The activity and interconnectivity of neurons, a key type of brain cell, are crucial to the brain's ability to compute and process information. However, recent studies suggest that astrocytes, a different type of brain cell, may also play an important role. This project combines experiments and computational modeling to study how astrocytes contribute to brain function. Astrocytes affect many aspects of neuronal activity and communication, providing a potential mechanism by which they can alter signaling in the brain. The computational modeling and mathematical analysis within the project will enable a deeper biological understanding of these astrocyte-neuron interactions, generate new ideas for why they may be important for information processing in the brain, and suggest ways to integrate these principles into artificial intelligence systems. In conjunction with the modeling will be experiments to observe and manipulate astrocytes in living brains. In so doing, the project will validate new ideas about astrocytes' roles in the brain, providing an enhanced understanding of neural circuits and brain function. The scientific premise of this project is the "contextual guidance" hypothesis, which postulates that astrocytes act as switchboards that transmit information about the environment and the physiological state of the organism to neurons and networks thereof. As such, astrocytes may act as a force multiplier that can expand the repertoire of dynamics that neurons can realize, thus enabling computation. The project will explore two ideas in this regard: (i) that astrocytes actively modulate neuronal dynamics in response to signals sensed from the environment, and (ii) this modulation enables neuronal networks to tailor their dynamics in response to context-specific circumstances. To substantiate these ideas, the project will investigate the role of astrocytes in neuromodulatory systems and subsequent effects on neuronal activity and synaptic plasticity. Furthermore, the project will examine network-level interaction between neurons and astrocytes, exploring features like "tiling," where astrocytes overlay neuron clusters to influence signal routing. In addition to scientific insights, the research will examine how brain-inspired computing may be enhanced by new artificial neural network designs that incorporate astrocytes, with a focus on context-dependent computational paradigms. Additionally, the project includes initiatives to engage trainees in interdisciplinary neuroscience research and exchange, including new mini-courses that bridge neuroscience, engineering and artificial intelligence. A companion project is being funded by the French National Research Agency (ANR). 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 · 2025-10

Cyberattacks on computer networks pose growing threats to critical infrastructure, businesses, and personal data across the United States. Computer security systems that monitor network traffic to detect suspicious activity are essential for protecting against these attacks, but they face a significant challenge: detecting sophisticated attacks often requires analyzing data from multiple organizations, devices, or locations simultaneously. However, sharing network data raises serious privacy concerns because this information can reveal sensitive details about individuals, businesses, and government operations. This project addresses this challenge by developing advanced privacy protection methods that allow organizations to work together to detect cyberattacks without exposing sensitive information. This work serves the national interest by strengthening cybersecurity defenses across critical infrastructure, supporting economic competitiveness through improved data protection, advancing national security through enhanced threat detection capabilities, and enabling compliance with privacy regulations while maintaining robust cyber defenses. This project develops privacy-preserving techniques for graph-based intrusion detection systems that model network traffic and device relationships as interconnected graphs. The research activities include developing specialized cryptographic protocols for essential operations such as sparse matrix multiplications that are fundamental to graph-based analysis. The project will utilize different data partitioning strategies and computational models to perform most operations locally on unencrypted data, minimizing the computational overhead of cryptographic protocols. The team will implement selective revelation of intermediate computational results under differential privacy protection to improve system efficiency while maintaining privacy guarantees. The research extends these protocols to protect the complete training process of graph convolutional neural networks used in intrusion detection, providing comprehensive privacy protection with enhanced computational efficiency. Additionally, the project will support privacy-preserving data provenance in graph-traversal-based detection systems by modeling graph traversal algorithms as matrix operations implemented through the specialized cryptographic protocols. The team will validate these approaches using real-world network datasets and evaluate their effectiveness in collaborative intrusion detection scenarios while measuring privacy preservation and computational performance across diverse network environments. 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 · 2025-10

The growing demand for smart and autonomous systems has driven a surge in the deployment of edge devices. However, the limited computational resources and energy constraints of these devices pose significant challenges for deploying complex deep neural networks (DNNs). Optimizing DNNs for edge devices is crucial to unlock their full potential and enable a wider range of innovative applications. This project’s novelties lie in developing a new generation of tools that can automatically generate hardware accelerators for edge devices while satisfying latency and hardware platform constraints. This project’s impact is to enable high-performance DNN models with high accuracy and fast response to be synthesized in constrained hardware such as Virtual Reality (VR)/Augmented Reality (AR) or assistive robotics will positively change social perception and confidence towards using these future ubiquitous systems. Our approach integrates multi-task learning, neurosymbolic Artificial Intelligence (AI), and high-level synthesis to create accelerators that meet strict latency and hardware platform constraints. In particular, this project introduces NeuHLS, a neurosymbolic approach for merging, compressing, and synthesizing DNNs. NeuHLS’s primary objective is to develop a flexible and efficient framework that balances accuracy, hardware utilization, and latency. In addition, the synthesized hardware must maximize the number of DNN weights implemented using software tunable parameters to allow for flexible fine-tuning at runtime. The proposed NeuHLS toolchain consists of three phases. The first one aims to merge a set of single-task DNNs into one multi-task DNN by sharing representations from different single-task DNNs, hence reducing the model size. Next, the toolchain exploits recent advances in symbolic knowledge distillation to compress the multi-task DNN into a neurosymbolic model, which can then be processed by novel neurosymbolic high-level synthesis techniques that optimize the deployment while balancing accuracy, hardware utilization, and latency. The team of researchers will evaluate the toolchain using existing benchmarks and real-world application deployment in the various domains of autonomous systems. 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 · 2025-10

