University Of Chicago

NSF Awards · FY 2025 · 2025-07

As artificial intelligence (AI) continues to proliferate rapidly through society, there is a growing interest in incorporating core statistical principles, such as uncertainty quantification, into predictive systems to enable reliable and trustworthy AI decision-making. This research program aims to bridge the current conceptual gap between statistics and AI by ensuring that machine-assisted predictions are statistically valid, thereby supporting their safe and effective application to complex scientific challenges in data-rich domains such as imaging, personalized medicine, business analytics, marketing, and economics. The research agenda is organized around two overarching objectives, unified by a common thread: generative models in which data are viewed as stochastic outputs of computer programs. Conducting predictive inference in these models poses significant challenges due to their inherently opaque, black-box structure. The first objective is to develop a novel Bayesian inferential framework for generative models, leveraging modern machine learning tools such as deep learning and Bayesian Additive Regression Trees (BART). This work will lay the methodological and theoretical foundations for a new class of “generative Bayes” techniques that enable statistically principled inference in complex generative systems. The second objective focuses on advancing practical methodology for computerized adaptive testing (CAT), aimed at enhancing computer-human interactions through dynamically tailored questioning that adapts in real time to the respondent’s skill level with applications. The overarching goal of this research is to integrate modern machine learning tools into statistical modeling while establishing rigorous theoretical foundations that justify their practical use. The first project will develop a novel generative Bayesian framework for quantile-based learning using Bayesian Additive Regression Trees (BART). The outcome will be a flexible generative toolkit capable of simulating from a wide range of conditional distributions–core components for addressing numerous inferential tasks, including prediction. This work will chart a new path for nonparametric modeling of conditional distributions (such as posterior and posterior predictive distributions) via quantile learning under minimal assumptions. The methodological advances will be supported by a comprehensive frequentist-Bayesian theoretical analysis to assess the fidelity of distributional reconstructions. The second project will provide new theoretical insights into widely used sparsity-inducing priors, such as the horseshoe prior, by evaluating their performance from a predictive perspective. These contributions will deepen our understanding of the predictive properties of sparse Bayesian models and further enhance their applicability. The third project aims to bring machine learning techniques, particularly Q-learning, into the realm of computerized adaptive testing, thereby extending classical item response theory models to enable more responsive and individualized assessment tools. Together, these three projects will significantly advance the frontiers of nonparametric Bayesian methodology, theory, and applications. 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 · 2025-07

PROJECT SUMMARY In developing B cells, and in the germinal center (GC), genomic mechanisms of diversity must be strictly coordinated with those of proliferative selection. Failure to do so risks genomic instability and leukemic transformation. In the bone marrow (BM), this is done by ordering B lymphopoiesis into alternating and mutually exclusive states of either stochastic immunoglobulin gene (Ig) recombination or cell proliferation with selection. Likewise, GC B cells are divided into three subpopulations that occupy different niches and compartmentalize the incompatible functions of proliferation, somatic hypermutation and selection. In both the BM and GC, transit between proliferation and DNA mutation states requires large reordering of both genomic accessibility and transcription. In the BM, the epigenetic reader BRWD1 orchestrates the radical change in enhancer landscapes when cells exit proliferation and initiate Igk recombination. We have recently demonstrated that it does this by converting static to dynamic cohesin at topologically associating domain (TAD) boundaries. Dynamic cohesin then extrudes DNA across TADs to appose promoters and enhancers for gene activation and to contract Igk for recombination. We now provide preliminary evidence suggesting that BRWD1 is recruited to these TAD boundaries by specialized CTCF sites flanked by DNA GAGA sequences and that GAGA motifs are important for both BRWD1 recruitment and function. Finally, we demonstrate that deleting Brwd1 in GCs leads to disordered subsets and diminished function. Overall, our findings support a model in which BRWD1 orchestrates essential B cell functions, in both the BM and GC, by regulating dynamic cohesin-mediated chromatin topology to both determine enhancer landscapes and poise immunoglobulin genes for recombination. Specific Aims: Aim 1. To define BRWD1 GC functions and determine if these are associated with cohesin conversion. Hypothesis: We hypothesize that BRWD1 and cohesin conversion are used, and reused, to rewire B cell gene topology across cell state transitions. Aim 2: Determine the mechanisms and functional importance of BRWD1-mediated GAGA motif recognition. Hypothesis: BRWD1 recognizes GAGA motifs and this is important for both BRWD1 chromatin binding and chromatin remodeling. Aim 3. Determine how the recruitment of BRWD1 and cohesin are coordinated for cohesin conversion. Hypothesis: CTCF assembles BRWD1 with cohesin at TAD boundaries to orchestrate cohesin conversion.

NIH Research Projects · FY 2025 · 2025-07

PROJECT SUMMARY The overarching goal of this proposal is to integrate a pharmacy-technician based model for delivering long- acting injectable cabotegravir-rilpivirine (LA-CAB/RPV) into community-based federally qualified health centers (FQHCs) that care for underserved people with HIV (PWH). LA-CAB/RPV, given every 4 or 8 weeks, is a novel HIV treatment option that positively impacts quality of life and has been shown to improve viral suppression in PWH with adherence challenges. However, rollout has been slow due to complex delivery logistics such as navigating insurance coverage and procuring medication from designated specialty pharmacies. Despite strong interest from PWH and providers, this complexity has prevented many FQHCs, which provide HIV care regardless of ability to pay or insurance, from being able to offer LA-CAB/RPV. Preliminary data from academic HIV clinics demonstrate that a LA-CAB/RPV delivery model which places a pharmacy technician in a key role, along with multidisciplinary case conferences (coalition-building), robust relationships with the specialty pharmacies that supply LA-CAB/RPV (network weaving), and tools for quality monitoring, allows programs to effectively initiate and maintain PWH on treatment. In this study, guided by two implementation science frameworks, the Consolidated Framework of Implementation Research (CFIR 2.0) and Proctor’s implementation outcomes, we will adapt and test the pharm-tech based model in five FQHC clinics in Chicago, assessing cost and equity impacts. Aim 1 will consist of formative work to adapt and refine the pharm tech- based model for the FQHC setting via rapid ethnography and implementation mapping. Aim 2 involves a hybrid Type III implementation-effectiveness study. To evaluate the effect of the PHARM ART model, we will compare the LA-CAB/RPV initiation rate in implementation clinics vs. a group of contemporaneous control clinics using a controlled interrupted time series design over three phases: before the PHARM ART model (pre- implementation phase – 24 months prior to initiation of PHARM ART), during active facilitation by the study team (implementation phase – 21 months), and with cessation of active facilitation (sustainment phase – 21 months). Using electronic medical record data and iterative qualitative and survey research, we will delineate mechanisms by which the PHARM ART model works in the FQHC setting, assess the quality of implementation, and evaluate clinical effectiveness outcomes. In Aim 3 we will estimate cost, budgetary impact and cost-effectiveness of PHARM ART and associated strategies. We will adapt an existing mathematical model to evaluate epidemiologic and economic impact across populations, applying distributional cost- effectiveness analysis (DCEA) to assess health equity impacts, including scenarios with varying levels of baseline viral suppression. The proposed work will provide important information on implementation strategies and cost/equity considerations to guide decisions by clinics and funders to invest in capacity-building.

