PI: Mark Lurie, Professor of Epidemiology

Co-PIs: Thomas Trikalinos, Professor of Health Services, Policy and Practice and of Biostatistics, Director of the Center for Evidence Synthesis in Health; Jason Gantenberg, Research Scientist and Assistant Professor of the Practice of Epidemiology

Indoor spaces provide the physical substrate for biological contagion. The SARS-CoV-2 highlighted the importance of indoor environments as sites of disease spread, receiving attention due to their association with superspreading events as well as increased interest in how air flow dynamics and ventilation influence pathogen spread. However, indoor environments also structure the dynamics of the social and contact networks that develop within. As these networks evolve, they influence outbreak potential and severity. Altering network structure—for instance, by instituting limits on building occupancy, reducing class sizes, or assigning individuals to cohorts—can be a means of preventing outbreaks or of mitigating them once a pathogen has begun to spread. While interest in human mobility spiked during the pandemic as an important data source to inform epidemiologic and mathematical modeling, most data focused on movement at the city scale or higher. Comparatively little characterized mobility and social networks in indoor spaces. In Fall 2023 we conducted the MAPPING@Brown study, which is, to our knowledge, the largest Bluetooth-based study of indoor mobility to date. During the study, we collected over 130 million Bluetooth reports from participants’ smartphones. To make optimal use of these data, we propose to augment our prior calibration studies with additional experiments aimed at developing efficient algorithms for processing Bluetooth signals and improving the accuracy of localization based on these signals. We also plan to test an alternate method for localization (ultra wide band) to investigate its suitability for expanding mobility studies to other congregate settings.

PI: Jennifer Merrill, Associate Professor of Behavioral and Social Sciences

Co-I: Elizabeth Aston, Associate Professor, Department of Behavioral and Social Sciences

Alcohol use among young adults (YA) is a significant public health problem. Low-risk drinking guidelines (LRDGs: limits on number of standard drinks [SDs] one should consume per day/week) can assist people in monitoring consumption. To follow LRDGs, individuals must be aware of them and understand the concept of SDs. Labeling alcohol containers with SDs and LRDGs could increase consumers’ ability to make informed choices about drinking. However, research on effects of SD/LRDG labeling is nearly absent in the US and has yet to approximate influences of repeated exposure to labels on real-world consumption. Our longer-term goal is to build evidence for the impact of SD/LRDG labeling on hazardous drinking. Here, we will assess YA understanding/usage of labeling information and support for inclusion of SD/LRDG labeling on alcohol products, via focus groups with YA who drink (Aim 1). Then, a pilot sample of YA will be randomized to repeatedly (over 8 weeks) receive alcohol packaging label stimuli with either: (a) control: industry standard labeling (% alcohol/volume), (b) addition of SD labeling, or (c) SD+LRDG labeling. To inform a future trial, we will test initial signal for experimental labeling conditions on alcohol demand and knowledge SDs/LRDGs (Aim 2) and feasibility of our study design (Aim 3). Findings regarding consumer sentiment toward SD/LRDG labeling and potential downstream effects on consumption can inform policy maker decisions on whether to mandate such labeling. The proposed work will place Brown at the forefront of research that has direct impacts on policies related to public health.

PI: Nicola Neretti, Associate Professor of Molecular Biology, Cell Biology, and Biochemistry, Associate Director for the Center on the Biology of Aging
Co-PI: Jeffrey Morgan, Professor of Pathology and Laboratory Medicine, Donna Weiss '89 and Jason Weiss Director of the Center for Alternatives to Animals in Testing, Professor of Engineering

In response to the National Institute on Aging’s 2024 workshop on 3D in vitro tissue systems, Dr. Nicola Neretti (MCB) and Dr. Jeffrey Morgan (School of Engineering) are spearheading a collaborative project at Brown University to advance research in the biology of aging. This effort aims to address limitations in traditional animal and 2D cell culture models by focusing on two key areas: cellular senescence and rejuvenation through partial reprogramming. While animal models and standard 2D cultures are foundational to aging research, they fall short in capturing the complex, tissue-specific processes observed in human aging. The development of 3D in vitro systems, such as spheroids, presents a promising solution. These models better replicate in vivo environments by enabling cell interactions and nutrient diffusion, crucial for studying senescence and tissue remodeling. This application aims to test the application of 3D spheroid models in two high-impact areas: investigating cellular senescence and exploring cellular rejuvenation via partial reprogramming.

