University Of California, San Diego

NSF Awards · FY 2025 · 2025-01

A prominent paradigm used in both mathematics and computer science is the “structure vs randomness'' paradigm. At one end, it aims to identify structure in mathematical objects that is useful for their mathematical analysis and for the design of efficient algorithms for their manipulation. At the other end, it aims to use randomness and random-like properties for the analysis of objects that lack structure. Randomness is a common mathematical tool used throughout mathematics and computer science. One can sometimes identify “random like'' properties of mathematical objects that are as useful as true randomness for their study and analysis. In the best-case scenario, randomness can be identified as the absence of structure, and the two notions of structure and randomness can be seen as complementary. The project will expand the bridge between mathematics and computer science and may have a practical impact on widely adopted algorithms. Integration of research and education is a key component of the project. This project focuses on a novel approach towards this “structure vs randomness'' paradigm, whose main goal is to obtain significantly improved quantitative bounds for many important problems in mathematics and computer science. At the core of this new paradigm are two new concepts: spreadness and mixing. Spreadness is a quantification that there are not too many elements in any structured subset of the ambient universe and mixing is a measure for how two objects behave jointly via some common operation. The project aims to develop both the theory and applications of this new approach. On the theory side, we aim to build a versatile framework that can both connect existing applications and extend them to new ones. On the applications side, the proposal highlights exciting potential applications across several fields - additive combinatorics, communication complexity, graph theory, and fast combinatorial algorithms. 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 · 2025-01

The EEG-DaSh project aims to enable Artificial Intelligence (AI)-driven breakthroughs in the realms of electroencephalography (EEG) and magnetoencephalography (MEG) by providing the necessary data and associated tools to process them. EEG and MEG are widely used methods for recording brain activity, which can be harnessed through AI to gain deeper insights into brain function and neurological disorders. EEG-DaSh addresses the current lack of large-scale, standardized datasets by providing annotated data from over 25 partner labs and reformatting data from existing archives. The project also focuses on educating scientists on utilizing AI in EEG and MEG research and offers free and open high-performance computing resources to support such efforts. The EEG-DaSh Gateway Project has three main objectives: First, create a comprehensive data-sharing platform for neuroelectromagnetic data featuring a user-friendly web interface for easy data management. This platform adopts a new data format optimized for machine learning and deep learning applications and uses the Hierarchical Event Descriptor (HED) system for standardized data annotation. Second, the project develops advanced automated processing pipelines, accessible through high-performance computing resources, to facilitate machine learning, deep learning, and EEG model fitting. These features include cutting-edge techniques for feature extraction, measures of critical brain dynamics, measures of connectivity between regions of interest, and customized neural fields and neural mass models for data augmentation. Deep learning example scripts that are compatible with EEG-DaSh will also be provided. Third, the project organizes workshops and provides in-person support to promote data sharing, train researchers, and foster collaboration in neuroscience. These efforts enhance the understanding of brain function and accelerate the development of AI applications in neurotechnology. A companion project is being funded by the US-Israel Binational Science Foundation (BSF). 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 · 2025-01

PROJECT SUMMARY Human milk is considered the recommended sole source of human nutrition worldwide for at least the first six months of life and is the recommended continued source of human nutrition for at least the first two years of life. Yet, from a scientific perspective this basic human tissue type has been grossly understudied. The UC San Diego Human Milk Research Biorepository (HMB) was established in 2014 as a first-of-its-kind environmental epidemiology cohort (EEC) consisting of indexed human milk samples and associated clinical data that is available to the scientific community to aid our understanding of this critical source of infant nutrition. This open cohort study has accumulated over 110,000 aliquots of human milk samples to date from across the nation along with associated broad-based longitudinal data on infant health and development. Accordingly, it provides unprecedented current and future value to researchers and trainees as well as public health entities to better understand the short, intermediate, and long-term role of human milk in human health and disease. However, the existing cohort is lacking diversity in several critical areas, including racial/ethnic diversity and socioeconomic diversity. This proposal aims to increase the diversity of the HMB EEC by partnering with local lactation groups and academic and/or community partners in targeted diverse settings to enhance the recruitment of diverse populations into the study. Namely, we will recruit 300 new participants who represent low or low-middle income households and/or who self-identify as Black or Hispanic/LatinX. We further aim to improve the accessibility of the HMB EEC data to the scientific community through key enhancements to the study website and resource infrastructure. Finally, we will engage in efforts to raise awareness of the HMB EEC as a research resource that is available to lactation scientists who themselves represent populations that are underrepresented in medicine or who focus their research on questions pertaining to minority health and health disparities. This will be accomplished via the establishment of partnerships with historically Black colleges and universities and other minority-serving institutions to explore opportunities to enhance the diversity of the workforce involved in breast milk research. The HMB EEC provides unmatched value to the research community and public health entities to better understand the role of breast milk in child health and development across the lifespan. By enhancing the diversity of the cohort and improving the accessibility of the data to the scientific community at-large and specifically to diverse groups of human milk researchers, this EEC will provide greater benefit not only in the area of minority health and health disparities-related breast milk research but will also provide a unique cross-sectional snapshot of the “health” of the breast milk supply across the nation that can be monitored over time, by location, and by specific demographic groups.