As large language models (LLMs) transform how we interact with information, this project explores how such generative AI technologies can be tightly integrated into the core of data analytics systems. The goal is to make advanced artificial intelligence (AI) capabilities a seamless part of data processing pipelines, allowing organizations to ask richer questions and get faster, more insightful answers from their data. By enhancing a prior system called EnrichDB—which successfully combined machine learning with database queries—this research aims to expand its capabilities to work with modern LLMs, improving how data is enriched, interpreted, and acted upon. Beyond technical innovation, this work could have real-world impact in areas like disaster response, where quick and intelligent analysis of fast-changing data is critical. The research will also contribute to workforce development through student involvement and collaborations with public agencies. This proposal aims to explore mechanisms to seamlessly embed large language models into data processing in the context of large scale data analytics applications. A key contribution will be the design of a middleware component, the LLM Extender, that mediates between user analytics tasks and a set of LLMs available through API-based access with diverse performance, cost, and latency profiles. The middleware will support automatic selection of appropriate LLMs per task, optimizing for user-defined tradeoffs among quality, cost, and latency. The integration of the LLM Selector into EnrichDB will allow real-time query execution to dynamically invoke LLM-based enrichment functions. This preliminary work will assess the feasibility of embedding LLMs into large-scale, real-time data analytics and lay the groundwork for future agentic data systems. The research will also explore applications in disaster planning and response, leveraging existing partnerships with local agencies during the Great California Shakeout. 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 · 2025-10

Artificial intelligence (AI) is becoming a key partner in critical human decision making — from diagnosing diseases to everyday driving. To ensure these partnerships work effectively, this project develops new ways to accurately measure and track how well both people and AI perform tasks over time. These assessments will help determine in what situations it is better for a human to take the lead in decision-making and in what situations the AI can be trusted. Equally important is creating AI systems that respect and reinforce human goals, like fairness, teamwork, and preserving a sense of personal control. By combining techniques from statistical modeling and cognitive science, this project will improve how people and AI collaborate, making these interactions not just effective but also aligned with human values and expectations. The research project is interdisciplinary in nature, building on Bayesian learning and cognitive modeling to systematically study and optimize human-AI interaction. It focuses on three core research activities: (1) developing Bayesian inference frameworks to dynamically evaluate the evolving abilities of both humans and AI agents; (2) creating adaptive optimization algorithms to manage decision policies under uncertainty, incorporating costs and constraints inherent in real-world human-AI collaboration; and (3) exploring human-centered aspects, integrating subjective metrics such as perceived fairness, teamwork, and agency into multi-objective optimization frameworks. Behavioral studies across various tasks—including image classification, natural language question answering, visual target tracking, and simulated navigation—will validate theoretical models and algorithms. Additionally, the project will create openly accessible datasets to support reproducibility and facilitate further research on collaborative human-AI systems. 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 · 2025-10

In today's rapidly evolving intelligent systems landscape, spanning various smart operations such as traffic and energy grid management to deploying sensors for environmental monitoring, optimizing resource allocation is crucial for efficient decision-making and policy formulation. Despite numerous advances in the field of optimization theory, current methods for optimal resource allocation problems often falter, especially in large, distributed networks where devices must coordinate without the presence of a central authority, resulting in slow solutions and high communication demands. This project addresses these fundamental limitations in combinatorial optimization problems, specifically submodular maximization under matroid constraints—an NP-hard problem that regularly appears in optimal discrete resource allocation. While efficient mathematical approaches have been developed to achieve near-optimal solutions in centralized form for submodular maximization under matroid constraints, these solutions often incur significant computational costs. In distributed settings, in addition to these computational costs, the optimal resource allocation process is further exacerbated by high in-network communication costs and local data disclosure issues. Our objective is to develop practical solutions that reduce in-network communication costs, maintain favorable trade-offs in optimality gaps, and accelerate solutions. The project will deliver practical distributed algorithms with analytically characterized optimality gaps and proven practicality measures. The project will also provide vital educational and training opportunities for students from K-12 to graduate level, preparing the future workforce in optimal decision-making for autonomous systems. This research aims to significantly advance the theory and practice of distributed submodular maximization under partition matroid constraints. We will focus on the continuous relaxation approach, which has improved the optimality gap for centralized solutions from 0.5 to 0.63. Our primary goal is to develop practical distributed algorithms based on this approach that maintain strong optimality guarantees while drastically reducing communication and computational costs. We will achieve this through two key technical avenues: first, by exploring methods like hard-thresholding and localized gradient approximations to limit data exchange among agents; and second, by accelerating convergence rates using novel variance reduction techniques based on control variates and advanced algorithms that reuse gradient information. The project emphasizes rigorous formal analysis to characterize optimality gaps and will validate these algorithms through extensive computer simulations and experimental studies, including applications in multi-agent robotic coverage and smart grid systems. Outcomes will include a catalog of provably correct distributed optimization strategies and open-source software tools. 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 · 2025-10

