University Of Southern California

NIH Research Projects · FY 2024 · 2024-09

PROJECT SUMMARY According to the Centers for Disease Control, hearing loss is the third most common chronic physical condition in the US, surpassing diabetes, and cancer in prevalence. Most cases of hearing loss occur due to damage to the cochlear amplifier. The term "cochlear amplifier" refers to the chain of processes that amplify low-level vibra- tions within the cochlea, thereby enhancing the sensitivity and dynamic range of hearing. While many details of the amplifier's mechanisms remain unclear, it is well-established that electromotility, the voltage-activated length changes of cochlear outer hair cells (OHCs), plays a crucial role. The OHC-driven amplification is vulnerable to common factors such as excessive noise, ototoxic drugs, aging, and congenital defects. Various insults can affect distinct stages of the amplification process, inhibiting electromotility or reducing OHC stimulation, leading to sensory hearing loss. However, current diagnostic tests, such as recordings of otoacoustic emissions (OAEs), have limited ability to provide precise information on the site of damage within the cochlea and its functional consequences. My research aims to address this gap by developing a diagnostic test of cochlear amplification. We propose that recordings of cochlear microphonics (CM), which represent the summed electrical fields of stimulated OHCs, can assess the functional state of local cochlear amplification and identify the site of damage within the amplifier. The proposed test combines the place-specific properties of cochlear two-tone suppression with recent findings on intracochlear motions that reveal broad regions of excitation and suppression in OHC vibrations. We hypothesize that the suppression of CM responses is controlled by local gain changes in cochlear motions; hence, we named it gain-sensitive CM (gCM). Thus, the gCM test should be capable of detecting coch- lear regions with dysfunctional OHCs and evaluating the dynamic range of the amplifier at a specific cochlear location. Furthermore, we predict that gCM, when measured in low- vs. high-intensity regimes, will exhibit differ- ential sensitivity to cochlear amplification loss caused solely by dysfunctional electromotility vs. by disruptions in the processes driving it. Currently, no other hearing test can achieve this level of diagnostic precision, which is essential for developing and testing individualized treatment options. In the short term, we aim to validate the gCM in mice by comparing it to direct, albeit invasive, measures of cochlear gain in the organ of Corti vibrations as well as to more established OAE tests. To evaluate the hypothesized site- and place-specificity of gCM, we will use both healthy and hearing-impaired mice where damage is either limited to a specific stage of the ampli- fication process or to a specific cochlear region. The results of this study have the potential to revolutionize the treatment of sensory hearing loss by providing an objective and noninvasive test to identify the nature and loca- tion of cellular damage responsible for amplification loss that outperforms OAE tests. Consequently, this project could enhance the precision of diagnosing sensory hearing loss in humans and facilitate the development and implementation of targeted intervention strategies in both research and clinical settings.

NIH Research Projects · FY 2025 · 2024-09

PROJECT SUMMARY The craniofacial skeleton is composed of multiple tissue types that must be precisely integrated during development to ensure normal structure and function. While the mechanisms orchestrating this process are key to revealing the etiology of congenital disorders with craniofacial differences, our current understanding of the craniofacial skeleton is largely based on studies that focus on bone and cartilage. Conversely, the mechanisms that regulate differentiation and patterning of diverse soft connective tissues, inter-tissue crosstalk, and the formation of the connections between different tissue types in the head remain largely unknown. The scope of the proposed study is comprehensive and forward-thinking, as it investigates the craniofacial skeleton as a multi-tissue system, addressing critical gaps in our understanding of craniofacial connective tissue development and maintenance in four complementary research programs. These programs leverage Dr. Amy Merrill’s expertise craniofacial development to decode the molecular identities of connective tissue cells, reveal their role in joint development and maintenance, identify the signals that bridge connective tissue and bone, and establish novel associations between connective tissue development and human disease. Utilizing both disease-first and gene-first approaches from this distinct perspective, alongside an expert team of human geneticists, this research aims to identify new genes variants that cause human craniofacial disorders. Cutting-edge techniques in spatial transcriptomics and mouse genetics will be used to identify genes and pathways regulating assembly of the craniofacial skeleton, engineer new tools that target and track understudied skeletal tissue types, and establish powerful models of human congenital disorders. The increased flexibility and duration of the R35 grant will facilitate these technological advances, breaking new ground to uncover fundamental programs that regulate skeletal connectivity in craniofacial development and disease. Dr. Merrill is uniquely poised to successfully lead this study at the University of Southern California in the Center for Craniofacial Molecular Biology, a rapidly growing research center with exceptional resources directed by Dr. Yang Chai. She has an outstanding record of research productivity and impact in the field of craniofacial biology and currently holds two R01’s from NIDCR. Dr. Merrill has shown excellence in mentoring, which is reflected by the success of her trainees, recognition with the prestigious USC Mentoring Award in 2022, and role as Co-Investigator on USC’s NIDCR T90 Training Grant for Craniofacial Biology. She has also shown sustained commitment to professional service within the research community, serving in various leadership roles, including President-elect of the Society of Craniofacial Genetics and Developmental Biology. Dr. Merrill’s contributions to research, mentoring, and service position her on an upward trajectory to advance our understanding of craniofacial development, ultimately paving the way for diagnosis and treatment of congenital craniofacial disorders.

NIH Research Projects · FY 2024 · 2024-09

Lewy body dementia (LBD) is the second most common cause of dementia after Alzheimer’s disease. In particular, Parkinson’s Disease (PD) progression has been associated with LBD and cognitive decline in a number of cognitive domains including executive function, attention, processing speed, episodic memory, and visuospatial processing. The predominant clinical motor features of PD are bradykinesia, resting tremor, and muscular rigidity; however, the prevalence and severity of the nonmotor effects of PD have significant detrimental effects on quality of life. While conventional pharmacological and surgical treatments of PD are effective in improving motor symptoms of PD, they do not improve cognitive deficits and may even have long-term deleterious effects on verbal fluency and cognition. Chronic high frequency deep brain stimulation (DBS) in the subthalamic nucleus (STN) and internal segment of the globus pallidus (GPi) is efficacious for improving motor symptoms of PD. Current stimulation parameters are optimized for motor benefit, with frequencies in the high gamma (100-180 Hz) range. Interestingly, increased low frequency oscillations (i.e. theta rhythms (4-8 Hz)) have been implicated in a range of cognitive functions, including spatial and episodic learning and memory. There is growing evidence that low (theta) frequency STN stimulation preferentially improves executive function compared to standard-of-care gamma DBS (cDBS). Indeed, we have generated data in PD patients with STN DBS that indicate “on” theta stimulation improves hippocampal-based verbal fluency compared to “off” or “on” gamma stimulation. Unfortunately, low frequency (theta or beta; 5-30 Hz) stimulation is not beneficial for motor symptoms. However, recent advances in stimulation programming allows for theta burst stimulation, which integrates high frequency (gamma 50-200 Hz; trains of 5-25) bursts of stimulation repeated at theta (5-10 Hz) frequency intervals. This theta burst stimulation increases theta oscillation activity. Moreover, there is evidence that STN theta burst stimulation is not only safe, but also has comparable motor outcomes compared to conventional gamma frequency STN DBS. Overall Goal: In light of our recent findings that theta stimulation improves cognitive function in PD patients with STN DBS, we hypothesize that chronic theta burst stimulation will confer a long-term cognitive benefit while concomitantly maintaining the motor benefits of gamma stimulation. The proposed randomized double-blind phase 2 clinical trial will focus on determining if 1) short-term and 2) chronic theta burst STN stimulation will improve both cognitive and motor measures; and 3) determining if theta-burst DBS and cDBS result in differing acute and chronic functional brain connectivity. The interpretation of data from this research will improve understanding of the acute and chronic effects of theta burst DBS on cognition and motor function. If successful, this study has the potential to develop a novel STN stimulation paradigm to treat both motor and cognitive PD symptoms as well as understand the effects of theta burst DBS compared to gamma DBS on fMRI functional connectivity measures. As this study is aimed at modulating cognitive networks, the ultimate goal is to develop novel stimulation parameters to treat chronic cognitive dysfunction in PD dementia and more broadly Lewy body dementia. Moreover, data collected, and collaborations developed will lay the foundation for a definitive Phase 3 clinical trial utilizing theta-burst stimulation for cognition and motor symptom improvement.