PI: Carlos Giovanni Silva-Garcia, Assistant Professor of Molecular Biology, Cell Biology, and Biochemistry

Aging is undeniably accompanied by a decline in organismal functions and a host of prominent hallmarks, including epigenetic alterations. These age-associated epigenetic changes, such as DNA methylation, histone modification, chromatin remodeling, RNA modifications, and non-coding RNA regulation, are crucial in regulating aging and contribute significantly to age-related diseases. We recently demonstrated that constitutive activation of the histone acetyltransferases PCAF-1/GCN5/KAT2B is sufficient to promote longevity in wild-type animals. In this seed award, we seek to identify the mechanisms by which this acetyltransferase PCAF-1 mediates longevity. Thus, we will define the gene network changes that regulate histone acetylation-dependent longevity. Our work will also study the specific tissues and developmental time requiring histone acetylation to promote a healthy lifespan. Understanding how the epigenetic landscape modulates aging will provide new avenues for developing strategies to promote healthy aging.

PI: Christopher I. Moore, Associate Director of the Carney Institute of Brain Science, Professor of Neuroscience, Professor of Brain Science

Co-PI: Ahmed Abdelfattah, Robert J. and Nancy D. Carney University Assistant Professor of Brain Science, Assistant Professor of Engineering

Key Personnel: Dr. Kevin Turner, Postdoctoral Researcher, Neuroscience; Dr. Eric Salter, Postdoctoral Researcher, Neuroscience; Akshay Nagar, Doctoral Student, Engineering

In this Seed Award, we build on our recent discovery that the Blood-Brain Barrier (BBB) shows rapid, local moments of permeability during behavior and in response to input from Dopamine neurons (‘Plume Events’). These new BBB dynamic Events are ~2 orders of magnitude faster and more local than currently believed. Here, we test the Hypothesis that Plume Events are triggered by fast and local voltage changes in blood vessels, a kind of ‘action potential’ in the vasculature on millisecond time scales. These fast electrical dynamics can position the BBB to mediate real-time communication between brain and body, and provide focal metabolic supply and/or targeted waste clearance. To test that fast vascular dynamics trigger Plume Events, we will optimize new genetically-encoded voltage indicators for vessels: These methods will serve our Hypothesis testing, and more generally open a new domain of study, a unique view of the fast electric network dynamics of mammalian vessels.

PI: Mandar Naik, Assistant Professor of Molecular Biology, Cell Biology and Biochemistry (Research)

Co-PI: Jonathan Kurtis, Stanley M. Aronson Professor of Pathology and Laboratory Medicine

Plasmodium falciparum PfGARP is an ideal drug target for the development of anti-malarial treatment. It is expressed only during the trophozoite stage of the parasite life cycle and exported to the exofacial surface of trophozoite-infected RBCs. Currently, there is no structural information available on PfGARP, which is predicted to be an intrinsically disordered protein. Similarly, the functional roles of PfGARP have remained poorly understood. It was thought to regulate parasitemia by sensing host proteins through an unknown mechanism. Our preliminary data has identified the elusive role of PfGARP and found that it interacts with a human chemokine leading to robust phase separation, which could be a critical checkpoint during the progression of the parasite lifecycle and a leading cause of sequestration of infected RBCs known for past many years. This seed project will perform structure-function relationship characterization of PfGARP and its interaction with the host chemokine protein. The results from this study will be critical for the ongoing anti-malarial drug development efforts at Brown University.

PI: Mark Johnson, Professor of Biology

Crop reproduction is exquisitely sensitive to high temperature and yields of staples like rice, wheat, corn, and fruit crops like tomato are reduced by even short heat waves. By studying rare cultivars of tomato that set fruit at high temperature, we have established that thermotolerant pollen tube growth is an essential component of crop resilience. The pollen tube is a single cell that delivers sperm to female gametes within the flower by extending over 1000 times its initial width in ~8 hours. Without successful pollen tube growth fertilization fails and the crop is not produced. Our work in this area establishes Brown as a center for development of climate change adaptation strategies at the cellular and molecular level. We are defining the molecular basis for thermotolerant pollen tube growth and the next phase of our work will be to define the gene variants and molecular pathways that can be used to achieve crop resilience. Our work suggests that the cell wall integrity signaling pathway is critical and this seed award will allow us to pursue two new questions that allow us to compete for two new grants: 1) How does the cell wall integrity pathway control thermotolerant pollen tube growth? and 2) How do the floral cells the pollen tube grows through contribute to thermotolerant pollen tube growth? Seed funding will allow us to develop functional spatial transcriptomics to address these questions.