NIH Research Projects · FY 2026 · 2025-01

PROJECT SUMMARY Cerebral blood vessels are critical to deliver blood, containing oxygen and nutrients, to every neuron and glial cell in the brain. In addition, the endothelial cells comprising central nervous system blood vessels possess unique properties, termed the blood-brain barrier (BBB), that allow them to tightly regulate the movement of ions, molecules, and cells between the blood and the brain, thus, tightly regulating the extracellular environment of the brain. Studies have indicated that neurovascular dysfunction, including hypoperfusion and BBB dysfunction, may be an important component of Alzheimer’s disease pathogenesis. This vascular dysfunction could lead to accumulation waste products such as amyloid, leakage of serum components to the brain, and increased neuroinflammation, all of which could have important impacts on the neurodegeneration that is observed in patients with Alzheimer’s disease. Despite the potential importance of neurovascular dysfunction, very little is known about the molecular changes to cerebral blood vessels in patients with Alzheimer’s disease, and how these changes may affect neurovascular function and the progression of Alzheimer’s disease pathology. We have performed proteomics on blood vessels isolated from human patients with Alzheimer’s disease and from age and sex matched controls. Interestingly, we found downregulation of pathways involved in angiogenesis and barrier formation, which may account for the hypoperfusion and BBB dysfunction, respectively, observed in Alzheimer’s disease patients. The proteomic data also showed robust downregulation of the transcription factor SOX18 in the blood vessels of patients with Alzheimer’s disease. We further used mice models to downregulate Sox18 expression in brain endothelial cells and found very similar changes to those found in Alzheimer’s disease patients, including a downregulation of angiogenesis and barrier formation pathways. Therefore, we hypothesize that downregulation of SOX18 in brain endothelial cells is key to the neurovascular dysfunction observed in patients with Alzheimer’s disease. In this grant we will first examine the role of Sox18 in modulating neurovascular function in mice, including angiogenesis and barrier formation. We will then use human induced pluripotent stem cell models of the BBB to identify the molecular mechanisms by which SOX18 regulates neurovascular function. We will then determine whether modulating Sox18 expression can alter the disease course in multiple mouse models of Alzheimer’s disease. These experiments will allow us to understand the role of SOX18 in neurovascular function, and determine whether targeting SOX18 could limit the pathogenesis of Alzheimer’s disease.

NIH Research Projects · FY 2025 · 2024-12

Project Summary: Genetic susceptibilities and environmental insults together shape risk of developmental disorders. Identifying convergent pathways among diverse risk factors is critical to developing therapeutic targets. One such candidate pathway is oxidative stress. However, the role of oxidation-reduction (redox) signaling in neurodevelopment is poorly understood. We will address this gap by characterizing redox signaling during normal forebrain neurodevelopment in mice and its response to environmental exposures. Using a genetically encoded redox sensitive fluorescent biosensor, we will determine the redox state of developing neurons at critical developmental timepoints. Embryos will then be exposed to the known oxidant herbicide paraquat and acetaminophen to determine the sensitivity of neuron subclasses to oxidative stress during different developmental stages. A commonly used antipyretic in pregnancy, acetaminophen metabolites deplete cellular antioxidants, making cells more susceptible to oxidative stress. The response of developing neurons to redox modifying compounds like paraquat and acetaminophen are of substantial clinical and

NIH Research Projects · FY 2026 · 2024-12

Abstract While radiotherapy (RT) is a mainstay in the treatment of glioma, most patients will experience radiation-induced neurocognitive decline due to injury to critical brain tissue. The pathogenesis of radiation-induced neurocognitive decline is not well understood and the potential risk factors for neurocognitive decline including tumor-related, other treatment-related, and patient-specific parameters are unclear. In this proposal, designed to respond directly to RFA PAR-21-329 on neurotoxicity of cancer therapies, we will perform a prospective longitudinal study to clinically characterize RT-associated cognitive decline using a novel patient-centered approach (cognitive phenotyping), uncover the mechanisms underlying these changes using quantitative multimodal imaging (MMI), and identify risk and resilience factors that moderate the relationship between RT-induced brain injury and cognitive decline. This work will guide future interventions for cognitive-sparing radiotherapy that aim to optimize neurocognitive outcomes in patients with glioma. Specific Aims: 1. To characterize neurocognitive heterogeneity and identify cognitive phenotypes in glioma patients before and longitudinally after RT. 2: To analyze mechanisms underlying radiation-induced cognitive decline by measuring MMI biomarkers of structural and microstructural damage and vascular sufficiency over time. 3: To identify tumor-, treatment-, and patient-related factors (including demographic, cerebrovascular health, and social determinants of health; SDH) that impact a patient’s vulnerability to RT-induced brain injury and increase risk for cognitive decline. Study Design: We will prospectively enroll adult patients with glioma receiving fractionated partial brain radiotherapy on an open longitudinal, observational study (NCT05576103) with total target accrual of n=250. We will recruit English and Spanish-speaking patients who are diverse in race, ethnicity, and socioeconomic class from a broad geographical region. Patients will undergo standard of care clinical MRI, including diffusion, 3D volumetric, and perfusion imaging along with comprehensive neurocognitive, mood, and quality of life assessments at discrete endpoints: baseline (pre-RT); 3-, 6-, 12-months post-RT, then yearly for up to 5 years. Socio-demographic, clinical, tumor-related, and treatment variables will be collected, along with assessments of tumor progression. Cognitive phenotypes and MMI signatures of brain health will be derived pre-RT and their trajectories will be measured up to 5 years post-RT. Finally, risk and resilience factors related to vascular health and SDH will be explored as potential moderators of RT-induced brain injury and cognitive decline. This study has strong implications for public health because it will identify modifiable factors which impact treatment-related cognitive decline in patients with glioma. This study also strives to improve representation of minority patients in clinical trials, which will increase generalizability and the clinical impact of our findings.