NIH Research Projects · FY 2024 · 2024-09

PROJECT SUMMARY (OVERALL) The University of Southern California (USC) CLIMAte-Related Exposures, Adaptation, and Health Equity (CLIMA) Center's mission is to build a community of transdisciplinary scientists and robust research infrastructure to advance its theme of community-engaged, solution-oriented climate change (CC), adaptation, and health research. The goal is to inform climate action policies for health equity that strengthen local adaptive capacity, reduce vulnerability, and increase resilience across the life course. A Methods Development Research Core (MDRC) will develop innovative, highly spatiotemporally resolved models of exposure to urban heat islands, wildfire smoke, and increasingly frequent, concurrently, or consecutively occurring compound climate events and investigate their overlap with neighborhood adaptation strategies (e.g., air conditioning use, tree canopy shade, greening interventions) and vulnerabilities (power outages) to identify priorities for increasing CC resilience. These exposure and adaptation measures will be linked to large and diverse electronic medical record (EMR) Data Lakes from USC-affiliated hospital systems, with novel geo-enrichment with neighborhood social and environmental determinants of health in Research Project 1 (RP1), and with lifetime residential histories of young adults with detailed cardiovascular health assessment starting from childhood in Research Project 2 (RP2). RP1 will assess heat extremes and wildfire influenced particulate matter air pollution impact on acute heart failure hospitalization and rehospitalization risk in adults, while informing new directions in biostatistical methods. RP2 will assess the lifetime heat stress and wildfire smoke exposures impact on cardiovascular health measures (blood pressure, pulse rate) and allostatic load as a measure of biological resilience, and as a cardiovascular risk indicator earlier in life. The two RP's will investigate how these effects differ by social vulnerability and adaptation factors (e.g., air conditioning use, power outages, power grid resilience, urban heat islands, tree canopies etc.). The Community Engagement Core (CEC) will facilitate solution-oriented translational research and multi-directional communication with policymakers and the public. It will engage environmental justice communities using an intergenerational approach, including community participatory action research and education to engage youth in the co-design, monitoring, and ground-truthing of adaptation strategies. The CLIMA Center will create strong transdisciplinary research teams, capacity, and culture urgently needed to assess the complex, cascading impacts of CC hazards and exposures on health equity and adaptive capacity over the life course, starting with cardiovascular health and disease as a sentinel public health burden. Given California's diverse communities and array of climate adaptation, mitigation, and health equity policies, it provides an opportunity to interrogate real-life effectiveness, co-benefits, and gaps of policies already in place; inform policies nationwide; and use CLIMA results to inform how to protect the most vulnerable and strengthen CC resiliency.

NIH Research Projects · FY 2025 · 2024-09

Abstract An estimated 43 percent of children under age 5 in low- and middle-income countries (LMICs) will not reach their full developmental potential due to poverty, stunting, or inadequate psychosocial stimulation. Parenting interventions that coach parents on responsive caregiving can effectively improve ECD outcomes in LMICs, at least in the short-term. However, the very few programs that have examined their sustained effects find that early program impacts tend to fade over time. To date, the only parenting interventions demonstrating sustained impacts five or more years after the end of their programs from a LMIC are two small pilot studies, both featuring weekly individual home visits over two years on a total of 134 stunted or low birthweight children from urban Kingston, Jamaica. Such intensive delivery models are prohibitively expensive to scale in rural and resource-poor LMIC settings. Community-based group meetings are a more cost-effective and scalable delivery model, and can also allowforpeer-to-peer learning and the formation of social support networks. Though growing evidence from LMICs shows that group meetings are at least as effective as individual home visits to improve ECD in the short- term, evidence on their ability to sustain impacts over time, and the mechanisms underlying such impacts (e.g., social support), is still very limited. With NICHD support (R01HD090045), we demonstrated that an 8-month, group-based ECD parenting intervention delivered by CHWs in rural Kenya significantly improved short-term ECD outcomes and parenting practices, and the program was highly cost-effective. In complementary work (R21HD098508), immediately after the main intervention, in half of intervention villages, we added 9 bi-monthly “booster” group meetings to reinforce key messages over two years. In a recent two-year follow-up assessment, we found that early impacts from the original 8-month intervention were sustained when children were ages 3.5 to 5, and the poorest families benefited the most. The booster extension, despite being severely disrupted by the COVID- 19 pandemic, had small additive effects on children’s socioemotional outcomes and parenting behaviors. Building on these results, we now propose to follow-up our original study cohort roughly 5 years after the end of the intervention to measure longer-term sustained impacts on our sample of targeted children who will be roughly 6.5 to 8 years old, as well as their younger and older siblings to uncover potential spillover effects. We will re-enroll households from our original sample across 60 villages to collect an additional survey wave to measure children’s cognitive, language, executive function and socio-emotional skills, as well as parenting behaviors, knowledge and beliefs, social networks, and social norms around parenting to examine potential mediating pathways driving any sustained impacts and any spillover impacts onto siblings. The goal of our study is to help fill a gap regarding evidence on the long-term effectiveness of ECD programs on children’s outcomes from a scalable and cost-effective model of delivery to help inform policy.

NIH Research Projects · FY 2024 · 2024-09

PROJECT SUMMARY/ABSTRACT 4D transthoracic echocardiography (TTE) offers an excellent means to study complex heart anatomy and measure cardiac dynamics via myocardial strain imaging. While TTE-based 4D strain imaging overcomes many shortcomings of traditional strain analysis, and modern speckle-tracking algorithms enhanced reproducibility, 4D echocardiography currently lacks the necessary spatial and temporal resolution for comprehensive flow analysis. Likewise, modern systems do not still allow the user to adjust the number and location of the acquired planes, which is crucial for assessing heart valves and cardiac fluid dynamics enabled by echocardiographic particle image velocimetry (Echo-PIV). There is an unmet clinical need to study 4D flow fields in heart chambers, especially in the right ventricle (RV) via echocardiography, which is ubiquitous, real-time, and less expensive than cardiac MRI (CMR). The significance of 4D flow information in the RV and how it alters clinical diagnosis and therapy has been recently illuminated. To date, no breakthrough in transducer technology has yet evolved to offer selective control over 3D spatial information acquired about tissue strain and complex flow. Our team’s multidisciplinary expertise in hardware, software, and clinical ultrasound supports our overarching goal to develop a novel generation of broadband echocardiography transducers, which uniquely utilize only the elements around a matrix array periphery. Combined with multi-line transmit focusing and coherence-based receive beamforming methods, they will provide volume rates substantially higher than existing systems. Not available on commercial systems, the combined advantages of broadband capability, high-speed volume acquisition, and user-adjustable multiplanar acquisition will enable rapid 4D strain imaging and enhance detection, tracking, and visualization of microbubbles for 4D flow measurements. Combining both labs’ complementary expertise will devise a novel generation of 4D TTE ultrasound systems using matrix arrays with a beamformer to enable operator-controlled multiplanar acquisition at high volume-rates. SPECIFIC AIM 1. Develop a novel generation of 4D TTE probes with broadband multi-row/multi-column boundary array that will enable high-speed adjustable multi-planar acquisition. SPECIFIC AIM 2. Devise multi-line transmit-focusing and coherence-based receive beamforming methods. SPECIFIC AIM 3. Validate 4D TTE system in vitro and clinically by enabling MP-Echo-PIV and multiplanar strain imaging.