Chronic stress during pregnancy is a serious and often underestimated threat to maternal and fetal health, linked to complications such as preeclampsia, preterm birth, and low birth weight. Unfortunately, current methods for monitoring stress in pregnant women are inadequate: questionnaires provide only subjective, infrequent data, lab tests are invasive and sporadic, and existing wearable devices capture only general physiological signals that are easily confounded by other factors. This project aims to fill that gap by creating a novel wearable tool for comprehensive stress monitoring. The device will continuously track both physiological stress indicators (e.g., heart rate variability) and molecular stress biomarkers (e.g., hormones and inflammatory markers) in real time. By enabling early and accurate detection of harmful stress levels, this project’s goal will allow timely interventions to protect maternal and fetal health. In doing so, it aligns with NSF’s mission by promoting scientific progress, advancing national health and welfare, and fostering interdisciplinary training. The project main objective is to develop an innovative wearable research platform that integrates novel molecular sensors with physiological sensors to enable comprehensive stress assessment during pregnancy. The team will create new electrically regenerable molecular sensors that can repeatedly detect multiple key stress biomarkers (e.g., cortisol and inflammatory cytokines) in interstitial fluid. These sensors will be embedded into a flexible wristband alongside modules for monitoring physiological signals like heart rate variability and skin conductance. The integrated device will simultaneously and noninvasively measure molecular and physical indicators of stress in real time. Finally, the platform will be tested on human subjects under controlled stress conditions to validate its accuracy and reliability. This interdisciplinary work will demonstrate a transformative wearable sensing paradigm, laying the groundwork for future continuous stress monitoring. 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 · 2025-10

Machine learning-based approaches are increasingly replacing manually designed heuristics in network controllers such as those used for adaptive bit rate video streaming and for Internet congestion control. However, these black-box learned models introduce new challenges for network operators. Key concerns include selecting appropriate training data, testing under realistic and variable network conditions, ensuring debuggability, and maintaining safety during online deployment. This project aims to develop novel explainability tools and techniques to enable effective human-AI collaboration during the design and deployment of learning-based network controllers. This project plans to pursue three synergistic thrusts to realize trustworthy learning-based network controllers. First, this project focuses on developing a concept-based explainer that interprets model behavior using high-level, human-understandable concepts that are easy for an operator to understand. Second, it plans to create a hybrid explainability framework that integrates concept-based, low-level feature-based, and predictive future performance-based explanations to provide a comprehensive and multi-layered understanding of controller decisions. Third, the project will investigate how these explanations can enhance key operational tasks, including data curation for training, test coverage evaluation, debugging, and root cause analysis. Together, these thrusts aim to bridge the gap between black-box models and practical network management needs. The domain-specific explainability solutions developed in this project will offer high-level insights into learned models, enabling network operators to better understand, trust, and manage learning-based controllers in real-world networks. By improving interpretability, these tools aim to lower the barrier to deploying high-performance learning-based solutions in practical network environments. In addition, they are expected to enhance the robustness and reliability of these controllers, enabling safer adoption of learning-based solutions in critical network applications. This project also involves close collaboration with industry partners and the development of a web-based platform to promote broader engagement, dissemination, and adoption of the developed tools. 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 Metastatic brain tumors represent the most common form of malignancy in the central nervous system. In women with metastatic breast cancer, up to 35% of patients will suffer from brain metastasis and will have little to no therapeutic options to treat them. Current treatments are often complicated by the protected environment of the brain with its blood-brain barrier, distinct anatomical structure, metabolism, and unique immune composition. Given the high prevalence of breast cancer in women in the United States, this poses a significant unmet clinical need. Immune modulation represents a promising therapeutic approach, but the efficacy of these treatments is often hindered by the presence of suppressive immune cells, such as regulatory T cells (Tregs). Tregs are a population of immunosuppressive cells that serve to control immune responses and prevent autoimmunity. Tregs have been reported to be present in brain metastases and have been used as a prognostic factor in tumor progression and patient outcomes. Systemic depletion of Tregs can promote anti-tumor immunity, but this approach also confers unwanted systemic autoimmunity. Thus, specific mechanisms to target Treg function in the brain without completely depleting them is necessary. My preliminary data demonstrates that Tregs accumulate readily within brain metastases and their depletion confers a complete elimination of tumor mass. This response results from an enhanced effector T cell infiltration and an enriched antigen presentation program in microglia, the brain-resident macrophage population. Co-depletion of microglia in Treg depleted mice results in a complete reversal of this tumor rejection. This emphasizes the central role that microglia play in the anti-tumor response in brain metastases and demonstrates a need to target Treg-microglia interactions in this setting. To this end, the proposed study will aim to understand and target Treg modulation of microglia, allowing for the potent and brain-specific targeting of Treg function. First, I will determine if antigen presentation is a central microglial function suppressed by Tregs to identify targets to inhibit Treg function specifically within the brain. I will also assess the therapeutic potential of deleting CD73, an ecto-5’-nucleotidase expressed on Tregs that converts AMP into adenosine, which has been shown to suppress microglia function through A3R, an adenosine receptor exclusive to microglia. This work will pave the path to the generation of novel therapeutics to impair Treg-mediated suppression in the brain and generate desperately needed treatments for patients with brain metastases.