NIH Research Projects · FY 2026 · 2024-12

PROJECT SUMMARY Recent research suggests that primate-specific APOBEC3C-mediated C-to-T DNA and ADAR1-mediated A-to- I RNA editing represent essential drivers of human cancer progression. Though many DNA and RNA editing sites have been identified, the functional relevance of APOBEC3C and ADAR1 deaminase deregulation, especially in primary patient samples, is still unresolved. By elucidating the relative roles of APOBEC3C and ADAR1 in myeloproliferative neoplasm (MPN) progression to secondary acute myeloid leukemia (sAML), this represents a unique opportunity to understand how DNA and RNA editing deregulation drive cancer stem cell generation and therapeutic resistance. While pro-inflammatory cytokine-responsive ADAR1 (adenosine deaminase acting on RNA 1) and APOBEC3 (apolipoprotein B mRNA editing enzyme, catalytic polypeptide like type 3) base deaminases restrict viral replication5 and LINE element retrotransposition, base deaminase deregulation has been linked to both genomic and epitranscriptomic (post-transcriptional modification) instability. We have shown that pro-inflammatory cytokines activate primate specific APOBEC3C-mediated cytidine to thymidine (C-to-T) deamination of DNA and ADAR1 p150 isoform-mediated adenosine to inosine (A-to-I) deamination of double-stranded RNA (dsRNA) thereby fueling transformation of pre-LSCs to LSCs in myeloproliferative neoplasms (MPN). The overall goal of this proposal is to continue our investigation of malignant deaminase activation in directing human pre-LSC evolution to sAML LSCs with the ultimate aim of informing the development of effective strategies that predict and prevent transformation to rapidly fatal sAML. We will investigate if ADAR1 shRNA knockdown reduces telomere length, TERT expression and alters the self-renewal program in pre-LSCs and LSCs in stromal co-cultures and humanized mouse models. In Aim 2, we proposed to examine the detailed mechanisms of how ADAR1 drives WNT/beta-catenin signaling in primary patient samples in stromal co-cultures and humanized mouse models as well as reporter and confocal fluorescent microscopy. Lastly, we will test if activation of both APOBEC3C and ADAR1p150 fuels pre-LSC evolution by whole genome sequencing, Tapestri single cell DNA, and proteomic sequencing platform. In addition to vastly expanding our knowledge of A-to-I editing function in progenitor cell maintenance, this research program will inform the development of malignant ADAR1 editase detection and inhibition strategies that may help to prevent progression of MPNs to acute myeloid leukemia.

NIH Research Projects · FY 2026 · 2024-12

SUMMARY Formation of long-lived memory T cells is a critical feature of the adaptive immune system which enables the efficient control of recurrent infection. Tissue-resident memory T (TRM) cells are a unique and transcriptionally heterogenous population of memory T cells that provide host protection against pathogens through their positioning within barrier tissues and rapid effector response. Current treatments for viral infections lack the ability to enhance the TRM cell protective response largely because our understanding of TRM biology lags that of circulating memory T cell populations. To dissect the transcriptional regulators of TRM across distinct tissues, I have chosen to focus on T cell factor 1 (TCF1), a transcription factor that is critical for the formation of central memory T cells yet is largely unstudied in the context of TRM cell biology. TCF1 has been shown to repress lung TRM, however its role in TRM cell formation and maintenance in other non-lymphoid tissues such as the intestines, salivary gland, kidney, liver, and pancreas are unknown. Furthermore, TCF1 is expressed in multiple isoforms yet the role of the individual isoforms of TCF1 have never been investigated in the context of TRM cell formation in any tissue. Additionally, there is a growing body of evidence which suggests that TCF1 expression during the first week of an acute viral infection defines cells that are fated to become memory cells versus terminal effector cells. However, the mechanism behind TCF1 regulation of this fate decision remains largely unknown. This proposal seeks to address these gaps in knowledge by investigating the hypothesis that TCF1 has distinct roles in regulating CD8+ T cells throughout the course of infection from effector CD8+ cell expansion, memory precursor formation, and TRM cell formation and maintenance. To investigate this hypothesis, we will complete the following two specific aims: (Aim 1) Determine the role of TCF1 in the formation and maintenance of CD8+ TRM cells across distinct tissues; (Aim 2) Investigate the role of TCF1 in early effector CD8+ expansion and memory precursor cell formation across distinct tissues. Identification of the molecular mechanisms regulating TRM cell formation and maintenance will build on our understanding of the transcriptional regulation of TRM and will have a positive impact on therapeutics designed to modulate tissue-specific TRM. The experiments and career development opportunities outlines in this proposal will greatly prepare the applicant for a successful career in academic research.

NSF Awards · FY 2024 · 2024-12

Faint dwarf galaxies are sensitive to dark matter and early universe physics. The investigators will develop a model that comprehensively predicts the population of dwarf galaxies surrounding the Milky Way as a function of dark matter properties and early universe dynamics. Comparing these predictions to the data will yield constraints on dark matter particle models and cosmic inflation. The key product of this work is an efficient and accurate model for the Milky Way’s dwarf galaxy population in non-standard cosmologies. This work will train early career researchers in broadly-applicable computational methods and support undergraduates from underrepresented groups in physics through the National Society of Black Physicists and Society for the Advancement of Chicanos/Hispanics and Native Americans in Science programs. Dwarf galaxies occupy the smallest dark matter halos that can form stars. These systems are extremely sensitive to unknown dark matter and early universe physics. To date, the faintest dwarf galaxies have exclusively been detected as satellites of the Milky Way. This situation presents a theoretical challenge: to discover fundamental physics in these data, we must model the specific formation history of our Galaxy’s dark matter halo and satellite population. The investigators will perform new suites of cosmological simulations constrained to match the Milky Way, with initial conditions appropriate for a variety of dark matter and early-universe scenarios, which will be used to calibrate the Galacticus galaxy formation model. The calibrated model, Galacticus-MW, will rapidly and accurately predict the Milky Way satellite population as a function of the beyond-CDM linear matter power spectrum. Comparing these predictions to the data will yield the first near-field measurement of small-scale matter clustering, along with constraints on dark matter properties (including its particle mass, interactions, and production mechanism) and on cosmic inflation. This work will enable cosmological inference using dwarf galaxy populations detected by Stage IV surveys over the next decade. 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 · 2024-12