NIH Research Projects · FY 2025 · 2024-09

Project Summary: The regeneration of blood and immune cells relies on hematopoietic stem and progenitor cells (HSPCs) that reside in the bone marrow. HSPCs not only sustain homeostasis of the blood pool by constantly producing the precise amounts and types of blood cells as needed, but they also respond rapidly to injuries such as bleeding or infection. During these processes, HSPCs constantly sense and react to signals from various bone marrow cell types. Their spatial proximity to the sources of these signals change as they migrate during differentiation and proliferation within the tightly packed bone marrow. The spatial organization of individual HSPCs forms a dynamic landscape that modulates their intercellular communication. In the proposed project, we will study the dynamic spatial configuration of HSPCs during hematopoiesis and its impact on intercellular signaling, proliferation, differentiation, and injury response. The central hypothesis of this proposal is that cell fate transitions during hematopoiesis are associated with the migration of HSPCs through distinct bone marrow micro-environments where they are exposed to signals that promote their expansion and differentiation. We propose to use MEMOIR and seqFISH technologies to (1) map the spatial organization of HSPCs in the bone marrow, (2) investigate the dynamic changes of HSPCs’ spatial positioning during hematopoiesis and upon the demand of a specific blood cell type, and (3) determine the impact of HSPCs’ spatial context on their intercellular signaling and fate choices. Our findings will introduce a novel, spatial perspective of blood and immune cell regeneration and significantly impact many biomedical fields including immunology, hematology, tissue engineering, aging, and cancer. Moreover, this work will provide an experimental and conceptual framework for analyzing spatially defined intercellular communication in other tissue contexts.

NSF Awards · FY 2024 · 2024-09

Reinforcement Learning (RL) is a powerful Artificial Intelligence technique that enables machines to teach themselves optimal decision making by repeatedly learning from the consequences of their actions. RL has demonstrated impressive gains in building autonomous agents in a variety of scientific and engineering domains such as autonomous driving, robotics, and game playing to name a few. However, training these agents is extremely time consuming. Moreover, existing research on reducing the execution time of RL is inaccessible to RL application developers as they require expertise in designing careful compute orchestration among different types of hardware devices (such as Graphics Processing Units (GPU), Field Programmable Gate Arrays (FPGA)) available in modern data centers. To address this issue, the objective of this project is to develop an easy-to-use library that can enable automatic deployment of RL applications on a data center composed of GPU, FPGA, and AI accelerators. By abstracting away the complexities of deployment and orchestration of computations on data centers, the project will significantly increase the productivity of AI application developers leading to more robust AI agents with faster development cycles. Existing RL libraries only consider homogeneous platforms that are composed of a single type of hardware device. Thus, there is a demonstrated need from researchers in the Computer & Information Science & Engineering (CISE) communities, including AI system development, RL algorithm, and domain application user communities (e.g., scientific computing and cyber-physical systems) for a library that can enable seamless deployment of RL applications on emerging heterogeneous platforms (composed of multiple types of hardware devices) while achieving high performance (for example, reduced training time). To address this need, this project will leverage novel algorithmic, architectural, and memory optimizations across heterogeneous devices to create a performance portable library that will enable automatic system composition for high-throughput RL on heterogeneous cyber infrastructures. The library will build upon and harden the research artifacts developed in the recent NSF-funded work of the investigators on accelerating RL on CPU-FPGA platforms. It will be portable to various heterogeneous cyber infrastructures and support a wide range of RL hyperparameters, algorithms, and policy models. Furthermore, it will offer APIs to facilitate productive development and seamless integration with existing RL ecosystems. Additionally, the project will include the following community interactions and sustainability plans: 1. Interactions with key processor, GPU, and FPGA vendors (AMD, Intel and NVIDIA) to integrate the proposed library into their software development tools (AMD-Xilinx Vitis, Intel oneAPI and NVIDIA CUDA-X). 2. Collaborations with NSF Open Cloud Testbed, and NSF NCSA to integrate the library into their cyber infrastructures. 3. Making the library compatible with existing RL frameworks (e.g., RLlib) and RL simulation toolkits (e.g., Gymnasium). 4. Ensuring the availability of the library to a broader audience by collaborating with commercial cloud service providers such as Microsoft and Amazon. 5. Demonstrating end to end applications in various domains through collaborations with NSF AI Institutes including ACTION, AgAID, and ICICLE. 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 · 2024-09

PROJECT SUMMARY Alcohol use disorder (AUD) and overweight/obesity (OOB) are highly prevalent conditions that contribute substantially to global disease burden. Alongside recent increases in the prevalence of AUD and alcohol- attributable mortality, projections indicate that up to one half of U.S. adults will live with obesity by the end of this decade. Because each condition confers increased risk for numerous chronic health conditions, those with co-occurring alcohol use disorder and overweight/obesity (AUD-OOB) are at increased risk for negative health outcomes. The emergence of highly effective incretin-based therapies for diabetes and OOB, including glucagon-like-peptide 1 (GLP-1) receptor agonists, is rapidly expanding the range of effective treatments for those with OOB and cardiometabolic disorders. Moreover, preclinical findings and clinical observations suggest that GLP-1 receptor agonists can reduce alcohol consumption. Tirzepatide, the first dual GLP-1 and GIP (glucose-dependent insulinotropic polypeptide) receptor agonist, shows superior efficacy for weight loss compared to the most effective GLP-1 receptor mono-agonists. Tirzepatide (Zepbound) received FDA approval for the treatment of overweight/obesity in November 2023. Recent findings further show that tirzepatide has protective effects on cardiovascular risk outcomes and projected risk of cardiovascular disease. Based on epidemiological trends and the high prevalence of metabolic conditions in those with alcohol use disorder, an increasing number of heavy drinkers are likely to receive treatment with tirzepatide. Should tirzepatide prove effective for reducing alcohol consumption, the identification of a treatment for co-occurring alcohol use disorder and overweight/obesity could achieve significant long-term health impact. The aim of this study is to expedite clinical research on dual GLP-1/GIP receptor agonists in participants with AUD-OOB by conducting a Phase II randomized trial of tirzepatide. Adults with AUD-OOB will be recruited for a double-blind, between- subjects trial using dose-escalating treatment with tirzepatide (Zepbound) over 8 weeks. Prospective changes in alcohol use, weight, and cardiometabolic outcomes will be measured at weekly clinic visits. By expediting clinical research on tirzepatide, a new treatment poised for widespread clinical use, this project will position the field for larger clinical trials of dual GLP-1/GIP receptor agonists as potential treatments for co-occurring AUD and OOB.