NIH Research Projects · FY 2025 · 2025-09

PROJECT SUMMARY/ABSTRACT Bile acids have been implicated as a contributor to atherosclerosis and is commonly found to be elevated in patients with metabolic disorders and cardiovascular disease (CVD). Recently, we observed extrahepatic bile acid release in high-fat/high-sucrose (HFHS) fed and LDLR KO pig models of atherosclerosis. To reveal novel targets to treat BA abnormality in patients with CVD, we aim to use the deep-sequencing technique, ribosome profiling (ribo-seq), to investigate translational control of BA synthesizing and trafficking genes in colon, kidney, liver, and muscle in HFHS-fed and LDLR KO atherogenic pig models (Aim 1). Additionally, Ribo-seq can be used to identify small open reading frames (smORFs) which encode microprotiens. These microprotiens are often overlooked and underappreciated as key players in biology and are highly enriched in mitochondria and often found to perturb metabolic processes. Thus, we aim to filter Ribo-seq data for smORFs encoding mitochondrial microproteins and determine their impacts on cellular metabolism using untargeted metabolomics/lipidomics followed by relevant stable isotope tracing to identify affected enzymes/pathways.

NIH Research Projects · FY 2025 · 2025-09

Project Summary Obstructive sleep apnea (OSA) is known to be correlated with cardiovascular diseases (CVD), yet controlled clinical trials showed that CPAP treatment was insufficient to alleviate CVD risk in all patients. Current tools fail to identify the mechanistic pathways involved in OSA-related CVD pathophysiology, and residual CVD risk may persist despite CPAP treatment. Conversely, some studies have identified OSA as possibly protective against myocardial infarction preconditioning consequences, highlighting significant knowledge gaps in understanding the mechanisms linking OSA and CVD. This research project aims to elucidate the complex relationship between OSA and CVD through an innovative "heart-on-a-chip" platform combined with validated intermittent hypoxia (IH) protocols. Specifically, we will develop and validate physiologically relevant IH protocols using endothelial cells, comparing responses against patient tissue samples. These validated protocols will then be applied to our "heart-on-a-chip" model to investigate the direct impact of OSA-induced IH on cardiac structure and function, providing insights into OSA-specific cardiovascular diseases. Additionally, we will examine whether different patterns of OSA-associated IH can provide cardiovascular protection through preconditioning mechanisms prior to ischemic insult and identify the underlying pathways. This comprehensive approach addresses current limitations in OSA research by providing physiologically relevant in vitro models that can directly test competing hypotheses about OSA's cardiac effects. Our research has the potential to identify novel therapeutic targets and improve the clinical management of OSA patients at risk for cardiovascular complications.

NIH Research Projects · FY 2025 · 2025-09

PROJECT SUMMARY Obstructive sleep apnea (OSA) is known to be correlated with cardiovascular diseases (CVD), yet controlled clinical trials showed that CPAP treatment was insufficient to alleviate CVD risk in all patients. Current tools fail to identify the mechanistic pathways involved in OSA-related CVD pathophysiology, and residual CVD risk may persist despite CPAP treatment. Conversely, some studies have identified OSA as possibly protective against myocardial infarction consequences, highlighting significant knowledge gaps in understanding the mechanisms linking OSA and CVD. This research project aims to elucidate the complex relationship between OSA and CVD through an innovative "bed to bench" approach, integrating clinical, in silico, and in vitro techniques. Specifically, we will develop and validate a novel mathematical model to quantify tissue-specific hypoxia burden in OSA patients, correlating this with clinical outcomes and traditional OSA metrics. This model will be validated using superoxide expression in microvascular tissue from patient skin biopsies and further refined with large-scale sleep study databases from the Sleep Heart Health Study (SHHS). Second, we will optimize in vitro microvascular models, including an endothelial cell model and an advanced Vascularized Micro-Organ (VMO) platform, to study the direct effects of intermittent hypoxia (IH) on the microvasculature. The IH protocol will be developed to match gene expression profiles of IH-exposed endothelial cells to patient biopsy profiles to ensure physiological relevance. Third, we will employ a "heart-on-a-chip" model to investigate the direct impact of OSA- induced IH on cardiac structure and function, providing insights into OSA-specific cardiovascular pathways. Throughout these aims, we will leverage data from the National Sleep Research Resource to inform our models and validate our findings against real-world patient outcomes. This comprehensive approach addresses current limitations in OSA research by providing more accurate risk assessment tools and physiologically relevant in vitro models. Our strategy has the potential to uncover new biomarkers, identify novel therapeutic targets, and ultimately improve the clinical management of OSA patients at risk for cardiovascular complications.