NIH Research Projects · FY 2026 · 2025-07

PROJECT SUMMARY Cancer remains a global health challenge despite significant advancements in diagnosis and treatment. In the past two decades, immunotherapy has dramatically changed the landscape of cancer treatment, with immune checkpoint inhibitors (ICIs) targeting PD-1/PD-L1 and CTLA-4/B7-1 showing notable success. However, these therapies benefit only a subset of patients, underscoring the need to explore additional immune regulatory mechanisms. The T cell immunoglobulin and ITIM domain (TIGIT) has been identified as another promising checkpoint for immunotherapy, with preclinical studies showing potential. However, clinical trials have revealed that anti-TIGIT monotherapy offers limited benefits for patients with advanced solid tumors. The intricate network of TIGIT signaling, characterized by its extensive crosstalk and interaction with other co-inhibitory and co- stimulatory pathways, suggests that targeting TIGIT alongside other pathways using bi- or trispecific antibodies might enhance antitumor efficacy. The primary challenge in leveraging this approach is target determination due to the complexity of TIGIT network, which is influenced by a delicate balance of various components, each contributing differently to immune regulation. Developing a quantitative model that accurately reflects the role of each element within the TIGIT pathways as well as their collective impact is crucial for devising effective targeted therapy combinations. However, the field faces substantial technical hurdles, especially in achieving precise, noninvasive, and simultaneous control of multiple protein components in T cells to assess their collective impact on immune suppression. To address this, we propose the development of a chemogenetic toolbox that allows for the controlled expression of multiple proteins. This toolbox will allow us to dissect the intricate immune suppression mediated by the diverse elements in the TIGIT network. This strategy utilizes destabilization domains (DDs) which, upon binding to specific small molecules (SMs), can modulate protein levels in a dose- dependent manner. Despite the availability of some DD:SM pairs, their application is limited by issues such as interference with endogenous systems, poor SM permeability, and lack of orthogonality. To address this and to study TIGIT signaling, our project is structured around three main objectives: (1) engineering new and orthogonal DD:SM pairs through molecular docking and advanced protein engineering to overcome existing limitations (Aim 1), (2) developing a high-throughput assay to unravel the complex immune inhibition orchestrated by various components within the TIGIT network and to establish a quantitative model (Aim 2), and (3) Generating bispecific antibodies that target critical elements within the TIGIT network to augment the immune response (Aim 3). Through this research, we aim to unlock new insights into TIGIT signaling and establish a solid framework for assessing the most effective therapeutic combinations, ultimately enhancing the arsenal of immunotherapy, and potentially improving outcomes for cancer patients.

NSF Awards · FY 2025 · 2025-07

This I-Corps project focuses on a minimally-invasive medical device designed to prevent strokes in individuals with irregular heart rhythms who are unable to take blood-thinning medications. As the U.S. population continues to age, the prevalence of irregular heart rhythms, particularly atrial fibrillation, is rising significantly. This condition increases the risk of stroke due to clot formation in a small outpouching of the heart known as the left atrial appendage. While oral medications to prevent clotting are effective for many, a substantial portion of patients (up to one in five) are ineligible due to contraindications such as high bleeding risk. Without a safe and effective alternative, these patients face elevated risks of life-threatening complications. Existing mechanical closure devices are rigid and metallic, often poorly suited to the highly variable anatomical structure of the left atrial appendage, and have been linked to adverse outcomes. This project explores a new approach involving a soft, adaptable closure method intended to improve patient safety, reduce healthcare costs, and increase access to life-saving stroke prevention for a vulnerable patient population. This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of the technology. This solution is based on the development of a percutaneously delivered catheter system that accesses the heart through the vascular system and deposits a gel-based material into the left atrial appendage. Unlike traditional implantable metal devices, which are round and rely on mechanical fixation in an irregularly shaped structure, this system enables complete, conformal sealing by filling the space with a biocompatible gel that hardens in place. The design is engineered to minimize risks such as perforation, incomplete closure, and post-operative bleeding, while adapting to patient-specific anatomy. The technical advancement lies in the directional control of gel delivery, in-situ curing capability, and the elimination of long-term foreign metallic implants. 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 · 2025-07

Scientists are increasingly incorporating machine learning (ML) and artificial intelligence (AI) techniques into their applications to accelerate and enhance scientific research and discovery across a wide range of disciplines. For example, machine learning has been successfully integrated into tools for weather forecasting, earth sciences, astronomy, high-resolution imaging, genomics, and molecular biology. However, the ever-growing size of scientific datasets results in prohibitive hardware resource costs, significantly complicating the deployment of these applications on high-performance computing platforms at scale. Lossy compression — a data reduction technique that significantly reduces dataset size by removing redundant or less important information — has proven effective for many scientific datasets, including those from cosmology and structural biology. Despite its promise, integrating lossy compression into AI-driven scientific applications remains a non-trivial challenge, requiring broad expertise in data compression and machine learning, as well as a deep understanding of application requirements, system considerations, and their interactions. These complexities hinder the adoption of this powerful data reduction technique in scientific applications. The overarching goal of this project is to address this gap by providing a cyberinfrastructure that seamlessly and adaptively integrates lossy compression into deep learning pipelines within scientific applications. This integration will reduce memory usage and communication overhead, enabling AI-for-Science applications to scale to massive datasets. The design includes several key innovations. First, it features a user-friendly interface that allows users to define accuracy requirements — which may evolve during application execution — and to instantiate different compressors, supporting both customization and extensibility. Second, it provides a software layer that integrates with popular deep learning frameworks, such as PyTorch, enabling compression to be applied to existing neural network models with minimal code modifications. Third, it incorporates an adaptive execution engine that dynamically selects the appropriate compressors and error bounds based on the desired accuracy, data characteristics (e.g., smoothness, value range, sparsity), model structure, and system configuration. The cyberinfrastructure will support both existing and emerging machine learning accelerators and will be released as open-source software, accompanied by documentation and training materials to promote adoption within the scientific and computing communities. Ultimately, this project has the potential to benefit the broader community by enabling scalable, AI-driven scientific discovery. 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 · 2025-07