NIH Research Projects · FY 2025 · 2024-09

Project Summary The goal of this project is to assess the role of cost-associated barriers to new and future novel AD/ADRD therapeutics, with a particular focus on older adults with high AD/ADRD burden. Our work in the first (planning) phase of the project will include semi-structured interviews with a cohort of Latino adults at high risk for AD/ADRD that was created as part of the team’s ongoing work on representation in clinical trials. We will also interview an existing group of Latino patients with AD/ADRD and their caregivers. The interviews will inform a subsequent population-based survey to be fielded in the Understanding America Study, an online panel study housed at the University of Southern California. Both the interviews and the survey will elucidate cost and non-cost barriers to the use of new novel dementia therapeutics. The survey will more fully and systematically capture the contribution of environmental, sociocultural, behavioral, and biological factors to the demand for and use of these new medications for all older adults in the US. The second, implementation stage of our work will analyze existing Medicare claims and Encounter data to understand real world use patterns. We will then use the team’s well-established dynamic microsimulation to estimate the downstream impacts of use on cognitive and physical health, costs, and quality of life of older adults with dementia. The dynamic microsimulation model will be informed by the work conducted in all other parts of the project both in terms of parameter inputs and simulated scenarios.

NSF Awards · FY 2024 · 2024-09

Subsurface flow and transport systems within the Earth experience complex interactions among rocks, fractures, and fluids. Subsurface system dynamics affect groundwater aquifers, geothermal energy, hydrocarbon resources, geologic carbon storage, and other important Earth resource management needs. A significant challenge in predicting the behavior of these subsurface systems arises from the need to specify highly uncertain rock physical properties, which are inherently heterogeneous and exhibit variability across multiple scales. Advances in artificial intelligence (AI) offer unprecedented capabilities for processing large and diverse multimodal datasets. In this project, advanced AI models will be developed to capture salient spatial and temporal relations in subsurface flow and transport systems by integrating physical principles and existing multiphysics data. This work has immediate impact on significant societal geoscience applications, such as groundwater aquifers and geothermal energy recovery, as well as climate change and environmental sustainability. The project will attract and educate students underrepresented in STEM fields and equip the next generation of the geoscience workforce with AI skills. The objective of this project is to develop novel domain-aware, robust, and interpretable AI solutions for capturing and predicting subsurface flow and transport dynamics. Specifically, Physics-Informed Causal Deep Learning Models (PINCER), will be developed to detect and exploit spatial and temporal relations in subsurface flow and transport systems by integrating physical principles and multiphysics data. This research is focused on developing: (1) novel deep learning architectures that honor the general structure of fluid flow equations while accounting for uncertainties and allowing for learning and adaptability based on observed monitoring measurements; (2) physics-informed causal deep learning models for succinctly capturing and predicting fluid flow and transport dynamics; and (3) flexible deep learning-based inference of heterogeneous rock flow properties from incomplete multiphysics monitoring data. PINCER presents a paradigm shift from traditional data-driven approaches or model-based techniques to a hybrid solution that combines the benefits of both methods. It makes novel contributions in several AI research areas, such as causal analysis, physics-informed AI, and interpretable AI. It also advances geoscience research by developing more efficient and robust modeling and prediction of fluid flow and transport processes in subsurface environments. The PIs will broaden the impact of their work by training students, disseminating research findings (including new datasets and open-source software), and developing outreach programs. 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 · 2024-09

There are uncountably many possible shapes in the world, and computers cannot store, represent, and display them all. Instead, one typically discretizes a shape. This means that a shape is constructed of many smaller, simpler sub-shapes, for example, triangles. A computer can easily store a triangle by storing the locations of each of its corners; a large complex shape can then be represented as a collection of triangles. Traditional discretization approaches like this—named discretization because they turn a shape into a discrete collection of triangles—are powerful in their simplicity, but they have critical drawbacks. Finding a discretization of a shape in terms of triangles is a difficult and computationally-intensive process, and it is easy to accidentally create an invalid collection of triangles (for example, because it has holes) that is invalid for computational use. Moreover, such a discretization will always be a mere approximation of a true shape, and computations performed on these discretizations can suffer from artifacts introduced by the discretization process. Finally, discretizations are often hard to use in modern machine learning applications based on neural networks, because the discretization process is hard to differentiate, an integral step of training a neural network. This project will overcome these problems by developing discretization-free methods for processing shapes on computers. New methods for the animation of computer graphics characters will be developed that circumvent the traditional step of discretizing the interior of a character before an animation can be computed. The project will also develop discretization-free interpolation methods—when information is given at certain points on a shape (for example, climate readings on isolated weather stations), these methods will be able to interpolate this data over an entire shape for visualization and computation purposes. Lastly, the project will develop discretization-free representations of vector fields, which model data such as hair on a character, wind on the surface of the planet, or electric fields. The primary outcome of this research will be the development of discretization-free methods that will enable smart geometry methods of the future. Furthermore, these awards will fund the education of graduate students at the Massachusetts Institute of Technology and the University of Southern California. A broad variety of mathematical, engineering, and application-oriented challenges will be tackled in the course of carrying out this research. In particular, design of robust algorithms for geometry processing requires solution of partial differential equations (PDEs) as well as PDE-constrained optimization problems on curved domains, with nonlinear objective terms and constraints coupling together multiple unknown functions. The key hypothesis in this work is that neural function representations are well-suited to geometry processing applications, since they are smooth, capable of representing a broad variety of functions, easily differentiable, and compatible with modern machine learning representations, but they will need to be tailored to the needs of this application by making them conform to input geometries and constraints of geometry processing problems. To accomplish this broad goal, the project is divided into three thrusts reflecting applications described in the previous paragraph. As a model problem for animation problems, custom fields will be used to optimize for skinning weights on volumes, a key computational challenge in pipelines for 3D deformation. Extending to cage-based animation, more complex constraints will then be added to the neural fields for geometry processing by considering the problem of optimizing for generalized barycentric coordinates, whose reproduction property is not well-captured by standard machine learning architectures. Finally, non-scalar problems in geometry processing such as frame field design and geometric flows will be considered for which conventional mesh-based algorithms are numerically stiff. Each thrust of the project centers around practical open problems in computer graphics. 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 · 2024-09

ABSTRACT People with serious mental illness (SMI) experience dramatically worse health and longevity than the general population, which is exacerbated by extremely high rates of loneliness and social isolation. Both social isolation (lack of social supports) and loneliness (the subjective experience of isolation) are highly predictive of early mortality. Although social isolation and loneliness are inherently relational, influenced by contextual factors, and prevalent among people with SMI, very little is known about how to design community mental health environments that foster social connection (“social architecture”). Thus, the goal of this K01 is to lay the foundation for an innovative, impactful research career focused on reducing social isolation among people with SMI through environmental design. My research plan employs a rigorous, mixed-method design to: 1) identify environmental features associated with two behavioral mechanisms of action: activity engagement and social interaction; 2) examine associations between activity engagement, social interaction, and self-reported loneliness and support; and 3) feasibility test the co-design and implementation of an environmental modification to support social connection. To do this, I will first conduct socio-spatial observations of activity engagement and social interaction in four mental health Clubhouses in Hawaii. These are community-based psychosocial rehabilitation centers that are highly interested in reducing loneliness among their members. Observation data will be paired with survey data (N=150) to examine associations between observed activity engagement and social interaction and self-reported loneliness and social support. Second, spatial observations and social network data will be visualized to refine a conceptual model interlinking engagement and social interaction. These visualizations will be shared with participating Clubhouses to contextualize the identified patterns in engagement and social interaction and to co-identify potentially modifiable environmental features associated with them. Third, an environmental intervention co-design process will be conducted in one Clubhouse and the intervention will be implemented. Feasibility data on the co-design process and modification of the environmental intervention protocol will be collected through field notes and a final focus group and subsequently analyzed qualitatively. This research will be carried out with close mentorship from a team of highly accomplished senior scholars: Drs. Henwood, Wenzel, Salzer, Valente, Wilson, and Stark. Their work is directly aligned with my career and training goals, which focus on three intersecting areas of expertise: 1) socio-spatial methods (GIS, social network analysis and ecological momentary assessment), 2) theory linking environmental design, human behaviors, and health, and 3) environmental intervention co-design based on community-based participatory research principles. This K01 builds on my unique clinical and theoretical background as an occupational therapist and community psychologist and lays the foundation for an innovative, highly impactful career focused on reducing social isolation and loneliness among people with SMI.