NIH Research Projects · FY 2025 · 2025-09

PROJECT SUMMARY/ABSTRACT Targeted-protein degradation (TPD) is a novel modality that harnesses cellular proteolysis to eliminate disease-causing proteins. Proteolysis-targeting chimeras (PROTACs) are small molecule degraders that induce proximity of the proteasomal machinery with a substrate. Such approaches hold the promise to revolutionize modern medicine as they target classically undruggable families that lack active sites and offer safer therapeutic windows due to their catalytic pharmacology. However, PROTACs are limited by the capacity of proteasomes. A multi-step assembly of ubiquitin ligases precedes POI ubiquitination, further complicating degrader design and efficacy. Additional challenges include acquired lysine mutations, a small fraction of available E3 ligases, and an inability to effectively troubleshoot the events leading up to proteasomal degradation. Thus, several limitations must be addressed before the potential of degrader modalities can be fully realized. Lysosomes are the alternative degradative route and have been exploited for extracellular proteins by Lysosomal Targeting Chimeras (L YTACs). These exciting developments highlight the potential of lysosomes to combat the central limitations of proteasomal degraders, but are wholly restricted to extracellular proteins. Degraders of intracellular proteins rely on ubiquitin modifications that direct proteins to proteasomes. Recent work shows that arginine methylation modifications direct proteins to lysosomes during normal homeostatic maintenance in unmodified, natural conditions. Preliminary work shows that this pathway can be activated through proximity between a substrate and a protein arginine methyltransferase. We hypothesize that chemically inducing arginine methylation will be sufficient to target diverse protein classes for degradation in lysosomes. The proposal's main goals are to characterize the mechanisms for applying this degradation route synthetically (Aim 1) and to identify endogenous substrates of this degradation pathway that may later be amenable to synthetic degradation (Aim 2). The outcome of this study will generate a degrader modality that is independent of ubiquitin and offer insight into endogenous networks of protein homeostasis. This research holds long-term potential to improve therapeutic tractability across diverse disease states.

NIH Research Projects · FY 2025 · 2025-09

PROJECT SUMMARY The national comprehensive societal cost from alcohol-impaired crashes is estimated at over $348 billion, with $296 billion attributable to ≥ .08 BAC cashes. In the last 10 years, the national alcohol impaired driving crash fatality rate has increased by 24%. Young drivers are the most vulnerable group and at highest risk of being seriously injured and/or killed in an alcohol impaired driving crash. Much of this vulnerability exists in the context of young drivers navigating life at a time when individual freedom and mobility via driving is high as is exposure and ease of access to alcohol and drugs. Prior to turning age 21, the Minimum Legal Drinking Age and Zero- Tolerance laws (making it unlawful for those <21y/o who drive to have a BAC ≥.02) are in effect for youth and young drivers with the intent of reducing harm and death due to negative consequences of alcohol use. However, when a young driver who drinks turns 21y/o, they are no longer subject to key effective alcohol prevention/public safety policies. Instead, when getting behind the wheel, they are subject to a BAC per se policy with a threshold of ≥.08 g/dl. Unfortunately, this is a higher BAC limit with a well-established greater risk of serious injury and fatal crash (i.e., BAC per se of ≥.02 before age 21 vs. a BAC per se of ≥ .08 upon turning age 21y/o). Given the current state of an increasing national alcohol crash fatality rate and high vulnerability of young adult drivers, there is a critical need and salient opportunity to innovate policy focused in the young-impaired driver domain. An early reduced risk exposure approach at the time young drivers turns 21y/o and can legally drink could yield measurable harm and fatality reduction effects. Except for the state of Utah, where the BAC per se policy is ≥.05, all US states are at a BAC per se of ≥.08. International studies prove reductions in BAC limits to .05 significantly reduce alcohol-impaired traffic injuries and deaths. Using epidemiologic, qualitative, and system dynamics research methods, this study will provide a robust examination and modeling of a conceptual national age-based Graduated-BAC per se policy (Grad-BAC) so that at the moment a young driver turns 21y/o through age 24y/o, the BAC per se would be ≥.05. Thereafter, at age 25y/o, the BAC per se would be ≥.08 (except for Utah, already at a BAC per se of ≥.05). This study will first examine longitudinal pre/post-age 21y/o driving after drinking behavior as well as state-/national-level crash fatalities specifically among drivers 21-24y/o by BAC levels. Further, changes in fatal crash rates among drivers 21-24y/o pre- vs. post-BAC per se policy of ≥.05 in Utah and in neighboring states will be evaluated. Next, national public survey and focus groups of state policy leaders will be conducted to assess support for or against the Grad-BAC per se policy vs. a national BAC per se of ≥.05 for all drivers ≥21y/o. Finally, the construction of a novel and comprehensive simulation system dynamics model will facilitate the examination of potential effects of the Grad-BAC policy among 21-24y/o that drive after drinking. This study will effectively leverage historical and recent landmark findings in impaired driving research and policy. Findings will prove to be highly innovative and pivotal in informing new policy and prevention efforts.