PROJECT SUMMARY Nine out of ten women in the US take at least one medication during pregnancy. However, due to ethical and practical constraints, pregnant women are typically excluded from clinical trials, resulting in a lack of data on medication safety during pregnancy and potential impact on the child. Epidemiological studies are increasingly using real-world data (RWD), including electronic medical records (EMRs) to study treatment effects during pregnancy, but these studies have relied primarily on structured data. This can miss additional context and information that can be gained from unstructured data sources (i.e. clinical notes), potentially resulting in limited analytic precision and potential bias. To investigate how maternal medication history affects child health outcomes, we will use both structured and unstructured data from a large repository of EMRs of mothers and their children. Specifically, in Aim 1, we will build a registry of mother-child pairs containing high-quality phenotype data, which we will extract from EMRs using large language models (LLMs). We will use a paired mother-child data mart that will include all individuals who were pregnant between 2012 and the end of the study, based on the clinical records at UCM, as well as the clinical records of their children. We will utilize LLMs to enhance our ability to extract additional medication exposures and phenotypes from clinical notes. In Aim 2, we will identify robust correlations between maternal medication exposures and early life health outcomes. We will perform a broad binary statistical association of maternal medication exposures to outcomes using a medication-wide association study (MedWAS), or modified PheWAS, approach. We will then perform rules-based refinement of select exposures and outcomes, which will include the application and development of exposure ascertainment and phenotype algorithms. In Aim 3, we will use causal inference models to investigate the effects of maternal medication exposures on early life health outcomes. For a subset of associations identified in Aim 2, as well as associations established in the literature, we will utilize causal inference machine learning approaches to estimate both binary and continuous (e.g., dose-dependent), individualized treatment effects. Overall, this work will provide insights into the safety and risks of medication exposures during pregnancy, which may inform data-driven treatment strategies for pregnant patients. At the same time, this work will provide me abundant opportunities to grow my computational research skills while benefitting from the support of diverse collaborators and mentors spanning clinical, informatics, and genomics expertise.

NIH Research Projects · FY 2025 · 2025-07

ABSTRACT: Advances in molecular biophysics often occur when tools, techniques, and ideas from the physical sciences are joined to biological research questions to create new insights and opportunities, revealing microscopic design principles within biology. BPHYS is an independent graduate training program at The University of Chicago founded on a training model in which an interdisciplinary bridge between the physical sciences and the life sciences is built within each student. The hallmark of the program is that our students are co-mentored by two faculty with complementary research interests and receive a Ph.D. degree from both the Physical Sciences Division and Biological Sciences Division. By making our students full members of both mentors' laboratory groups at the outset, trainees learn to identify and bridge intellectual gaps between disciplines. This allows students to overcome cultural barriers between fields and identify opportunities for innovation. Our students greatly value this experience and intellectual independence, and we attract daring, creative, and driven students. BPHYS trainees receive training in both biological and physical sciences through a combination of lab-based courses, didactic courses alongside students in traditional disciplinary programs, and program activities. Intense interdisciplinary didactic and practical training is implemented by the trainee through research on an innovative, student-designed, dual-mentored thesis project, which is investigated using both physical and biological approaches. The trainee is a full member of both mentors' laboratories and becomes adept at communicating across disciplinary boundaries. Program mentors actively recruit (and compete for) our BPHYS students because they forge new and lasting collaborations between groups. Due to this demand, we have increased the number of students who matriculate into the Program to ~10 per year with no drop in quality with further growth supported by our applicant pool and trainer funding profiles. Accordingly, this application requests 10 trainee slots to support 5 students in the Program in their first and second years.

NIH Research Projects · FY 2026 · 2025-07

PROJECT SUMMARY Bacterial secretion systems encompass a wide range of macromolecular nanomachines which are involved in a variety of functions crucial for cell viability. Despite their significance, our understanding of how the majority of secretion systems known to date are built and function inside a bacterial cell remains limited. In part, the latter is caused by the lack of high-resolution structural information pertaining to intact cells. At our lab, we aim at utilizing cutting-edge microscopy techniques, such as cryo-electron tomography (cryo-ET), and a number of other associated methodologies, to study the assembly, disassembly, and function of bacterial secretion systems. Those techniques enable direct visualization of macromolecular complexes within intact, hydrated- frozen bacterial cells at nanometer resolution. As a start-off, we will focus on the following two specific secretion systems: 1) the bacterial motility nanomachine, the flagellar motor (an example of type III secretion system), and 2) the type IV secretion system (T4SS), the latter being involved in a wide range of processes such as secreting effector proteins and genetic material transfer between bacterial cells (bacterial conjugation). Using cryo-ET, we will unveil the mechanism of how exactly bacteria assemble their seemingly complex flagellar motors in stepwise fashion, and how they disassemble them under harsh environmental conditions to save energy. Jointly utilizing cryo-ET and cryo-light microscopy, we will subsequently decipher the role of the T4SS in the process of intercellular DNA transfer, visualizing the accompanying T4SS structural changes. The results of our studies will deepen the understanding of how, in general, macromolecular complexes assemble and perform their functions under various environmental conditions.