ABSTRACT Alzheimer's Disease (AD) is a devastatingly progressive, fatal neurodegenerative disease. With approximately 30 million patients affected worldwide, it is the most common form of dementia and is predicted to grow exponentially. The pathology of AD is driven by Amyloid Precursor Protein (APP) and its proteolytic processed products, primarily the Aβ42 peptide, leading to amyloid plaques, Tau- containing neurofibrillary tangles and neuron death. Although there are now multiple FDA approved drugs to treat AD, including monoclonal antibodies (mAbs) that target Aβ plagues, unfortunately none of these target the root cause of AD, namely APP. Consequently, there remains a great unmet medical need to develop AD therapies that selectively target APP driven pathogenesis to prevent cognitive dysfunction and neuron death. The field of RNA therapeutics has come of age in the clinics. There are currently 16 FDA approved RNA therapeutics, including 4x Phosphorodiamidate Morpholino Oligonucleotides (PMO-ASOs), 7x Phosphorothioate Anti-Sense Oligonucleotides (PS-ASOs) and 5x siRNAs, with >50 ongoing early and late stage clinical trials. Due to their Watson-Cricket base pairing mechanism of target engagement, RNA therapeutics have exquisite on-target selectivity for all mRNAs, including APP, with minimal off- target genetic effects. Impressively, a single RNA therapeutic dose can achieve a 3 to 6 month pharmacodynamic response in the clinics. However, endosomal escape remains the rate-limiting problem to solve with only ~1% of PS-ASOs and 0.3% of siRNAs escaping from endosomes. Due to the endosomal escape problem, PS-ASO and siRNA therapeutics require excessively high doses resulting in cytotoxicity and safety profile concerns. In contrast to siRNAs and PS-ASOs, non-protein binding PMO ASOs can be dosed 20-times higher than PS-ASOs with no safety signals arising. Moreover, have longest metabolic stability and duration of response. To selectively target APP, we will develop new technology to generate a highly efficacious APP PMO therapeutic that efficiently escape the endosome by conjugation to a universal endosomal escape domain (uEED). We will test and optimize hAPP PMO- uEEDs in a humanized APP mouse model of Alzheimer's Disease.

NIH Research Projects · FY 2026 · 2024-12

PROJECT SUMMARY Heart failure is a result of aberrant matrix remodeling in the left ventricle; male patients experience ventricular wall thinning, whereas females experience ventricular wall thickening. While current pharmacologic agents palliate symptoms associated with heart failure, the morbidity of the disease remains high as these treatments do not provide a universal cure, necessitating a sex-specific understanding of heart failure. Mechanical cues and gonadal hormones regulate the sex-specific cardiac remodeling observed during heart failure through changes in the extracellular matrix (ECM). Resident cardiac fibroblasts that reside in the ventricular wall largely regulate aberrant ECM remodeling during heart failure, but the sex-specific regulation of cell-mediated ECM remodeling remains poorly understood. Both sex hormones and sex chromosomes have been implicated in regulating sex differences in cardiac ECM remodeling, but current methods do not decouple their synergistic effects. My proposal seeks to decouple the effects of sex chromosomes and sex hormones using gonadectomized animal models and biomaterials-based cell culture tools that recapitulate the cardiac ECM during fibrotic stiffening. Our laboratory tools enable our central hypothesis: sex chromosome combination and gonadal hormones synergistically regulate myofibroblast phenotypes, and X-chromosome linked genes and estrogen dosage will promote myofibroblast matrix remodeling. The hypothesis will be investigated through the following aims: Aim 1 will improve our understanding of myofibroblast populations that drive sex-specific matrix remodeling independent of gonadal hormones; Aim 2 will use engineered hydrogels to mimic sex-specific myofibroblast phenotypes in vitro; Aim 3 will investigate the role of estrogen and estrogen receptor expression on mediating the myofibroblast phenotype independent of sex chromosome combination. By leveraging animal models and biomaterial technologies, this study will unveil a robust understanding of the myofibroblast phenotype by decoupling the effects of sex chromosomes and sex hormones. If successful, the proposed research will significantly advance the understanding of sex-specific biology in myocardial fibrosis.

NIH Research Projects · FY 2025 · 2024-12

PROJECT SUMMARY Aging is a continuous degenerative process that begins in adulthood and is associated with systemic loss of function and increased risk of neurodegeneration. The circadian clock, which orchestrates the temporal organization of physiology and behavior in 24-hour cycles, regulates the rhythmic expression of up to 40% of the transcriptome and has a fundamental role in aging and longevity. The circadian clock decays with brain aging, which manifests as increased night-time wakefulness, poor sleep efficiency, and reduced amplitudes in body temperature rhythms. Furthermore, the expression of core clock genes decreases and shifts toward advanced phases in the human aging prefrontal cortex. High-amplitude circadian rhythms correlate with well- being and increased lifespan in animal models, while weaker circadian rhythms and fragmented activity patterns precede the development of dementia in older adults. Understanding the crosstalk between aging and the circadian clock and harnessing the power of chronobiology promises novel strategies to support healthy aging, promote longevity, and mitigate cognitive deterioration and disease in the elderly. This proposal requests partial support for the 15th UCSD Center for Circadian Biology Symposium, "Biological Time Keeping, Aging, and Disease,” to be held at the Scripps Institute of Oceanography auditorium in San Diego from March 21 to 22, 2025. The symposium, which attracts approximately 150 attendees worldwide, will focus on chronobiology, aging, and neurodegeneration to define priority topics and advance joint research. The meeting will feature 23 invited speakers, a poster session, and three workshops. The symposium speakers and session chairs are recognized leaders in their areas and were chosen to represent biologists, neuroscientists, chronobiologists, and data scientists specializing in aging to foster new cross-collaborations to advance the field. We aim to attract scientists from diverse backgrounds through targeted advertisement and, with NIH support, to offer travel fellowships to trainees and junior faculty, prioritizing those from under- represented groups. Training aspects of the meeting include four lunch workshops covering tools and data modeling for circadian analysis and aging research, career opportunities, and chronobiology-based work-life balance strategies aimed at supporting trainees and new faculty and to expand their expertise and networks.