NIH Research Projects · FY 2025 · 2025-09

Abstract The decline of urethral function with advancing age plays a crucial role in urinary incontinence in women. However, none of the current treatments address this decline, and the mechanisms by which aging impacts urethral physiology remain little known and largely unexplored. Our recently published work evaluated the effect of aging through functional, morphometric, and transcriptomics studies on the female mouse urethra. Similar to aged female humans, we demonstrated that aged mice have substantially lower functional leak point pressure and significant morphometric decline in striated muscle and elastin, with increased fibrotic connective tissues compared to young mice. Gene expression profiling of the whole urethra showed that myogenesis and fibrogenesis pathways were predominantly enriched in aged urethral tissue. Further investigation of the RNA- sequencing data revealed that aged urethral tissues had more fibrotic and extracellular remodeling gene expressions, such as connective tissue growth factor (Ctgf), compared to younger tissues. Connective tissue associated with striated muscle is primarily generated by resident Pdgfrα+ cells known as fibro-adipogenic progenitors (FAPs) with multilineage progenitor properties and a fibroblast-like phenotype. Immunofluorescence staining of Pdgfrα+ cells in the urethral striated muscle layer of both young and aged mice, coupled with RNAscope analysis in our preliminary studies, revealed an increased density of Pdgfrα+/Ly6a+ FAP cells and elevated Ctgf expression from these Pdgfrα+/Ly6a+ FAP cells in the striated muscle layer of the aged compared to young mouse urethra. These results suggest a critical role of Ctgf in FAP cells for urethral dysfunction. Therefore, we hypothesize that aged urethral tissue striated muscle dysfunction is attributed to increased fibrosis due to altered FAP cells that secrete factors such as Ctgf and that targeting Ctgf can improve urethral function. To test this hypothesis, Specific Aim 1 will focus on single-cell RNA sequencing of urethral tissues from three different age groups (young adult, middle-aged, and aged) to evaluate gene expression changes in FAP cells due to aging (Aim 1.1). Spatial transcriptomics will be used to define the FAP cell locations and their communication networks with surrounding tissues to better understand their role in female urethral dysfunction (Aim 1.2). Blocking CTGF activities has shown promise in treating fibrosis in other tissues and organs, and our current data shows that Ctgf is highly expressed in aged tissues. Therefore, in Specific Aim 2, we aim to evaluate the efficacy of a CTGF-antibody drug as a potential therapy for urethral dysfunction (Aim 2.1). By inhibiting the activities of CTGF, we aim to decrease the formation of fibrotic tissues in the urethra and preserve urethral function. The impact of our proposed work will result in a deeper understanding of female urethral dysfunction and provide new therapeutics for treating urethral dysfunction, offering new opportunities for advancements in this area.

NIH Research Projects · FY 2025 · 2025-09

PROJECT SUMMARY While recent immunotherapies for Alzheimer’s disease (AD) have shown promise, they come with significant side effects and may not work equally well for patients at different stages of AD. This necessitates a continued focus on the development of AD therapeutics, especially on identifying the earliest drivers of AD. First, it is critical that we know where in the brain to find the earliest events that contribute to AD pathogenesis. Evidence from human literature indicates that the default mode network, in particular the retrosplenial cortex/posterior cingulate gyrus (RSC/PCC)/precuneus regions, shows early amyloid deposition with concomitant elevations in metabolism, though the functional implications of this observation are unknown. Anecdotal evidence also suggests that spatial navigation may be amongst the earliest impairments observed in preclinical AD, which is consistent with dysfunction of RSC. These findings all point to a potentially important role of the RSC in AD pathogenesis, yet what is occurring in the RSC, when it occurs, and how it may contribute to AD-related disease has not been explored. In this proposal, we will define how neuronal hyperexcitability in the RSC contributes to the development of AD- related behavioral deficits in mice, and the molecular mechanisms that underlie this effect. Our pilot data provide strong support for a causative role of hyperexcitability in RSC layer 5 cells in the disease-associated loss of memory in mice, as silencing RSC layer 5 cells prevented the age-associated loss of memory recall in AD model mice. In this application, we will first define the effect of inhibition of RSC layer 5 cells on hippocampus-associated spatial memory and spatial navigation tasks, and test when hyperexcitability emerges in these cells in Aim 1. This will be done in 3 different mouse models, including amyloid and tau models. In Aim 2 we will explore the transcriptional changes that occur in the RSC during disease progression in three different mouse AD models. In Aim 3, we will assess the transcriptional changes that occur in the human RSC using human postmortem samples from control, mild cognitive impairment, and AD patients. We will focus on identification of conserved gene expression modules that change during the development of AD in the mouse and human brain, validating our results using spatial transcriptomics and immunohistochemical validation. This study will provide support for a critical role that RSC hyperexcitability plays in AD progression, and a mechanistic framework for how changes in intrinsic excitability and synaptic function contribute to the development of AD.

NSF Awards · FY 2025 · 2025-09

This I-Corps project is based on the development of an artificial intelligence (AI) drug development process to generate drugs specific to various diseases. Current approaches to drug discovery rely on costly screening or simulations that fail to capture the complex biological context in which drugs operate. These methods typically examine proteins in isolation rather than within their biological pathways, leading to high failure rates due to unforeseen off-target effects and poor efficacy. This technology addresses these challenges by integrating a systems-level understanding with structural biology through the Bone Morphogenetic Protein (BMP) pathway foundation model. BMPs are a group of proteins that play a crucial role in bone and cartilage formation, as well as in various other developmental processes. This approach may predict how drugs interact within complex biological networks rather than just with individual targets. The solution may create more accurate predictions of both therapeutic potential and possible side effects, potentially improving the drug discovery process and patient survival rates while lowering the average cost of drug development. This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of an integrated AI platform that combines systems biology with structural predictions for more effective small molecule drug discovery. This technology is based on a therapeutic generative diffusion model that proposes novel drugs and a transformer-based foundation model of systems biology to prioritize biological targets of interest. Current structure-based drug discovery systems often optimize small molecules according to a target of interest without considering off-target effects. The technology unites a novel systems biology foundation model for the Bone Morphogenetic Protein (BMP) pathway with a guided diffusion model for small molecule generation. This integration may enable the targeting of key protein interactions while minimizing common off-target effects like nausea or hair loss during chemotherapy. The key advance over existing solutions is the inclusion of systems biology data in the generation of novel therapeutics to reduce the probability of off-target effects, which are often responsible for the failure of clinical trials. This solution may allow pharmaceutical companies to repurpose existing drugs using knowledge from this foundation model, generate alternative therapeutics for existing targets to hedge development programs, prioritize experiments to reduce the time to market, and decrease failure rates in clinical trials. 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