NSF Awards · FY 2025 · 2025-07

Identifying innovative ways of removing particulates from fluids remains of practical interest in areas from industrial processes to purification of water from microplastics. Filtration is challenging because of filter clogging and the large amount of energy required to push water through filters. This project will investigate a filter design in which the filters are suspended in the fluid in a turbulent flow. The filters can then be readily separated from the fluid and emptied. This new concept has the potential to revolutionize engineering approaches to filtration. By suspending the filters in the flow, large pressures required to drive fluid through the filters are avoided and pumping is substituted by shaking of the fluid. The key step in this new process is to understand the way particles move through the filter’s pore with random flow forcing on one side. This research will introduce students to new avenues of interdisciplinary science research, communicate science to the public from open platforms, such as to museums, and develop teaching material to advance STEM education. The goal of this project is to understand the fluid dynamics particle transport through pores of comparable size in the presence of random flow on one side of the pore. This complex process involves the interaction between Stokesian dynamics at the pore and viscous and pressure fluctuations driving the motion on one side of the pore. Experiments are proposed using high speed imaging of the process and particle imaging velocimetry of the fluid velocities. Simulation modeling of transport through pores will complement the experiments. Understanding this process will shed light on an alternative approach to filtration in which particles are captured by suspended filters, and filtration occurs in a fluid by shaking them together. The transport of particles through the filter pores determines the efficiency of filtration. 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 · 2025-07

Rare natural hazards (for example, storm surge and hurricanes) can cause loss of lives and devastating damage to society and the environment. For instance, Hurricane Katrina (2005) caused over 1,500 deaths and total estimated damages of $75 billion in the New Orleans area and along the Mississippi coast as a result of storm surge. Uncertainty quantification (UQ) has been used widely to understand, monitor, and predict these rare natural hazards. The Gaussian process (GP) modeling framework is one of the most widely used tools to address such UQ applications and has been studied across several areas, including spatial statistics, design and analysis of computer experiments, and machine learning. With the advance of measurement technology and increasing computing power, large numbers of measurements and large-scale numerical simulations at increasing resolutions are routinely collected in modern applications and have given rise to several critical challenges in predicting real-world processes with associated uncertainty. While GP presents a promising route to carrying out UQ tasks for modern emerging applications such as coastal flood hazard studies, existing GP methods are inadequate in addressing several notable issues such as computational bottleneck due to big datasets and spatial heterogeneity due to complex structures in multi-dimensional domains. This project will develop new Bayesian GP methods to allow scalable computation and to capture spatial heterogeneity. The new methods, algorithms, theory, and software are expected to improve GP modeling for addressing data analytical issues across a wide range of fields, including physical science, engineering, medical science, public health, and business science. The project will develop and distribute user-friendly open-source software and provide interdisciplinary research training opportunities for undergraduate and graduate students. This project aims to develop a new Bayesian multi-scale residual learning framework with strong theoretical support that allows scalable computation and spatial nonstationarity for GP modeling. This framework integrates and extends several powerful techniques respectively arising in the literature on GP and that on multi-scale modeling, including predictive process approximation, blockwise shrinkage, and random recursive partitioning on the domain. This framework decomposes the GP into a cascade of residual processes that characterize the underlying covariance structures at different resolutions and that can be spatially heterogeneous in a variety of ways. The new framework allows for adoption of blockwise shrinkage to infer the covariance of the residual processes and incorporates random partition priors to enable adaptivity to various spatial structures in multi-dimensional domains. New recursive algorithms inspired by wavelet shrinkage and state-space models will be developed to achieve linear computational complexity and linear storage complexity in terms of the number of observations. The resulting GP method will guarantee linear computational complexity in a serial computing environment and also be easily parallelizable. This Bayesian multi-scale residual learning method provides a new approach to addressing GP modeling issues among spatial statistics, design and analysis of computer experiments, machine learning, and nonparametric regression. 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 · 2025-07

Two-dimensional geometry has been a cornerstone of mathematics since antiquity. In addition to geometry on the xy-plane, mathematicians today study geometry on surfaces such as the crust of the Earth or the glaze of a donut. Most of these surfaces exhibit hyperbolic geometry, a type of non-Euclidean geometry in which parallel lines spread apart. Because of its ubiquity, two-dimensional hyperbolic geometry is a common meeting point for many branches of mathematics. Recently, the PI strengthened a connection linking hyperbolic geometry with certain transformations of two-dimensional Euclidean geometries, which were motivated by the kinetic motion of particles and the flow of electrons. The goal of this project is to further investigate this connection between non-Euclidean and Euclidean geometries in order to better understand both worlds. The project will also support professionalization, mentorship, and enrichment activities for a broad range of students. The research component of this project has three main prongs of attack. The first is to further develop (and complicate) the link between the earthquake and horocycle flows, dynamical systems on moduli spaces of hyperbolic surfaces and quadratic differentials, respectively. This will allow the PI both to leverage decades of progress on quadratic differentials to study hyperbolic surfaces, and to leverage new geometric features afforded by hyperbolic space to study quadratic differentials. The second prong is to study the structure of optimal Lipschitz maps between hyperbolic surfaces and their related renormalization dynamics. The final prong is to use all of these tools to investigate related counting problems on both flat and hyperbolic surfaces. 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 · 2025-07

PROJECT SUMMARY/ABSTRACT The hippocampus plays a crucial role as part of a distributed brain network supporting memory for life events. This exploratory project seeks to develop an evidence base for the use of intracranial EEG (iEEG) to investigate how noninvasive stimulation can be used to modulate hippocampal network neural activity and thereby test its role in memory. Hippocampal Indirectly Targeted Stimulation (HITS) is a noninvasive stimulation approach that uses transcranial magnetic stimulation of parieto-occipital regions of the hippocampal network. Previous studies indicate that HITS can selectively alter hippocampal activity observed through fMRI and enhance memory performance. The critical challenge addressed by this exploratory project is understanding neural mechanisms for how HITS affects network activity and behavioral performance. iEEG offers superior temporal and spatial resolution, which will provide detailed information on how HITS affects memory-related neural activity. The research has three primary aims: (1) investigating the selectivity of HITS effects on iEEG activity of the hippocampal network, (2) testing whether HITS impacts the endogenous theta- rhythmic oscillatory activity of the hippocampal network, and (3) determining if HITS specifically modifies memory-related hippocampal iEEG activity. These aims are addressed through a series of experiments involving HITS and detailed behavioral testing in neurosurgical patients undergoing iEEG monitoring for clinical treatment of epilepsy. This exploratory research would establish the HITS-iEEG method and provide initial mechanistic insights, with experiments designed to provide information to motivate future research on how HITS could be used to address the specific hippocampal network abnormalities associated with memory impairments in different neurological and psychiatric disorders. Relevant to the goals of PA-21-219, the hippocampal network is disrupted in disorders that are of joint interest to NIMH and NINDS, including brain injury, epilepsy, Alzheimer’s disease, schizophrenia, major depression, and PTSD. Positive outcomes from this exploratory study would support future research on HITS applications to these disorders, including using detailed knowledge of HITS effects on hippocampal network neural activity to develop targeted interventions for the specific network abnormalities responsible for memory and cognitive impairments. Detailed knowledge of how HITS affects brain network function will also motivate scientific applications for the use of HITS as a tool to test theories of the brain-behavior relationships supporting memory for life events.