NIH Research Projects · FY 2025 · 2024-09

PROJECT SUMMARY / ABSTRACT Over 50% of youths with type 2 diabetes (Y-T2D) develop diabetic kidney disease (DKD) during adolescence, increasing the risk of early dialysis and death. Latinx and African American youth are disproportionately affected but under-studied, and current treatment options for T2D in adolescents, projected to dramatically increase over the coming decades, are limited. Thus, it is critically important to identify modifiable risk factors and novel therapeutic approaches for preventing DKD in youth. Our strong preliminary data links per- and polyfluoroalkyl substances (PFAS), a group of ubiquitous artificial chemicals used for more than 60 years in consumer and industrial products, with kidney injury in proximal tubules in Y-T2D, a key component of DKD. PFAS toxicity in proximal tubules may be partly due to organic anion transporters (OATs) localized to these cells; Tubular OATs reabsorb PFAS from urine filtrate, concentrate PFAS in tubular cells, and decrease renal PFAS clearance, potentially amplifying PFAS associated nephrotoxicity. Our overarching biological hypothesis, supported by our preliminary data, is that PFAS increases tubular injury and DKD risk and that decreasing OAT reabsorption in proximal tubules may mitigate this risk. This first of its kind study will examine this hypothesis by building on existing data in two independent, meticulously phenotyped, longitudinal cohorts of youth with T2D of predominantly Latinx or African American race/ethnicity. To identify associations and potential therapeutic targets, we will leverage the NIDDK-funded Renal Hemodynamics, Energetics, and Insulin Resistance in Youth Onset Type 2 Diabetes Study (Renal-HEIR/HEIRitage). This study has existing measures of OAT activity measured using single cell RNA sequencing from kidney biopsies at baseline and gold-standard PAH clearance, a measure of tubular OAT function, in addition to existing outcome data, including tubular injury biomarkers, albuminuria, and hyperfiltration at both visits. To examine associations in a larger cohort with annual visits and longer follow-up duration, we will leverage the NIH/NIDDK funded TODAY/TODAY2 study, the largest longitudinal cohort of youth with T2D with comprehensive measures of kidney function (e.g., annual visits for 15 years) with the same outcomes as Renal-HEIR. In both cohorts, we will assess PFAS exposure and clearance by measuring longitudinal serum PFAS and urine PFAS at baseline; we will assess biomarkers of OAT function in urine and plasma and address reverse causality using pharmacokinetic (PK) models. We will use innovative statistical approaches, including a latent unknown factor analysis to identify how OAT activity associates with changes in PFAS over time, and a novel causal inference and mediation framework based on latent factors to examine whether decreases in OAT activity, induced by sodium/glucose cotransporter-2 inhibitors, decrease the risk of DKD via alterations in renal PFAS clearance. This groundbreaking study will deepen our understanding of the interplay between OATs, PFAS, and DKD in youth with T2D and inform targeted clinical interventions to prevent PFAS associated DKD, particularly in marginalized communities.

NIH Research Projects · FY 2025 · 2024-09

PROJECT SUMMARY. Inorganic nanoparticles (NPs) hold great promise as targeted drug delivery systems but tailoring their pharmacokinetics (PK) to specifically target regions of interest remains a challenge. This limits the clinical translation of NPs due to poor delivery efficiency and safety concerns associated with off-target accumulation of NP-based formulations. Due to the interactions of NPs with biological components, driven by their structural properties, customizing the PK of NPs requires a quantitative understanding of the effect of NP structural properties on their whole-body biodistribution, which in turn also governs their safety profile. Therefore, to enable rational design of inorganic NPs to achieve organ targeting and safety, we propose to leverage artificial intelligence to develop a toxicology-integrated physiologically-based pharmacokinetic model (PBPK-Tox) capable of accurately predicting the whole-body exposure and safety of novel nanomaterials, based solely on their structural properties, dose, and route of administration. For this, we will (1) develop a PBPK-Tox model based on diverse datasets from literature, (2) establish the quantitative relationship between NP properties, exposure, and toxicity, and (3) experimentally test the model predictions of rational design for organ-specific targeting. Our proposed modeling framework will enable efficient preclinical development of inorganic NPs (and accelerate their clinical translation) by providing rational design guidelines through in-depth computational investigation of biological and physicochemical variability on biodistribution and safety of NPs.

NIH Research Projects · FY 2025 · 2024-09

PROJECT SUMMARY Despite global improvements in cardiovascular disease (CVD) rates, CVD morbidity and mortality are increasingly common in young adults aged 18 to 39. CVD research often excludes younger individuals, resulting in prevention strategies that are tailored to older populations, as well as algorithms that underestimate young adult CVD risk and clinical guidelines that systematically undertreat young patients. Scientific consensus indicates that CVD risk begins to accumulate in early life, with vulnerabilities and behaviors in early adulthood determining the lifelong trajectory of cardiovascular health. Therefore, prevention strategies that target novel primordial indicators (i.e., early vulnerabilities emerging prior to traditional CVD risk factors) are urgently needed to prevent or delay CVD progression in young adults. The American Heart Association’s Life’s Essential 8 guidelines emphasize sufficient amounts of physical activity (PA; ≥150 min/week moderate-to-vigorous physical activity [MVPA]) and sleep (≥7 h/night) as core CVD prevention strategies. However, the majority of young adults fail to reach these recommended levels. These guidelines also draw from an evidence base that considers PA and sleep behaviors in isolation from their inherent compositional and circadian characteristics which might also affect cardiovascular health. Other weaknesses in the extant literature—including the use of inaccurate self- report measures and primarily cross-sectional examinations of aggregated PA/sleep on end-stage CVD outcomes—provide an incomplete understanding of the relationship between PA/sleep and CVD. Therefore, this training grant will examine the longitudinal, objectively-measured, within-person effects of compositional 24-h behaviors (light PA [LPA], MVPA, sedentary behavior [SB], sleep) and circadian rhythm on two primordial CVD indicators (stress, fatigue) in a cohort of free-living young adults. Data from N=239 participants in the recently- completed TIME Study will be utilized. 24-h behaviors and circadian rhythm were continuously measured with wrist-worn accelerometers (Fossil Sport smartwatches), and stress and fatigue were collected with daily ecological momentary assessment (EMA) prompts, over the course of one calendar year. We will then apply a series of novel combined multilevel model/isotemporal substitution analyses to examine within-person main effects of 24-h behavior compositions and circadian rhythm metrics on daily stress and fatigue, while controlling for between-person effects. The potential moderating effects of relative prior-day sleep deprivation, inactivity, and sedentariness, and current-day circadian rhythm metrics, will also be examined on these relationships. To our knowledge, this is the first largescale intensive longitudinal study to examine the within-person compositional effects of 24-h behaviors and circadian rhythm on early indicators of CVD risk in young adults. Our findings will help inform the development of primordial prevention efforts and integrated, tailored behavioral guidelines that address how unique individuals might decrease sub-clinical CVD risk in the context of their everyday lives.