This I-Corps project focuses on the development of a digital transaction analysis system. The technology uses advanced algorithms to quickly analyze large amounts of fast-moving data. It creates real-time summaries of complex transaction data, making it possible to calculate detailed insights quickly and accurately, something that most current data analysis systems struggle to do on time. The key innovation is a set of algorithms that are proven to be accurate, fast, and efficient with memory. The algorithms are also designed to work well across multiple computers at once. This makes the approach better and faster than current systems, which often rely on basic averages or slow processing. The project provides a data analytics platform that leverages advanced algorithms to process petabyte-scale datasets and high-velocity data streams in real-time. By creating compact data summaries that retain essential information while dramatically reducing computational costs, the technology enables fraud detection systems to make accurate decisions within milliseconds while handling billions of transactions. This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of the technology. This solution is based on the development of advanced algorithmic techniques that derive insights from high velocity data streams. Specifically, the technology employs mathematical constructs to create real-time summaries of massive transaction flows. This technique allows for the efficient computation of complex analytical measures. The capabilities are often infeasible for existing systems to achieve accurately and within the strict latency requirements of real-time environments. The scientific advance lies in the use of algorithms that offer provably optimal guarantees for accuracy, speed, and memory efficiency, coupled with inherent parallel computation capabilities essential for distributed processing. The approach improves on traditional methods that rely on simple statistical aggregates or slow batch processing. Users benefit from the adoption of this technology through superior automated decision-making. 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 2026 · 2025-09

PROJECT SUMMARY/ABSTRACT Thyroid Eye Disease (TED) is an inflammatory condition that can lead to severe vision problems due to cytokine production, fibroblast activation, and glycosaminoglycan deposition that cause vascular congestion and tissue edema. Insulin-like growth factor 1 (IGF1R) plays a key role in TED's development, impacting endothelial cell migration, fibrosis, and neovascularization; thus, TED patients can be prescribed teprotumumab, an IGF1R inhibitor. However, there is no standardized diagnostic or therapeutic tool for TED, making it crucial to develop rapid imaging methods and quantifiable biomarkers to offer personalized patient care. Spatial Frequency Domain Imaging (SFDI), developed at UCI's Beckman Laser Institute (BLI), is a promising solution. SFDI uses several wavelengths to assess tissue properties (oxy and deoxyhemoglobin, oxygen saturation, as well as absorption and reduced scattering coefficients) up to 5 mm in depth. SFDI has successfully predicted surgical outcomes and treatment responses in other medical conditions. Dr. Ediriwickrema’s research aims are to use SFDI to non-invasively identify optical tissue properties in TED, particularly those related to vascular congestion and tissue edema (Aim 2A), as compared to age-matched controls (Aim 1). She will also establish an optical signature profile for TED patients receiving teprotumumab (Aim 2B). The team will also conduct clinical pathology correlations (CPC 1, 2A, 2B) on surgical specimens from both healthy and TED patients. This K23 proposal provides comprehensive training for Dr. Ediriwickrema and enables her to develop noninvasive diagnostic and therapeutic response metrics for TED and similar inflammatory conditions. Dr. Bernard Choi, a recognized expert in wide-field imaging applications in the characterization and treatment of microvascular-related diseases, will mentor Dr. Ediriwickrema. Secondary mentors, Drs. Anthony Durkin, Anand Ganesan, Krzysztof Palczewski, and Kristen Kelly bring extensive expertise in the development and application of in vivo SFDI techniques, multiphoton microscopy for assessing cutaneous inflammatory changes, and advanced high-resolution molecular imaging techniques, such as 2-photon imaging. Additionally, Drs. Vivek Patel, Vasan Venugopalan, and Maria Estopinal will provide valuable guidance in TED management, optimization of SFDI-derived tissue optical parameters, and histopathologic evaluation of periocular inflammatory disease. Dr. Ediriwickrema will benefit from the rich scientific environment at UCI, including resources in the Department of Ophthalmology, BLIMC, Institute for Clinical and Translational Science (ICTS), Center for Translational Vision Research (CTVR), and Biostatistics Epidemiology and Research Design (BERD) Unit. This K23 training will equip Dr. Ediriwickrema with the skills to apply SFDI effectively to establish objective and continuous optical tissue biomarkers that characterize TED activity and response to treatment, and enable her to become an independent physician-scientist specializing in oculofacial inflammatory diseases.