NSF Awards · FY 2025 · 2025-07

Ultimately, people want to feel connected to people on their block, in their neighborhood, and in their city or town. That connection can in large part depend on interpersonal interactions, whether experienced as pleasant or uncomfortable, and these interactions are often shaped by people’s positive or negative attitudes. What is known about attitude formation is largely based on individual experiences and how those individual experiences influence attitude formation. This project advances what is known about attitude formation and attitude changes, by examining how geographic contexts shape the formation of attitudes and influence the quality of interpersonal interactions. This project tests three questions derived from an interdisciplinary mathematical model of attitude formation. This model draws inspiration from social psychology and urban science to explicitly account for interactions at different spatial scales. Specifically, the project tests: (1) How much do attitudes vary across place and spatial scale? (2) How much can the nested contexts of cities and neighborhoods explain about attitude formation? (3) How much variation in attitudes across spatial scales and places is related to the quality of local interpersonal interactions? To support training for students and researchers, the project also includes the development of an online course and a set of example data analysis pipelines demonstrating the application of these methods. This research develops novel integrative methods in attitude research, offering new theoretical frameworks that promote national prosperity and welfare. This project is jointly funded by Social Psychology and the Established Program to Stimulate Competitive Research (EPSCoR). 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 · 2025-07

PROJECT SUMMARY Circulating cell-free DNA (cfDNA) and RNA (cfRNA) offer enormous potential as a new class of biomarkers for developing non-invasive liquid biopsies applicable to numerous diseases and conditions. Recent years have seen intense study of cfDNA and cfRNA as tools for non-invasive prenatal testing, solid organ transplantation, cancer screening, and tumor monitoring. However, early cancer detection has been hindered by the low concentration of tumor-derived cfDNA in plasma, resulting in decreased sensitivity and specificity. While epigenetic-based cfDNA sequencing approaches have been developed; however, their effectiveness remains limited by the low abundance of target analyte at early stages. Cell-free RNA (cfRNA) presents a promising alternative, yet its practical application of cfRNA faces hurdles: rapid degradation of cell-free mRNA, insufficient quantities in clinical samples, and cellular contamination altering mRNA levels, thus compromising the detection specificity. Moreover, the study of epitranscriptomic markers from cfRNA remains challenging due to the lack of robust sequencing methods. To address these critical knowledge gaps, this project aims to (i) profile the transcriptomics and epitranscriptomics of plasma cfRNA, (ii) develop epitranscriptomic patterns for microbiome- derivative cfRNA as biomarkers for colorectal cancer (CRC) diagnosis. As described in the preliminary data in this application, I have developed an innovative method called "Low-Input Multiple MEthylation Sequencing" (LIME-seq) to monitor abundance and site-specific modifications across multiple RNA species from ultra-low- input cfRNA (sourced from only 600 μL plasma). Applying this method to human plasma samples, I have successfully uncovered the presence of tRNAs and specific ncRNAs as major cfRNA components originating from both human and microbiome sources. Importantly, I discovered that methylation patterns in microbiome- derived cfRNA represent an effective class of non-invasive biomarkers with unprecedented potential for early CRC diagnosis, offering a distinct advantage over microbial abundance profiles obtained from cfRNA/cfDNA sequences. These methylation patterns accurately reflect microbiota activity, surpassing traditional abundance- based methods and providing a more prescise approach to study human microbiota. Leveraging the LIME-seq technique, I will conduct a comparative analysis between the transcriptomic and epitranscriptomic landscapes of specific microbiomes and that of plasma cfRNA. This will validate the presence of certain microbiome-derived cfRNA and confirm the RNA methylation sites. I will also demonstrate the utility of microbiome-derived cfRNA profiles sourced from plasmain CRC diagnosis, with a particular focus on adenoma and stage I. The projected outcomes of this study will generate both diagnostic and predictive biomarkers from methylation sites of microbiome-derived cfRNA and provide a non-invasive method for assessing microbiota. Furthermore, a more ambitious vision for this research could be to identify novel microbiome targets, which could become potential therapeutic strategies for CRC prevention and treatment.

NIH Research Projects · FY 2025 · 2025-06

Abstract Following exposure to antigens, naïve B cells differentiate by going through germinal center reaction into plasma cells or memory B cells. Plasma cells secrete antigen-specific antibodies which aid in the clearance of invading pathogens by direct neutralization, by activating the complement cascade or by interacting with other immune cells by binding to Fc receptors. Long lived memory B cells can be reactivated to differentiate into plasma cells following secondary infection. Together, memory B cells and long-lived plasma cells form the basis for life-long B cell-mediated protection against various infections. To evade host protective immunity all successful pathogens have evolved numerous mechanism to directly or indirectly manipulate B cells. Pathogens could manipulate B cells by using them as a reservoir, by interfering with B cell maturation and by modulating B cell survival. Mouse mammary tumor virus (MMTV) is an exogenous retrovirus which persist indefinitely in mice of susceptible strains. B cells represent the primary and indispensable target for this virus and yet persistently infected mice do not make significant titers of virus-neutralizing Abs. Unlike mice from MMTV-susceptible mice I/LnJ mice even though become virus-infected produced potent virus-neutralizing Ab responses leading to the virus elimination. Using a positional cloning approach, we discovered the negative immune regulator of the major histocompatibility class II (MHCII) peptide presentation, DO is dysfunctional in mice from the I/LnJ strain and its dysfunction is a sole reason why these mice generate virus-neutralizing Ab responses. DO is a non-classical Major Histocompatibility Class II (MHCII)-like molecule which has evolved as a negative regulator of yet another MHCII-like molecule, DM. DM catalyzes peptide loading of MHCII on antigen presenting cells (APCs) for presentation to CD4 T cells. In the absence of DO there is a significant increase of high-affinity peptides presented by MHCII as DM is uninhibited. Whereas naïve B cells express high levels of DO, DO is downregulated during generation of Ab responses to allow for presentation of high affinity peptides and entering the germinal center reactions. We hypothesis that non-lytic viruses that infect B cells prevent downregulation of DO in germinal center B cells to avoid presentation of high affinity virus-specific peptides. The studies described in this basic science proposal will determine whether a B cell infecting retrovirus, such as MMTV impedes the mechanisms controlling DO levels in germinal center B cells to prevent the development of isotype switched, virus-specific Ab responses.