NIH Research Projects · FY 2026 · 2024-12

PROJECT SUMMARY Microbiomes are crucial to the physiological functioning of host organisms. Understanding what regulates microbiome assembly and function can therefore have critical implications for host health. The central hypothesis of this proposal is that host phylogenetics is a key driver of microbiome assembly and function, which in turn, has physiological implications for the host. Specifically, I aim to 1) assess how host phylogenetics and host traits regulate microbiome assembly and function, and 2) assess the effects of host microbiomes on host cell physiology and consumer fitness. Unicellular algae are an ideal model system to address these questions because these microbes harbor an experimentally tractable exterior microbiome. Further, algae span an unusually broad phylogenetic distance across two domains of life, allowing us to test for the implications of over a billion years of host evolutionary divergence. Lastly, our prior work has shown that algal microbiomes are species-specific, but whether this specificity corresponds with host evolutionary history or has implications for host health is unknown. Therefore, I use this model system to test my hypothesis that microbiome composition and function will be most similar among host species that are closely related, and more divergent among host species that are distantly related. To test this, I submerged 31 strains of axenic algae in a diverse community of freshwater bacteria to allow each strain to assemble a microbiome. I am using samples from my study to assess bacterial abundance, community composition, and function of the microbiomes via amplicon and shotgun sequencing. To test whether host specificity has implications for host health, I will carry out reciprocal transplant experiments of whole microbiome communities among nine host strains that span phylogenetic distances from hosts within the same species to those from different domains of life. I will pair these microbiome swaps with measurements of host cellular stress, including profiles of nutrients and lipids. Considering that microbiomes can facilitate host acquisition of limiting nutrients, including nitrogen and phosphorus, I hypothesize that hosts with an assembled microbiome will have higher cellular %N and %P compared to hosts without a microbiome. Further, under the expectation that microbiome composition will correspond with host phylogenetics, I hypothesize that the magnitude of change between metrics of host cellular stress will correlate with phylogenetic divergence between hosts. Further, I will test whether these varied algal microbiome treatments affect consumer fitness using a zooplankton model system, thereby clarifying how the microbiomes of the foods we eat may affect human health. Overall, I aim to unravel the mechanisms behind microbiome assembly and the consequential implications of host-specific microbiome assembly on host health. Using the unparalleled phylogenetic breadth of hosts available in an algal model system, this study will clarify how hosts and their microbiome have co- evolved over evolutionary history and the implications of such host-specificity on host and consumer health.

NSF Awards · FY 2024 · 2024-12

Cloud service providers offer Field Programmable Gate Arrays (FPGAs) as a time-shared service for efficiently accelerating high-value workloads such as machine learning, genome sequencing, databases, encryption, and other applications with strict security requirements. While the hardware is time-shared between multiple tenants, there is generally believed to be no information leakage between subsequent users since the FPGA bitstream and memories are digitally erased after each tenant’s use. The project studies “FPGA pentimenti” data that leaks between subsequent users through analog effects. The project’s broader significance and importance are developing techniques for securing cloud infrastructure and promoting education and research in FPGA security. This project studies, characterizes, and develops mitigations for FPGA pentimenti. Specifically, this project investigates how data from previous users is leaked via an analog side channel due to bias temperature instability effects. This project establishes bounds of data-recovery capabilities within the cloud FPGA model and identifies effective defenses for all stakeholders. This project also characterizes techniques for extracting FPGA pentimenti. With this knowledge, this project develops mitigations to reduce or eliminate these analog side-channel attacks from the perspective of the manufacturer, cloud provider, and end-user. 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 · 2024-11

SUMMARY Type I interferons (IFN-I) are essential in the defense against viral infection and cancers. However, aberrant, or excessive IFN-I production can be pathogenic and lead to the development of autoimmune diseases. Plasmacytoid dendritic cells (pDCs) produce more IFN-I than any other cell type and are thus critical to establish systemic defense against viral pathogens. Conversely, pDC dysfunction is strongly associated with many interferonopathies. We and others have found that although pDCs initially produce exceptional amounts of IFN-I following a viral infection, they lose their capacity to produce these cytokines, becoming “exhausted”, a few days later. This phenomenon, which is conserved in mice and humans, provides the virus an opportunity to persist and exposes the host to increased risk of secondary infections. pDCs with an exhausted phenotype have also been reported in many tumors. However, the mechanisms underlying pDC IFN-I production in general as well as pDC exhaustion are still not well understood. Although many studies have identified individual regulators of pDC IFN-I production, such as the transcription factor IRF7, to date no genome-wide investigation to identify regulators of pDC function has been performed. This is likely because pDCs are rare and short lived, and the available pDC-cell lines are suboptimal. We have now generated an improved pDC cell line (4C1), which produces high levels of IFNα and IFNβ, both of which are dependent on IRF7 and toll-like receptor (TLR). We have also engineered this new cell line with an IFN-I reporter and constitutive Cas9 expression to allow for genetic screening of pDC IFN-I regulators. Our preliminary data indicate that this reporter-pDC cell line can reliably measure IFN-I production after stimulation and can be genetically manipulated in a straightforward manner. Furthermore, we have developed a protocol for inducing exhaustion in this reporter-pDC cell line, which recapitulates the exhaustion gene signature and the cell-intrinsic TLR7 dependency that we reported during in vivo pDC exhaustion. Combining our new tools along with the Genome-Scale CRISPR Knock-Out (GeCKO) library, we propose the first genome-wide screen for regulators of both human pDC function and exhaustion. Furthermore, using protocols well established in our laboratory for the manipulation of human primary pDCs, we propose to verify and further characterize the top regulators selected from the screen in pDCs from healthy volunteers. Finally, we will validate these results using in vivo mouse models of viral infection. Our proposed work will provide the first genome-scale analysis of IFN-I regulation in pDCs, and give us exceptional and novel insights into the mechanisms that regulate IFN-I production and subsequent exhaustion in this unique and critical innate cell. These insights can then be leveraged to design therapies which are targeted to modulate IFN-I production in multiple human diseases.