NIH Research Projects · FY 2025 · 2025-09

Project Summary Surface-enhanced Raman scattering (SERS) is a popular approach for molecular biosensing, and for a good reason: it enables the identification of molecular targets in a label-free fashion, down to the level of single molecules. This tremendously attractive capability has triggered developments that aim to use SERS for rapid screening of sequences of molecular residues, which would enable applications such as sequencing the nucleobases in DNA. SERS-based sequencing overcomes persistent problems with label-based sequencing methods, including the ability to read out longer sequences and the possibility of directly detecting epigenetic modifications to the nucleobases. Despite successful examples of Raman-based sensing of single molecules, the photophysics of single-molecule SERS measurements on plasmonic substrates limits spectral acquisition rates to the 0.1-10 Hz range, many order of magnitude below the rates needed to render SERS practical for DNA analysis and other biosensing applications where detection speed is of the essence. In this proposal, we overcome the fundamental speed limit in surface-enhanced Raman-based biosensing in particular and label-free single-molecule sensing in general. We achieve label-free detection of single molecules at acquisition rates that are at least 104 faster compared to SERS measurements. Such unprecedented detection speeds are obtained by combining surface-enhanced coherent anti-Stokes Raman scattering (SE-CARS) with the state-of-the-art design of dielectric enhancement structures. Through systematic tailoring of dielectric nano-antennas, and arranging them in a lattice, this project delivers a thermally robust enhancement platform that enables label-free single molecule detection with microsecond signal acquisition times. Our team is comprised of experts in coherent Raman scattering microscopy and advanced engineering of nanophotonic metasurfaces. The proposed work is innovative in that it uniquely fuses the accelerated signal rates of coherent Raman scattering with the latest insights in dielectric nano-antenna design. By pushing the limits of both Raman photophysics and the collective resonances afforded by metasurfaces, our innovation tackles a major obstacle in label-free biosensing and opens up new high- speed applications that have hitherto remained out of reach.

NIH Research Projects · FY 2025 · 2025-09

Project Summary Alzheimer’s disease (AD) currently affects ~6.7 million individuals in the U.S. alone and is expected to grow to 13 million by the year 2050. A major component of its devastation is the progressive loss of the patient’s ability to form memories. Treatments for rescuing memory function in AD patients do not exist, due in part to insufficient research performed to characterize the activity of memory-supporting neural circuits compromised by the disease. The effort to develop memory-restorative therapeutics that intervene in neural circuits requires vulnerable neuronal subtypes impacted during pathophysiological progression, and how they relate to memory performance. Neurons in the entorhinal cortex (EC) act as a gateway for sensory inputs feeding into the hippocampus. This EC-hippocampus circuit is critical for memory formation and retrieval. Despite its significance to AD pathophysiology, it remains unclear which circuit components are vulnerable in the EC of AD patients or animal models. Our laboratory’s pioneering previous research has developed exciting preliminary results that suggest a path forward. Our overall hypothesis is that circuit vulnerability of the EC, including the lateral entorhinal cortex and medial entorhinal cortex, to tauopathy causes associative memory impairment in AD. We will use two Tau mouse models to identify EC cell types vulnerable to tauopathy. The project is expected to yield advances toward the development of therapeutics to mitigate associative memory functions impaired in AD.

NIH Research Projects · FY 2025 · 2025-09

PROJECT SUMMARY/ABSTRACT Acute pancreatitis (AP) is a prevalent gastrointestinal disorder characterized by severe abdominal pain, leading to substantial hospital admissions worldwide. The current reliance on opioid analgesics for pain management in AP introduces risks such as abuse, dependence, and possibly exacerbated disease outcomes. Emerging evidence suggests that the endogenous lipid-signaling molecules anandamide and palmitoylethanolamide (PEA), which produce analgesia through non-opioid mechanisms, could offer a promising alternative for pain relief. However, their therapeutic potential is limited by rapid breakdown via the enzyme fatty acid amide hydrolase (FAAH). This project seeks to explore FAAH-regulated lipid signaling as a novel, opiate-sparing therapeutic target for managing AP pain. Preliminary studies have demonstrated that inhibiting FAAH leads to significant antinociceptive effects in animal models of AP, supporting the hypothesis that FAAH plays a crucial role in modulating AP pain through the degradation of anandamide and PEA. This research project is structured around three specific aims: First, to identify the receptor systems, including CB1 cannabinoid receptors, PPARα, and others, that mediate the analgesic effects of FAAH inhibition in AP. Second, to assess changes in FAAH expression and activity, alongside levels of anandamide and PEA, providing insight into the regulatory mechanisms at play during AP. A subsequent aim will delve into the cellular substrates mediating analgesia from FAAH inhibition, employing state-of-the-art techniques such as snRNAseq and CyTOF to profile cell populations and their transcriptional landscapes. The final aim addresses whether anandamide and PEA exert redundant or synergistic analgesic effects in AP. By revealing the intricate mechanisms of FAAH-regulated signaling and its impact on pain relief, this project stands to significantly advance our understanding of non-opioid analgesic pathways in AP. Success in these aims could lay the groundwork for the development of novel, effective, and safer pain management strategies for AP, reducing opioid reliance and its associated risks.