NIH Research Projects · FY 2026 · 2025-06

PROJECT SUMMARY/ABSTRACT Ulcerative colitis (UC) is an idiopathic chronic inflammatory condition of the large intestine. Current treatment targets are aimed at healing the inflamed mucosa, however, most patients who achieve mucosal healing do not have sustained remission and continue to have reduced quality of life and a variety of comorbidities. This proposal provides a paradigm shift in the approach to UC as disease that is not mucosal, but transmural in its nature. The broad objective of this research proposal is to test the hypothesis that transmural structural alterations (TMA) to the bowel wall in UC lead to poor clinical outcomes as compared to UC that has a preserved bowel wall structure. I will study this using a novel approach of in vivo assessment of the bowel by intestinal ultrasound (IUS), and determine the predictors for TMA, the associated outcomes with such changes, and whether these changes are reversible. If successful, this proposal will identify new therapeutic targets and endpoints in clinical trials. To accomplish this, I propose the following specific aims: 1) Determine the predictors of TMA to the bowel wall in patients with UC with either active or remitted mucosal inflammation. 2) Determine disease-related outcomes of UC patients who have TMA and transmural healing (TH) of the large intestine. 3)Determine the relationship between TMA of the rectal wall and rectal function. I seek to develop a career as an independent physician-scientist focused on outcomes research and disease modification in IBD. My career goals require me to obtain additional training in patient-reported outcomes, longitudinal data analysis and advanced bowel ultrasound techniques. The 5-year career development program outlined in this application will be based at the University of Chicago, which has attained distinction in the clinical care and research of IBD and is a leading IUS center in the US. Dr. David T. Rubin, Professor of Medicine, is my primary mentor and will provide expertise in endpoint development and application of emerging diagnostic tools for the management of UC. Dr. Eugene Chang, Professor of Medicine, is my co-mentor and an expert on the pathophysiology and immunology of IBD. An interdisciplinary Advisory Committee will provide guidance for me during this development period: Dr. John Hart, expert in IBD pathology, Dr. Mihai Giurcanu, expert in biostatistics, Dr. Maria Abreu, established and NIH funded translational scientist in IBD with IUS experience, Dr. Corey A. Siegel, expert in outcomes research in IBD, and Dr. Shintaro Sagami, expert in the use of US for UC. With the mentorship, training, and resources available to me, I will be uniquely equipped to undertake an independent research career in outcomes research in IBD by the completion of this career development award.

NIH Research Projects · FY 2026 · 2025-06

PROJECT SUMMARY/ABSTRACT Episodic memory impairment occurs when diseases impact the hippocampus and its network of interacting brain areas. Better understanding of brain mechanisms for memory is required to development treatments. This project will develop experimental approaches that use stimulation to manipulate brain-behavior relationships in humans and thereby directly test how the hippocampal network supports memory. Intracranial EEG (iEEG) provides opportunities to study the neural basis of memory with high spatial and temporal resolution. Many experiments have identified iEEG signals associated with memory, for instance by comparing instances of memory success to failure. However, the functional relevance of these signals remains unclear, as such comparisons can identify signals of extraneous, co-occurring, cognitive processes. A routine approach to address this limitation in the behavioral neurosciences is via bi-directional manipulations, i.e., exogenous enhancement or inhibition of the putative neural process and tests of corresponding enhancement or inhibition of behavioral performance. This approach has been underutilized in human iEEG memory research, in part due to difficulties in reliably influencing memory or neural signals given currently available manipulations, such as direct electrical stimulation (DES) through iEEG electrodes. As a route to achieve bi-directional manipulation experiments in human iEEG memory research, this project utilizes a noninvasive electromagnetic stimulation tool that reliably and robustly influences hippocampal network activity and memory performance, called “hippocampal indirectly targeted stimulation” (HITS). HITS will be used with iEEG for the first time to identify activity patterns of the hippocampal network related to memory behavioral performance enhancement versus impairment. To rigorously test for behavioral relevance, we will use a closed-loop machine-learning approach to identify DES parameters that reproduce the hippocampal-network iEEG activity patterns generated by HITS, and then test whether delivery of DES to enhance or inhibit these hippocampal-network iEEG activity patterns produces corresponding enhancement versus impairment of memory behavioral performance. Thus, bi- directional manipulation will be used to directly/causally test iEEG signals of memory behavior. Group-level modeling of DES parameters that mimic the effects of HITS will be used to broaden the accessibility of bi- directional manipulations for future research that tests memory mechanisms in larger iEEG samples across multiple sites that lack the technical capability for HITS. Across three sites, will test whether bi-directional manipulations of hippocampal-network activity involvement in memory performance are more successful for HITS-informed DES versus DES based on a priori hypotheses. This exploratory project will thus build a foundation for future studies to rigorously test hippocampal-network support of memory behavioral performance using bi-directional experimental manipulations and will establish a collaboration of multi- disciplinary researchers across multiple institutions to perform this innovative mechanistic memory research.