NIH Research Projects · FY 2025 · 2024-09

PROJECT SUMMARY Young adulthood is a critical period for alcohol initiation and changes in alcohol use patterns, including increased risk for alcohol misuse and alcohol use disorders (AUD). Among young adults, social contexts are strong determinants of alcohol consumption and alcohol-related consequences. As such, it is critical for research studies to investigate social factors as mechanisms to understand future AUD risk in young adults. Rigorous prior research has shown that an individual’s response to alcohol is greater when they are drinking with another person compared to when they are drinking alone. We refer to this phenomenon as “social facilitation of alcohol effects”. We propose social facilitation of alcohol effects may represent the rewarding aspects of social contexts combined with alcohol-related rewarding effects to produce an additive rewarding experience. Social facilitation of alcohol effects may be relevant to young adults, as this developmental period is marked by lifetime peaks in alcohol use and heightened sensitivity to social factors. In Study 1, participants will complete 4 counterbalanced lab sessions: alcohol (0.0 g/kg vs 0.8 g/kg) × social context (alone vs with friend). During each session, participants will complete measures of affect and subjective alcohol effect prior to and following beverage consumption. During social sessions, participants will also complete a standardized conversation task with behavioral coding. The goal of Study 1 is to extend prior research characterizing the phenomenon of social facilitation alcohol effects to investigate person-level predisposing factors to social facilitation alcohol effects and mechanisms of social facilitation alcohol effects. Based on prior studies demonstrating that extraversion and social alcohol expectancies may associate with greater social reward and positive alcohol effects, we propose these as predisposing factors to social facilitation alcohol effects. Based on prior research demonstrating that alcohol consumed in social settings increases affiliative behavior and emotional contagion (i.e., individual’s affect can alter the affect of others in a coordinated manner), we propose increased agreeableness, emotional responsiveness, and emotional contagion as mechanisms of social facilitation of alcohol effects. The same participants will then complete Study 2, which includes 28-days of Ecological Momentary Assessment (EMA) and 6- and 12-month follow ups. EMA will capture social context and EMA and follow-up will assess naturalistic alcohol use outcomes (i.e., alcohol use, binge alcohol use, alcohol consequences). Study 2 will extend findings by investigating the role of acute social alcohol response (i.e., social facilitation of alcohol effects measured in Study 1) on future alcohol use and consequences. The combined EMA/longitudinal follow-up will allow for granular assessment (including social context) of real-time alcohol use uncontaminated by retrospective recall (EMA) and the ability to measure alcohol use patterns up to one year later (follow-up).

NIH Research Projects · FY 2024 · 2024-09

Project Summary Suicide the second leading cause of death among adolescents ages 14-24; in 2022, approximately 18 adolescents died each day from suicide. Child death review (CDR) teams are one method used across the US and internationally to dive deeply into the preceding circumstances of adolescent death to form recommendations to prevent adolescent mortality. Adolescents spend a considerable amount of time in schools, and school representatives can help a CDR team identify crucial proximal and modifiable risk factors of adolescent suicide death. Shockingly, fewer than 3 in 10 CDR teams have participation from school representatives, representing a major CDR implementation gap and lack of actionable data for suicide prevention in the high-risk developmental period of adolescence. The goal of this exploratory sequential mixed methods study is to explore barriers and facilitators to school participation in CDR team meetings. The team’s location in Los Angeles County is a unique strength because CDR methods were first developed in Los Angeles in 1978, and in 2000 Los Angeles County began the only CDR devoted solely to adolescent suicide death. With substantial connections to community partners to conduct this work, our team is uniquely equipped to complete the proposed exploratory study, which (1) aligns with the developmental purposes of the R21 mechanism and (2) responds directly to the Notice of Special Interest in Mortality of Adolescents, Young Adults, and Other NICHD Priority Populations in the United States (NOT-HD-23-001). We aim to (1) explore facilitators and barriers to school representatives’ CDR participation for adolescent suicide mortality; (2) explore school attorneys’ perspectives about school representatives’ CDR participation for adolescent suicide mortality; and (3) describe school representatives’ experiences with, concerns about, and recommendations for how to engage them in CDR team meetings. We will conduct semi-structured interviews with (Aim 1) samples of school representatives (e.g., school psychologists, social workers, counselors) and to explore emergent themes related to their concerns, administrative processes, and decision-making about participating in a CDR team meeting. For Aim 2, we will interview school attorneys to explore how they advise schools about participating in CDRs, probing about concerns of liability and confidentiality. We will use the interview data from Aims 1 and 2 to conduct a web-based survey (Aim 3) of school representative to describe leading barriers, facilitators, and recommendations for CDR. School participation in CDR reviews is a major undeveloped area of research for improving identification of mechanisms, correlates, and modifiable risk factors of adolescent suicide mortality, and our team is well positioned to develop and test policy and practice interventions to improve CDR participation based on the findings of this study.

NIH Research Projects · FY 2025 · 2024-09

PROJECT SUMMARY/ABSTRACT Autism spectrum disorder (ASD) and attention-deficit/hyperactivity disorder (ADHD) exhibit significant clinical overlap, but the shared and distinct biology of these conditions remains incompletely understood. Both ASD and ADHD have been linked in part to the brain’s reward system, with prior work demonstrating that these conditions are associated with altered function and structure of individual reward structures. However, reward areas do not act in isolation but instead communicate extensively with many higher- and lower-order brain regions, including sensorimotor and cognitive control areas implicated in ASD and ADHD, respectively. To inform our understanding of the convergent and divergent neural mechanisms underlying ASD and ADHD, there is thus a crucial need to understand how the connectivity patterns of reward structures are altered in these conditions. The proposed project will use cutting-edge resting-state functional magnetic resonance imaging (rsfMRI) methods together with refined machine learning and statistical modeling approaches to investigate reward functional connectivity in ASD and ADHD. Analyses will consider both categorical diagnoses (ASD, ADHD, comorbid ASD+ADHD, neurotypical controls) and continuous ASD and ADHD symptoms across the population. Datasets will include the largest available samples in the world for both categorical diagnoses (N>6,000; ages 5-65) and continuous symptoms (N>11,000; ages 9-14). The aims of this project are as follows: Aim 1 will investigate the functional connectivity patterns of reward regions in ASD and ADHD using machine learning and multivariate statistical approaches. Aim 2 will assess the moderating impact of key sources of heterogeneity (sex, puberty, medication) on reward functional connectivity in ASD and ADHD using machine learning. Aim 3 will create normative reference curves of reward functional connectivity using data-driven modeling approaches and examine deviations from typical maturational trajectories in ASD and ADHD. Taken together, this work will substantially improve our understanding of reward circuitry in ASD and ADHD, as well as the shared and distinct neural underpinnings of these conditions. In the long-term, this project will contribute to the development and personalization of novel biologically-grounded treatments for ASD and ADHD. These studies are in line with the NIMH strategic plan to define the brain mechanisms underlying complex behaviors, and to examine mental illness trajectories across the lifespan. Additionally, this proposal will provide the PI with significant training in the following new areas: (1) refined machine learning methods, (2) advanced statistical modeling approaches, and (3) cutting-edge functional connectivity methods. This training will be completed under the mentorship of Drs. Paul Thompson, Jose-Luis Ambite and Vince Calhoun, who are world-renowned experts in their respective fields. As a whole, the proposed research and training will allow the PI to become a productive independent investigator using cutting-edge analytic methods to lead translational research on neurodevelopmental disorders and their heterogeneity to improve precision medicine approaches.