NIH Research Projects · FY 2025 · 2024-11

PROJECT SUMMARY Mitochondrial dysfunction and endoplasmic reticulum (ER) stress are hallmarks of pathologic aging and are intricately linked in the onset and pathogenesis of etiologically-diverse neurodegenerative disorders including Alzheimer’s disease (AD) and related dementias (ADRDs). This has led to significant interest in understanding how cells regulate mitochondria in response to ER stress. Intriguingly, the ER stress-responsive kinase PERK is localized to ER-mitochondria contact sites where it acts as an effector of both the unfolded protein response (UPR) and the integrated stress response (ISR). Additionally, PERK-dependent transcriptional and translational signaling modulates nearly all aspects of mitochondrial biology including remodeling of mitochondrial cristae and respiratory complexes to enhance energy capacity, regulation of mitochondrial proteostasis (i.e., protein import, chaperone activity, and proteolysis), and remodeling of membrane phospholipid composition to induce protective mitochondrial elongation. Through these mechanisms PERK protects mitochondria during ER stress; however, persistent PERK activation induced by severe or chronic ER stress leads to apoptosis. Thus, PERK signaling both promotes adaptive mitochondrial remodeling and dictates cell fate in response to varying levels of cellular stress. The importance of PERK in regulating adaptation and survival is further supported by clinical, genetic, and pharmacologic evidence demonstrating that imbalanced PERK signaling contributes to the pathogenesis of etiologically-diverse neurodegenerative diseases. Hypomorphic variants in the gene that encodes PERK (EIF2AK3) predispose individuals to tauopathies such as progressive supranuclear palsy (PSP) and late-stage AD. In addition, exogenous PERK activation mitigates tau pathology in PSP, further indicating that protective PERK signaling is insufficient in the pathogenesis of this disease. Collectively, these observations establish PERK as a critical regulator of mitochondrial adaptation to cellular insult and suggest that imbalances in PERK signaling contribute to mitochondrial dysfunction implicated in neurodegenerative disease pathogenesis. Using cell culture models derived from patients expressing a hypomorphic PERK variant, I will show that deficiencies in PERK signaling impair mitochondria and contribute to neurodegenerative phenotypes such as tau pathology (Aim 1). Further, I will demonstrate that pharmacologic activation of the ISR—a stress-responsive program comprised of the eIF2α kinases GCN2, HRI, PKR, and PERK—mitigates mitochondrial dysfunction and improves neuronal survival in a human neuronal model of PERK-deficient neurodegeneration (Aim 2). These efforts are significant as they will define a critical role for PERK in regulating mitochondrial adaptation during neurodegeneration and establish pharmacologic ISR activation as a potential therapeutic strategy against neurodegenerative disease—for which no disease-modifying treatments are currently available.

NSF Awards · FY 2024 · 2024-11

The field of program synthesis aims to create tools that can automatically create a program from a specification of desired behavior. Synthesis holds the promise of easing the burden on programmers (e.g., by finding solutions to tricky special cases automatically), and allowing non-programmers to create programs merely by indicating the outcome that they want the program to produce. Unfortunately, current synthesis tools do not scale up to large-scale programming problems, a situation that threatens to doom this promising technology to being a niche field. This project's novelties are ways to exploit compositionality in program synthesis in a way that allows one to create bigger programs out of smaller ones. The project builds on a recent framework called SemGuS (Semantics-Guided Synthesis), which offers a foothold on the expressibility and scalability problems of program synthesis. In principle, the framework can support the synthesis of software in layers, where implementation choices in one layer are hidden from other layers (and thus consistent with modular software design with information hiding). The goal of the project is to capitalize on the opportunity that SemGuS offers for extending synthesis to much larger systems than was possible heretofore. The work will lead to more scalable and general synthesis algorithms, with potential benefits to synthesis applications that are already in widespread use. Further development of the SemGuS framework has the potential to make synthesis more usable and programmable, and thereby allow users to carry out synthesis tasks without prior knowledge of existing synthesis tools. Most importantly, compositional synthesis will allow synthesis to scale to larger applications with more practical relevance. 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 · 2024-11

ABSTRACT Epithelial cells are exposed to a wide range of environmental pathogens due to their location at body surfaces. Aside from functioning as a physical barrier, epithelial cells also coordinate immune defenses through receptor-mediated signaling. However, we do not fully understand the mechanisms that shape pathogen recognition and immune signaling in epithelial cells. The overall goal of this proposal is to elucidate the molecular mechanisms that regulate the outcome of viral infection in epithelial cells. There is a critical need for an in vivo animal model that facilitates mechanistic studies of epithelial immunity. Infection with Orsay virus–a naturally occurring intestinal pathogen of the nematode Caenorhabditis elegans–provides an innovative approach to study cellular interactions between a virus and its natural host in the context of a whole-animal model. An obligate, intracellular pathogen, Orsay virus invades C. elegans intestinal cells and induces the activation of the Intracellular Pathogen Response (IPR), a transcriptional defense program that confers pathogen resistance (Sowa et al. 2020). This antiviral transcriptional response requires DRH-1, a homolog of mammalian RIG-I-like receptors (RLRs). RLRs are intracellular pattern-recognition receptors (PRRs) that detect viral RNA to initiate an antiviral immune response. Notably, DRH-1/RLR is one of the few PRRs conserved between C. elegans and humans. Despite the similarity between RLRs and DRH-1, C. elegans lacks sequence-based homologs to the downstream signaling components of the RLR pathway–including MAVS, IRF3, NF-κB and interferon. It is unclear which host determinants are involved in DRH-1-mediated activation of the IPR and where DRH-1 is required to induce the IPR. The proposed studies integrate molecular tools and a natural infection model to address the central hypothesis that DRH-1 signals through a non-canonical RLR signaling pathway and functions in intestinal cells to induce the IPR. The central hypothesis will be tested through two specific aims. Aim 1. Determine how DRH-1 signals to activate the Intracellular Pathogen Response. Aim 2. Define the tissue specificity and subcellular localization of DRH-1 during Orsay virus infection. The expected outcome of this project is a mechanistic understanding of how C. elegans DRH-1 coordinates host-virus interactions at the epithelial barrier. This work will highlight either the evolutionary conservation or rewiring of RLR signaling in epithelial cells. Collectively, the proposed studies will reveal a novel form of antiviral immunity in C. elegans. More broadly, the proposed studies will have positive translational impact by elucidating a novel antiviral pathway that may provide insight into the regulation of innate immune responses during viral infection in humans.