NIH Research Projects · FY 2026 · 2025-06

Project Summary The American Urological Association Symptom Index (AUA-SI) has been the primary patient-reported outcome (PRO) used in men with benign prostatic hyperplasia (BPH) for decades; however, it is not a comprehensive assessment of lower urinary tract symptoms that may present with BPH because it lacks items on urinary incontinence and bladder pain. The Symptoms of Lower Urinary Tract Dysfunction Research Network (LURN) developed general-use multidimensional symptom questionnaires in order to address the complex needs of the urology clinic, but these tools have yet to be validated in specific clinical populations. The overarching goal of this project is to use these questionnaires and other PROs to enhance clinical decision-making in men with BPH. This will be achieved by (1) integration of PRO assessments into the evaluation and management of patients with BPH; (2) determination of clinically meaningful differences in LURN Symptom Indices in men with BPH receiving known effective treatment; and (3) creation of care-coordination recommendations to facilitate using PROs in order to match evidence-based treatments to patients with persistent symptoms. PROs are a cost- effective strategy that, combined with advances in electronic medical health record technology, can improve patient care by allowing for more real-time surveillance and intervention, including the monitoring of both the core urologic symptoms of BPH and common comorbid symptoms such as sleep disturbance and depression. In order to effectively determine meaningful symptom changes in response to treatment, minimally important differences specific to men with BPH must be established. Moreover, previously demonstrated convergent validity of the questionnaires must be confirmed in this population. In addition to using PROs to assess treatment response, they can also be used to monitor short-term side effects and comorbid conditions over the course of treatment, a practice that can facilitate care coordination that will increase quality of care and quality of life for men with BPH. In Aim 1, we will use statistical methods to predict changes in urologic symptoms and assess associations between lower urinary tract symptoms and measures of sleep disturbance, depression, and pain, among others. In Aim 2, we will assess test-retest reliability and other psychometric properties of the LURN Symptom Indices, using a triangulation of methods to determine minimally important differences. In Aim 3, we will engage stakeholders, including patients and clinicians, and use qualitative methods to create care coordination checklists intended for physician extenders to follow up with patients undergoing treatment, including via telehealth. Aim 3 will also enhance the recovery and routine monitoring of men with BPH after surgery, who can experience distressing side effects and complications such as urinary leakage and pain. The ultimate result of this work will be better understanding of the associations between comorbid sleep disturbance, depression, anxiety, etc., with urinary symptoms in men with BPH, and improved care and reduced symptom burden through the integration of robust and validated PROs into multiple phases of urologic care.

NSF Awards · FY 2025 · 2025-06

Gen3 is an open-source software platform that enables researchers to manage, analyze, and share large-scale scientific data, which are critical hurdles that often slow scientific advances. Gen3 manages petabytes of data that includes millions of data objects that are findable, accessible, interoperable, and reusable for scientific research and education. While Gen3 has historically been used by larger organizations that have experience with setting up and operating large-scale cloud computing applications, this project expands the reach of Gen3 to bring these same powerful data sharing and analytical solutions to smaller research organizations, so that they can obtain the product's benefits with respect to scientific advancement and reproducibility. With funding from the Pathways to Enable Open-Source Ecosystems (POSE) program, the project establishes a mature and capable open-source ecosystem around Gen3. The project enables distributed improvements to the code base from the community, the creation of new scientific use cases, and a governance structure that makes the Gen3 development roadmap responsive to community needs. This project strengthens the communities of users and developers around this open-source data sharing and analysis platform. The project focuses on making the product easier to set up and maintain, prioritizing features needed by the research community. This effort makes the product available more widely to research and educational communities, including those organizations with limited technical and/or monetary resources. To accomplish these goals, the project 1) improves documentation including a new site and additional content, 2) creates new working groups and hackathons focused on particular topics, 3) creates a technical steering committee for Gen3, and 4) expands direct user support for timely response to questions and code contributions from the open-source 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.

NIH Research Projects · FY 2026 · 2025-06

Project Summary/Abstract Adolescents and young adults (AYA) with cancer face unique psychosocial challenges during their cancer journey, especially individuals diagnosed with a hematologic malignancy (HM), often resulting in clinically significant depression. While evidence-based depression treatments are available for AYAs without cancer, AYAs with cancer, in general, and AYAs diagnosed with HM, in particular, encounter mental health disparities in regard to evidence-based mental health treatment. Technology-assisted cognitive behavioral therapy (tCBT) has the potential to reduce these disparities, but existing tCBTs have major weaknesses, such as being text- heavy, academically oriented, and not engaging enough. Our team has developed and pilot-tested a fun and engaging tCBT intervention among AYA cancer survivors. While promising, the platform’s core components were not designed for the AYA cancer population. In this R34 application, we will first tailor the engaging tCBT program, MYTH, for the unique and specific needs of the AYA HM population to create MYTH-C (study Aim 1). Specifically, we will collaborate with an advisory panel composed of AYA cancer survivors, many diagnosed with HM, and stakeholders, i.e., oncologists, oncology nurses, and other allied oncological providers. The advisory panel will closely guide our refinement/tailoring of MYTH for AYA HM patients, i.e., MYTH-C. Concurrently, in Aim 2, we will collect essential data to design and plan the next level multicenter randomized controlled trial (RCT) across the three collaborating study sites: University of Michigan Health, University of Chicago Medicine, and MD Anderson Cancer Center. We will work with each study site’s staff to obtain essential clinical trial-related data, including but not limited to screening-to-enrollment data, anticipated site- specific patient volume, and how MYTH-C may interact with each study site’s clinicians’ day-to-day workflow. Finally, in Aim 3, we will conduct a pilot RCT to evaluate the feasibility and acceptability of delivering MYTH-C at the University of Chicago Medicine site. Patients will be randomly assigned to MYTH-C or treatment-as- usual (TAU) using the University of Chicago Medicine standard psychosocial care for AYA HM patients and post-treatment survivors. Participants will be evaluated at baseline, throughout the session, immediately post- intervention, 4 weeks post-intervention, and 3 months post-intervention using a set of process and clinical measures outlined in the application.

NSF Awards · FY 2025 · 2025-06

Scientists across various fields face the challenge of answering numerous related questions with limited or noisy data. For example, genomicists may need to assess thousands of genes using data from only a few subjects, while survey statisticians might analyze average incomes in many towns based on limited surveys. To address these challenges, researchers often borrow information from related questions, a process that requires sophisticated statistical reasoning due to varying underlying characteristics. Empirical Bayes offers a method to enhance individual question inference by sharing information, potentially improving statistical accuracy. However, it relies on strong modeling assumptions, limiting its applicability. This research aims to make empirical Bayes more powerful and accessible by integrating it with modern machine learning and causal inference, demonstrating its effectiveness under fewer assumptions, and developing methods to assess uncertainty more effectively. These advancements will help practitioners utilize empirical Bayes in various scientific and industry contexts without extensive statistical modeling. Additionally, the project will produce a monograph and provide training for students and preceptors on modern data science challenges using empirical Bayes. This research aims to advance empirical Bayes by developing new methodologies and statistical theories to address four key limitations: the assumption of known likelihoods, the treatment of nuisance parameter heterogeneity, the development of nonparametric inference methods, and the integration with machine learning. The project will create new inference methods for primary parameters by using empirical partially Bayes methods and Bayesian nonparametrics, enhancing frequentist Bayes multiple testing theory with guarantees like false discovery rate control. Additionally, the research seeks to learn the unknown Bayes rule through neural networks with a specialized loss function, incorporate James-Stein shrinkage for combining unbiased estimates with semisupervised machine learning, and integrate empirical partially Bayes inference with doubly robust double machine learning for large-scale causal inference. 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 · 2025-06