NSF Awards · FY 2024 · 2024-09

To better understand how climate may change in the future, scientists look at how it has varied in the past, a field called paleoclimatology. Doing so requires the use of proxies such as the rings of trees or the chemical composition of ice cores to infer climate variability over thousands to millions of years. Given the importance of these datasets, it is crucial that scientists have the ability to efficiently locate, access, and integrate data. Paleoclimate studies often require the integration of hundreds of datasets that are stored in unstructured files, currently requiring scientists to spend a significant fraction of their time manually reformatting the data before they can work with it. The goal of this project is to use artificial intelligence to identify tables in files automatically so these tables can be used more easily in analyses. The end goal is to make more data available to scientists, to reduce the time scientists spend on data wrangling, to decrease the time it takes to obtain scientific results, and to increase the reproducibility of scientific analyses performed by different research groups. Working with collaborators from the hydroclimate community, the project contributes to the understanding of how water resources have changed over the last two thousand years. The Table Understanding for Paleoclimate Studies project aims to leverage table understanding machine learning mechanisms to greatly reduce the time scientists spend searching, wrangling, and annotating paleoclimate datasets prior to analysis. The resulting deep learning model will be integrated into a Python toolbox that will allow scientists to identify and retrieve tables from unstructured files into their computational environment along with relevant metadata. A primary use case for this project is to extract tabular paleoclimate data from the National Oceanic and Atmospheric Administration World Data Service for Paleoclimatology and the PANGAEA repositories. Over the past decade, these archived centers have spent a significant effort standardizing and updating the metadata of the datasets in their possession, but many datasets remain difficult to extract programmatically. In addition, this project builds on an existing recommender system for paleoclimate datasets that will be updated with community-curated datasets to help with metadata annotation and correction. TUPS will directly engage with the paleoclimate community through tutorials, hackathons, and a workshop to ensure that tools and training materials meet the community scientific needs and to help train the next generation of paleoclimate scientists. This award by the Office of Advanced Cyberinfrastructure is jointly supported by the National Discovery Cloud for Climate initiative within the Directorate for Computer and Information Science and Engineering and by Geosciences Directorate’s Research, Innovation, Synergies, and Education and Atmospheric and Geospace Sciences divisions. 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 · 2024-09

Mathematics plays a critical role in students’ career choices as well as their success in STEM fields. Unfortunately, math performance often declines as students’ progress through the grade levels and the aftermath of the COVID19 pandemic has worsened the situation. Research has shown that when teachers have strong content and pedagogical content knowledge that they can provide better quality mathematics instruction to their students and improve student outcomes. The goal of this project is to enhance elementary school teachers’ capacity to improve students’ mathematics learning through a scaled professional development program that uses artificial intelligence (AI) to create a personalized, active learning environment for teachers. The professional development program focuses on key elements of content-specific expertise needed for teaching, such as enhancing teachers’ understanding of the foundational ideas behind numbers and operations concepts that are developed across grade levels, as well as teachers’ understanding of how students learn these concepts and how various instructional tools and practices can improve students’ learning. By creating a professional development that is adaptive to the individualized needs of teachers and is accessible to teachers anywhere and at any time, this work has the potential to change student outcomes at scale. The professional development program utilizes advances in AI to create an inquiry-based learning environment for teachers to enhance their understanding through solving problems of practice. Equipping teachers with the knowledge and skills crucial for quality teaching has the potential to improve mathematics teaching and learning at scale, which has the potential to reduce the opportunity gaps to quality teaching faced by underserved students. The specific focus of this project is to enhance elementary school teachers’ content and pedagogical content knowledge of numbers and operations using a multiple-AI-agent to guide teachers’ development of a conceptual understanding of the content as well as ways to make the content more accessible to their students. Rather than AI delivering the information, the AI tool will serve as a facilitator to create a learning environment in which teachers meaningfully engage with purposefully developed activities and learn through the process. The research questions the work addresses are: (1) In what ways can advances in AI be incorporated into the AI-based interactive professional development program? (2) How well does the AI-based professional development program enhance teachers’ content and pedagogical content knowledge of numbers and operations? and (3) How well does the AI-based professional development program enhance the quality of mathematics teaching and students learning of numbers and operation? The professional development program and AI tool will be developed through multiple iterations and inputs from several key stakeholders, such as teachers, teacher educators, and content experts. The study will use a mixed methods approach. The effectiveness of the fully developed program on instruction and student learning will be explored through a randomized controlled trial with 200 elementary school teachers. The final version of the program will be made available online. The Discovery Research preK-12 program (DRK-12) seeks to significantly enhance the learning and teaching of science, technology, engineering, and mathematics (STEM) by preK-12 students and teachers, through research and development of innovative resources, models, and tools. Projects in the DRK-12 program build on fundamental research in STEM education and prior research and development efforts that provide theoretical and empirical justification for proposed projects. 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 · 2024-09

In this EArly-concept Grant for Exploratory Research (EAGER) project, artificial intelligence (AI) methods that can learn from, and make predictions on, simulations of the physics of materials will be developed. The approach in this project will constitute an extension of the capabilities of recent AI platforms, of which OpenAI's ChatGPT, Microsoft's Copilot, and Google's Gemini, are among the best known. These AI platforms have caught the public's imagination and are widely used in virtually every field of human endeavor. However, their use in science and engineering is mostly based on the AI learning from vast volumes of scientific literature in the form of text and drawing conclusions through language prediction via underlying neural network-based large language models. Going beyond this, we will develop AI methods for such large language models to learn from both simulations and mathematical equations in an expansion of the way they learn from text. This capability will make it possible for our AI platform to learn complex processes in the physics of materials and make predictions that are too intricate to be easily attained by human experts. In particular, the focus of this project will be on the physics of battery materials. Thus, in addition to advancing the frontiers of AI, this project will make important contributions to the design of future batteries for sustainable and safe energy generation. Allowing AI to learn jointly from simulations and mathematics will be a significant departure from previous text-based learning in large language models in science, and has not been demonstrated yet. This project will educate students from diverse backgrounds in developing AI. In addition to ensuring equitable access, this is of considerable importance because diverse and inclusive human input will help mitigate some of the effects of biases in AI. Furthermore, close attention will be paid to constant testing to avoid harmful output from the AI. Coupled electro-chemo-mechanics in materials physics lead to emergent phenomena including phase transitions and instabilities. For materials discovery and design, it is of interest to not only solve forward problems, but also to explain what type of model best represents the observations. Such inverse problems encompass the task of inference, where the goal is to identify the mechanisms of the coupled physics that best explain data that, typically, displays time evolution. Some progress has been made in inference with applications to materials physics of batteries, bio- materials, and structural alloys, but there remains a gap. Existing methods of inference in physics select the best models from a library of candidates. While useful to explain phenomena with data- driven models, such inference does not lead to discovery of previously unknown physics. This EAGER project is to develop Generative Artificial Intelligence methods, specifically Large Language Foundation Models, that will be augmented to perform true discovery of emergent phenomena in mechanics-driven coupled physics problems. Specifically, leveraging experience gained in prior work carried out by the PIs on pre-training and fine-tuning large language models, a new direction will be sought out for multi-modal foundation models that learn directly from computational physics simulations and mathematical equations. The project has broad implications, moving toward true discovery of nonlinear mechanisms in systems with complexity that are induced by coupling with other physics. Applications of this include energy materials such as batteries, biomaterials, and other materials classes such as structural alloys. This new modality of large language models generalized to learn from simulations and mathematics is novel. It represents a significant departure from previous text learning-based uses of large language models in science, and has not been demonstrated yet. Its success will draw from the investigators’ prior experiences with autoregressive, attention-based foundation models, and their understanding of how to extend them to learning from time series and spatially related data. 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 · 2024-09