NSF Awards · FY 2024 · 2024-10

The overarching theme of the project is to systematically expand understanding of how deep neural networks (DNNs) work and why or when they are better than classical methods through the lens of "adaptivity." Adaptivity refers to the properties of an algorithm that take advantage of favorable structures in the input data without knowing that these structures exist. That is, adaptive algorithms are those that are free of tuning parameters and could automatically configure themselves to adapt to each input data. The anticipated outcome of the project includes a new theory that explains and quantifies the adaptivity of popular DNN models such as multi-layer perceptrons, self-attention mechanisms (namely, transformer models), and meta-learning. The theory could result in substantial savings in the statistical and computational complexity of these models, allowing them to be applied in resource-constrained settings and to have more environmentally friendly energy footprint. This project will also provide opportunities for students and postdocs to explore interdisciplinary research topics related to deep learning. Specifically, this project investigates (1) the "local adaptivity" of DNNs in estimating functions from noisy data; (2) the "relational adaptivity" of self-attention mechanism that parses a structure data point (such as an image or a chunk of text); and (3) the "task adaptivity" of multi-task and meta-learning algorithms that learn to share information across multiple tasks. The research covers some of the most popular DNN models. Technically the project leverages multiple branches of mathematics (such as function classes, nonparametric statistics, statistical learning theory, optimization, and compressed sensing) and involves innovations in the approximation-theoretic understanding, algorithmic insights, and statistical theory of DNNs. The new analytical tools to be developed are also of independent interest to the broader machine learning theory community. 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 2024 · 2024-10

This project is motivated by the increasing public concerns on privacy issues, new legislations and the high demand for privacy enhancing technologies such as differential privacy (DP) in applications from both private and public sectors. The overarching theme of the project is to address the pressing new challenges that arise as differential privacy transforms from a theoretical construct into a practical technology. The project advances the state-of-the-art of research in the area of DP, and contributes to privacy education. On the research front, the project develops new algorithms and analytical tools that enable more precise privacy accounting and higher utility in DP. On the education front, the project involves training future leaders in DP areas, creating educational materials and expanding an open-source software library called autodp that makes state-of-the-art differentially private computation more accessible. Collectively, the integrated research and educational activities contribute to ongoing collaborative efforts in building innovative applications of differential privacy. The project has three main components in use-inspired fundamental research. The first component unifies the recent breakthroughs in DP, such as, Renyi DP, moments accountant, f-DP and produce an intermediate functional representation that allows lossless conversions among these representations. The second component focuses on investigating the stronger privacy properties permitted by the structures of the actual data, and addressing the dilemma of interpreting worst-case privacy on average-case data. The third component focuses on using a public dataset to ``denoise'' the private data releases or to facilitate private machine learning. The outputs of the research will be broadly shared through integration in autodp library, and will be integrated in courses. 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 2024 · 2024-10

This project aims to serve the national need of mitigating the underrepresentation of women in science, and therefore contributes to scientific progress. Peer mentorship is one of the most effective avenues to increase the status of women in academia. By providing intense peer-mentorship to junior women scholars of international relations, the project will promote high-quality scholarship and contribute to an increased success of mentored women in academia. The project will promote these goals through the hosting of intense peer-mentorship workshops to foster networks, provide feedback and support, disseminate information, and encourage psychological resilience. The project also tracks the success of peer-mentorship programs through survey research and a collection of data on academic success. Studies that have evaluated the status of women in international relations over the past 30 years reveal significant gender gaps on numerous dimensions. The continued under-representation of female scholars at top research institutions and high ranks harms scientific progress. Recent research demonstrates that active mentoring, especially through workshops that foster networks, provide feedback and support, disseminate information, and encourage psychological resilience, are among the most promising avenues for change. The Journeys in World Politics workshop program has mentored young women scholars of International Relations (IR) since 2004. The project hosts annual three-day workshops that support 18-20 participants and includes research presentations by junior scholars, feedback from discussants, oral autobiographies by senior scholars, and career and gender discussion sessions involving topics such as networking, work-life balance, and navigating classroom gender dynamics. Beyond the workshops, the project maintains an active website and other forms of communication, arranges meetings at conferences, and thereby builds a broad network of women in the entire political science discipline. To track the success of mentorship workshops, the project collects more systematic data to evaluate the mechanisms through which mentoring programs increase long-term success rates for female political scientists. 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 2024 · 2024-10

Large language models (LLMs) are an artificial intelligence (AI) technology that promises to revolutionize programming by translating a user's informal intent expressed in natural language into computer code. This technology has the potential to democratize programming and allow anyone, regardless of their skills, to generate code from a simple task description. However, LLMs do not offer any guarantees about the quality of the code they generate, or whether the generated code actually does what the user intended. With LLMs becoming popular, it is thus crucial to build formal techniques that can produce code that provably matches the user's intent and convince the user that the code will do what is expected of it. Recent work has proposed grammar-constrained decoding as a way to enforce that the output generated by large language models belongs to the language of a user-provided formal grammar. This project will contribute new grammar decoding techniques that can align LLMs with formal specifications and enable efficient generation of high-quality code. To this end, this project will integrate program analysis and synthesis techniques from formal methods with structured prediction methods from natural-language processing. Concretely, the project will (1) develop grammar-aligned decoding, a suite of decoding algorithms that more faithfully capture the LLM's underlying distribution than existing grammar-constrained decoding methods; (2) adapt program analysis and synthesis techniques to encode a variety of formal specifications as grammars that grammar-aligned decoding can handle; and (3) develop interactive techniques that help users formalize their intent as specifications. 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 2024 · 2024-10