ABSTRACT Proprioception is the sense that allows animals to monitor their body position and movements; proprioceptive deficits lead to severe challenges in moving and maintaining posture. Proprioceptive feedback passes from sensory neurons through two central nervous system (CNS) pathways: local pathways targeting motor circuits, where proprioception adjusts ongoing movements, and projection pathways to the brain, where proprioception is used to learn and plan future actions. The fundamental differences between projection and local pathways remain largely unknown. Our long-term goal is to understand proprioceptive circuit activity during natural animal behavior, emphasizing the roles played by genetically-defined cell types, which will facilitate the discovery of stem-cell-based therapies for injury and proprioceptive dysfunction. This proposal's objective is to characterize fundamental differences between projection and local pathways. We focus on second-order neurons, CNS neurons that receive direct input from proprioceptive sensory neurons of each pathway. We will specifically address the following questions: does the brain receive minimally processed stimulus information or integrated representations of specific stimulus features? Is information presented to the brain in a behavioral state-dependent manner? Does the brain receive privileged types of information in comparison to local circuitry? We use Drosophila larva as a highly tractable model to study proprioception. This proposal deploys two major technical innovations: CRASH2p microscopy, which allows for volumetric imaging of neural dynamics in intact, freely moving, and behaving larvae, and connectomics, which allows for comprehensive reconstruction of synaptic connections between second-order neurons and their synaptic partners. Based on preliminary data, we will test the central hypothesis that local and projection second-order proprioceptive neurons differentially integrate and process naturally occurring self-movement stimuli. We test this hypothesis using complementary in-depth (functional, Aim 1) and in-breadth (anatomical, Aim 2) experimental strategies. The proposed research is significant because it will provide two advances that remain, to date, out of reach in other models. First, it will provide a comprehensive anatomical understanding of the diversity of second-order proprioceptive neurons and the networks in which they are embedded. Second, it will produce first-of-its-kind knowledge of the activity of second-order proprioceptive neurons in intact animals performing multiple behaviors and determine the role of a specific type of proprioceptor in shaping that activity. Thus, our work is expected to provide a new conceptual framework for how diverse second-order neurons integrate and process proprioceptive information and how the brain senses proprioceptive stimuli.

NIH Research Projects · FY 2025 · 2025-05

Project Abstract: Neural connectivity, the collection of synapses wiring neurons, is a major property of a nervous system, and a determinant of neural function. In humans, billions of neurons make trillions of synapses, and the proper function of this system depends on proper wiring. Incorrect wiring of neurons can lead to improper perception and various neurodevelopmental diseases. While it is generally accepted that the connectivity is determined by cell surface receptors that uniquely label neurons and mechanistically guide their wiring, how these surface receptors signal is mostly uncharacterized. As one of the four classes of major axon guidance cues, UNC- 6/Netrins guide the direction of growth for many groups of neurons as studied in mammals, the fruit fly and the nematode C. elegans. Netrins can provide both attractive and repulsive cues to growing axons, which is dictated by the neuronal receptor: UNC-40/DCC for attractive responses, and UNC-5 for repulsive responses. The repulsive UNC-5 receptor family do not resemble in its domain composition and structure any of the other classes of structurally characterized guidance receptors, including the attractive UNC-40/DCC. Therefore, it is not clear how UNC-6/Netrin complexes with UNC-5 leads to repulsive signals, as opposed to the attractive UNC-6–UNC-40 complexes, despite three decades of genetic studies showing drastic phenotypes of misdirected growth of neurons in unc-5 mutants in several model organisms. The need for a structural understanding of UNC-6–UNC-5 complexes are further exacerbated by the fact that expression of unliganded UNC-5 can induce apoptosis, while UNC-6–UNC-5 complexes lead to over-proliferation of cells and are associated with cancer. To address this mechanistic gap in our understanding, we will develop biochemical strategies to constitute UNC-6–UNC-5 complexes and engineer these complexes through structure-guided rational strategies and directed evolution using yeast surface display. We propose to use the molecular tools and reagents we will develop for biophysical and structural characterization of UNC-5–UNC-6 and UNC-5– UNC-6–UNC-40 complexes, and to systematically characterize UNC-5 function and signaling using genetics in C. elegans and an in vitro cell death model in mammalian cell culture.

NSF Awards · FY 2025 · 2025-05

For centuries, partial differential equations (PDE) have played a fundamental role in understanding physical and natural phenomena. Dispersive/wave equations model wave propagation phenomena which are ubiquitous in nature. They also describe the basic laws of quantum physics, which is one of the greatest achievements of the 20th century. This project studies fundamental questions about dispersive and wave equations by introducing ideas from probability theory. The results of the project will advance the mathematical theory of wave turbulence, which has important applications to plasma physics, nonlinear optics, and oceanography, and the analysis of Gibbs measures for Hamiltonian systems, which plays a key role in quantum field theory and statistical physics. Due to its scope and connections to physics and science, the project will also promote interdisciplinary interactions. As part of the project, the Principal Investigator (PI) is training junior researchers and contributes to maintaining the diversity in STEM disciplines at University of Southern California. This award supports work on five research projects (A-E). The first three projects are concerned with the mathematical theory of wave turbulence. In Project A, the PI extends the short kinetic time derivation of wave kinetic equation to longer kinetic times. This is a major step in the development of the theory, as it goes beyond the perturbative regime and will also shed light on the longstanding open problem of the long-time derivation of the Boltzmann equation. In Project B, the PI plans to generalize this derivation to cover the full range of conjectured scaling laws, which is physically well motivated and also leads to new mathematically interesting structures. New significant combinatorial structures and cancellations which are not present in the physics literature are expected to be discovered. Project C considers the wave turbulence problem for water waves, which has been studied since the 1960s by physicists. Mathematically, it is a quasilinear equation and substantial new ideas are required to obtain results similar to the ones available in the semilinear case. The last two projects concern Gibbs and other invariant measures in statistical physics and quantum field theory. Project D concerns the Gibbs measure for the 2D hyperbolic sine-Gordon equation, which is an important model that contains near-critical scenarios. Here, the goal is to further develop the random tensor theory introduced by the PI in earlier work. Project E investigates, through a combination of techniques from probability theory and integrable systems, the invariance of the white noise measure for the one-dimensional cubic nonlinear Schrödinger equation, which is critical but also completely integrable. 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.