Reliable concentration measurements of chemical elements require careful quality control. Part of this quality control depends on the use of reference materials of known concentration to assess the accuracy of measurements and intercalibrate between laboratories. For trace element concentration analyses of seawater, the supply of reference materials has become exhausted in recent years. This project aims to collect large samples of water from the Pacific Ocean and Gulf of Mexico to be used as a new set of reference materials. Trace metal concentrations will be measured by several experts using different methods and conditions. The concentration values will be compared against one another, and a consensus value will be developed by these experts using state-of-the-art statistics. The remainder of these large-volume water sample will be stored for distribution to the scientific community in the future. These consensus materials are expected to improve the quality and accuracy of seawater trace metal concentration data over the next decade. Several students will participate in the three research cruises and receive training on trace metal sampling and intercalibration studies. The project also provides support for an early career scientist. The primary objective of this project is to improve the quality of trace metal data in the ocean over the next decade through optimized methodologies and well-studied consensus samples that can be used to assess accuracy. Samples will be collected from the surface and 1000 m at Station ALOHA, which brackets common ranges of open ocean dissolved metal concentrations. In addition, two surface stations in the Gulf of Mexico that have lower salinities and higher organic content due to the influence of the Mississippi River outflow will be used to test the boundaries of sample storage and analytical intercalibration. Samples will be collected and homogenized within large volume tanks and dispensed into 500mL bottles for archiving and distribution to the community. These samples will be analyzed initially by fifteen laboratories worldwide for a suite of trace metal concentrations, to address two goals. First, the data will be compared statistically, via a collaboration with an expert statistician from the National Institutes of Standards and Technology, to calculate consensus concentration values for each element. This consensus concentrations will be reported on the GEOTRACES website, and the statistical best practices will be published. Second, this network of collaborators will explore common intercalibration issues that have arisen over the last decade by utilizing a range of analytical methodologies and conditions. The primary scientific impact of this project will be a set of well-studied consensus samples that can be used to monitor the accuracy of trace metal analyses of seawater over the next decade. This project will support one early-career scientist, one PhD student from the University of Southern California, and 5-15 graduate and/or undergraduate students from Texas A&M University who will participate in the staging and collection of samples from these expeditions. 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 2024 · 2024-09

Project Summary Lipid accumulation, inflammation and metabolic dysregulation are hallmarks of liver disease. Nonalcoholic fatty liver disease(NAFLD) has recently emerged as one of the leading global etiologies of liver disease as it affects approximately 25% of the global population. Nonalcoholic steatohepatitis (NASH) is the inflammatory subtype of NAFLD, and approximately 40% of patients will progress to this stage; with another subset proceeding towards carcinogenesis. Unfortunately, many of these patients will go undetected as they progress to hepatocellular carcinoma which has a median survival of just 11 months. It is thus crucial to better characterize the interface between NAFLD and NASH to best improve patient outcomes. To better understand the molecular signals driving this transition we have targeted the insulin signaling pathway where we have established a liver specific deletion model of PTEN, the negative regulator of the insulin signaling pathway, that recapitulates human disease progression. Our preliminary data has shown that dysregulation in eicosanoid metabolism correlates with PTEN loss such that prostaglandin synthesis is significantly enhanced while proresolving cyp450 associated eicosanoid synthesis is downregulated. These correlations between PTEN, prostaglandin, and cyp450 associated protein expression levels are preserved agnostic of PTEN deletion when examining publicly available data. The PI3K/AKT signaling pathway is unequivocally induced upon PTEN deletion and as such we aim to further investigate the role of AKT in driving this bioactive lipid metabolism dysregulation. Utilizing genetic knockout hepatocytes, our data has shown that AKT regulates eicosanoid synthesis in an isoform specific manner. Further, eicosanoid have long been implicated in playing a role in modulating macrophage chemotaxis and polarization. Our data has also shown significant enrichment of macrophages, and previous work demonstrated that depletion of this macrophage accumulation via AKT2 deletion attenuated disease progression. As such, we aim to investigate the role of AKT specific isoforms in regulating hepatic eicosanoid biosynthesis, and the effect this has on macrophage chemotaxis and polarization in liver disease. Completion of this project will show the potential therapeutic benefit that targeting eicosanoid signaling may have in NAFLD and NASH. This work will also elucidate the mechanistic roles of each AKT isoform in regulating hepatic eicosanoid metabolism and chronic inflammation to provide more insights for the scientific community at large.

NSF Awards · FY 2024 · 2024-09

Quantum information processing brings fundamental new paradigms for sensing, computing and communication. It will enable foundational innovations in many fields from physics to chemistry, to energy and finance, to secure digital infrastructure, and more. However, the ability to communicate quantum information effectively and efficiently across distance is a crucial pre-requisite for reaping the benefits of many quantum applications, including large-scale quantum computing, distributed quantum sensing and quantum-enhanced (secure) communications. Building on top of recent advances in developing a quantum network along two separate lines, this project will develop a new hybrid architecture that will reap the benefits of both continuous-variable and discrete-variable quantum networking, and design algorithms and protocols to enable high-rate and high-fidelity distribution of quantum entanglement resources across long distances and for various future quantum applications. The research will result in concepts and tools to empower the future quantum infrastructure and ecosystem. The project will help develop the future quantum workforce by actively involving and broadening participation from high school and undergraduate students in quantum-related research. The key innovation of this project is a new hybrid continuous-discrete variable quantum network architecture, and a suite of models, algorithms and protocols, for understanding and enabling high-rate high-fidelity entanglement distribution for quantum computing, sensing and communication applications. To achieve this goal, the proposed research encompasses both theoretical and practical considerations around: how to model and optimize physical processes and protocols for generating and manipulating continuous- and discrete-variable entangled states, how to measure and optimize performance metrics of a hybrid continuous-discrete variable quantum network system with scalability, and what application-level utilities might such a hybrid network enable compared to an infrastructure using either continuous- or discrete-variable components alone. To answer these questions, the expected contributions of this project are: (1) developing new models and protocols for hybrid continuous- and discrete-variable entanglement generation, manipulation and conversion; (2) designing a generic hypergraph-based framework for optimizing the design of a hybrid network for end-to-end entanglement distribution; (3) developing models and algorithms to support future quantum applications in computing, communication and sensing with the hybrid architecture. The outcomes of this research will inform and empower the development of future quantum network technologies, by significantly expanding the flexibility in combining and utilizing heterogeneous devices and technologies in an end-to-end manner. 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.