Machine learning is deployed across various domains (e.g., finance, education, hiring) with the assumption that model outcomes are accurate and authoritative. But in reality, the specific model that is deployed is just one option of many: previous work has shown that multiplicity – the existence of multiple equally good models – arises at many stages of the machine learning pipeline. Formally reasoning about multiplicity is challenging due to the large (potentially infinite) set of models one has to take into account. As such, existing techniques are currently only able to reason about certain forms of model-based multiplicity, and generally only with empirical guarantees. This project’s novelties are a set of approaches that increase the auditability of machine learning pipelines. These techniques consist of frameworks and formal techniques to understand how multiplicity in the dataset creation and modeling processes impacts the final learned model that is deployed. The project’s impacts are especially prominent in domains where the decisions of machine learned models directly affect humans --- understanding multiplicity is vital for developing machine learning models that are fair and robust. The investigators are involved with organizing outreach programs to expose high schoolers and undergraduates to computer science and topics in machine learning. This project investigates multiplicity for diverse model architectures across the whole machine learning pipeline including training data, model predictions, and model explanations. The research integrates formal methods and robust machine learning techniques to provide techniques to help answer the question of whether machine learning outcomes are reliable, or whether they are just an artifact of multiplicity. For instance, the investigators study algorithms to certify (deterministically or probabilistically, depending on the model architecture) whether a model’s prediction is robust under various sources of multiplicity. 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 2024 · 2024-10

Advanced manufacturing strategies for functional analytical biomaterials are needed immediately to help accelerate tissue model development and enable high-throughput screening platforms for biomedical and pharmaceutical technologies. For example, one of the primary factors that halts the advancement of new therapeutic drugs are their adverse reactions in organs such as the heart. To help accelerate drug development and de-risk the progression towards clinical studies, better cardiac tissue models are required that not only mimic mature human heart tissue but also have integrated analytical tools that can rapidly assess the mechanical behavior of the tissue with high spatial resolution and sensitivity in response to early developmental drugs. Because of the material tunability, speed of fabrication, and biocompatibility, three-dimensional (3D) optical bioprinting has become an ideal additive manufacturing strategy to achieve realistic tissue models; however, very little advancement has been made on the printing of high-resolution biomechanical sensors directly into fabricated tissue constructs. This Boosting Research Ideas for Transformative and Equitable Advances in Engineering (BRITE) project spearheads this advancement by supporting research that intends to develop additive manufacturing processes that can rapidly print, and/or erase, biomaterials capable of high density wireless mechanical sensing. Through a multi-faceted diversity, equity, and inclusion plan, this project will also boost underrepresented minority student retention and degree attainment in science, technology, engineering, and mathematics (STEM) fields by building a STEM awareness program at community colleges and identifying opportunity/equity gaps in rigorous STEM curricula. Technically, this project aims to conduct research focused on engineering photolabile ferroelectric bioinks that incorporate piezoelectric nanoparticles modified with electrochromic dyes within a hydrogel that can be used for additive and/or subtractive manufacturing in 3D stereolithography instruments. These bioinks will be used to 3D print nano-ferroelectric biomaterials with < 5 micro spatial resolution that have tunable mechanical properties, high piezoelectric coefficients, and can directly couple piezoelectric and optical signals. Through surface engineering and modeling, a strong understanding of how to control and optimize the polarization-strain relationship in optically printed ferroelectric nanocomposites will be generated. In addition, fundamental questions will be answered on how local piezoelectric strains fields can be engineered to induce molecular Stark effects and how cleavable chemical bonds can be leveraged to rapidly deconstruct and erase polymer nanocomposites using light. Lastly, this project will demonstrate that functional analytical biomaterials can be optically manufactured from the developed bioinks and 3D bioprinting instrumentation that are capable of high-density force read-outs and sensitivity via a piezoelectric-to-optical transduction mechanism. 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 2024 · 2024-10

This Smart and Connected Health (SCH) award will focus on creating a robotic system for diagnosis of abnormal tissues inside the abdominal cavity. Diseased abdominal organs often present a complex mixture of normal, abnormal but non-cancerous, and cancerous tissues. Existing medical imaging methods fail to offer useful diagnosis due to errors caused by breathing and the limits of imaging resolution and sensitivity. Diagnostic laparoscopy along with tissue biopsies can provide more detailed information to guide treatment but are limited due to subjective errors in visual inspection and errors from sampling a small amount of tissue. To solve these problems, this research project will develop a smart robotic system with multiple sensors and artificial intelligence. The robot will move through the abdominal cavity, analyze the size, shape, and chemical information of tissues, and identify abnormal tissues. The research will also include educational and outreach activities to promote STEM fields, especially among groups that are traditionally underrepresented in these areas. The goal of the research is to design, develop, and evaluate a multimodal robotic system equipped with flexible endoscopy, ultrasound imaging, and Raman spectroscopy for comprehensive cancer diagnosis in the abdominal cavity. The project is built upon three research thrusts: 1) developing a multimodal instrument for multiscale tissue diagnosis, 2) developing a mesoscale continuum robot for tissue surface scanning, and 3) developing multimodal fusion for comprehensive diagnosis. The first thrust integrates a balloon-based ultrasonic probe with a Raman spectroscopy needle to detect, classify, and stage tissue on the surface and deep inside organs. The second thrust integrates the sensing modalities with a tendon-driven continuum robot and equips the robot with the ability to scan tissue surface through data-driven modeling and model predictive control. The third thrust combines data from multiple sources to perform tissue identification and staging and builds robust models to handle missing/occluded data and improve overall accuracy. The robotic system and its individual components will be calibrated and demonstrated by performing navigation tasks and collecting data using gelatin, tissue, and abdomen phantoms. The robotic system may not only provide comprehensive diagnosis of heterogeneous and unstructured tissue environments but also improve the safety and accuracy of surgery through intra-operative diagnosis. This project will generate new knowledge and methods in biomechanics and mechanobiology by revealing multiscale tissue information and potentially identifying new biomarkers critical to cancer treatment. 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.