PI: Jung-Eun Lee, Associate Professor of Earth, Environmental, and Planetary Sciences

New England typically experiences steady precipitation year-round, with slightly drier summers. However, our preliminary analysis indicates a significant intensification of summer convective rainfall, beyond what temperature changes alone would predict. This proposal aims to investigate the causes of this rapid increase, using weather station data and high-resolution climate model simulations. We hypothesize that the strength of convection, driven by vertical velocity, is rising due to weakened baroclinicity associated with global warming. New England, situated in a latitude zone experiencing one of the most substantial decreases in temperature gradient and baroclinicity, is particularly affected.

PI: Loukas Gouskos, Assistant Professor of Physics

Co-PIs: Stephen Bach, Assistant Professor of Computer Science; Greg Landsberg, Thomas J. Watson, Sr. Professor of Physics

The Large Hadron Collider (LHC) generates an overwhelming volume of data—far surpassing what tech giants like Amazon and Google manage—-at a rate of around 40 million particle collision events every second, resulting in petabytes of raw data annually. To process this deluge of information, real-time decision-making algorithms must rapidly identify interesting events within microseconds. This SEED project aims to develop real-time selection algorithms using advanced Artificial Intelligence (AI) and Deep Learning (DL) methods to analyse this immense data load more efficiently than ever before. In a unique interdisciplinary collaboration between Brown University's Physics and Computer Science departments, this project will develop AI/DL models specifically optimized for the LHC's Level-1 Trigger (L1T) system. The L1T operates under severe resource constraints, utilizing hardware such as field-programmable gate arrays (FPGAs) to make split-second decisions. These algorithms will significantly improve the detection of rare, beyond Standard Model physics phenomena--potential discoveries that current methods might miss. The innovations developed here will not only push the boundaries of particle physics but also have transformative applications in fields such as astrophysics and medical diagnostics, where rapid, real-time data analysis is critical.

PI: Ugur Cetintemel, Khosrowshahi University Professor of Computer Science

Co-Pl: Zhicheng Jiao, Assistant Professor in Diagnostic Imaging

Key Personnel: Wael Asaad, Professor of Neurosurgery and Neuroscience; Grayson Baird, Associate Professor of Diagnostic Imaging; senior research scientist; Director, Brown Radiology Human Factors Lab; Shane Lee, Assistant Professor in Neurosurgery, Director of Technical Design and Quantitative Applications for the Center for the Applied Neurosciences AI Registry (CANARY)

Clinical data lakes integrate diverse data sources—such as electronic health records, medical imaging, lab results, and physician notes—offering immense potential for personalized medicine and evidence-based care. However, extracting value from these data lakes remains challenging due to their scale, heterogeneity, and the complexity of clinical information. Recent advances in AI, particularly large language models (LLMs), offer new possibilities for querying and interpreting such data, but their inherent unreliability and inefficiency limit adoption in critical healthcare settings.

This project aims to make AI-augmented processing of large clinical datasets both efficient and reliable, analogous to how traditional database systems revolutionized transactional workloads. We propose building a Prompt Execution Engine that enables clinicians to query multi-modal clinical data lakes using natural language. By combining the flexibility of AI-driven interfaces with the robustness of database systems, this system will simplify and scale access to complex healthcare data. A central focus is the integration of reliability guardrails to manage LLM-related errors, hallucinations, and inconsistencies—ensuring safe deployment in sensitive clinical environments.

The system's design and evaluation will be guided by three representative use cases: (1) data de-identification, (2) problem-centric summarization, and (3) imaging-based report generation. This multidisciplinary effort will lay the foundation for an AI-driven healthcare platform at Brown, enhancing its position in clinical informatics by improving access to crucial data insights for better patient outcomes.

PI: Jonghwan Lee, Associate Professor of Engineering, Assistant Professor of Brain Science

Co-I: Ana-Lucia Garcia, Assistant Professor of Neurology

Key Personnel: Liqi Shu, Instructor in Neurology (Research)

We propose developing artificial intelligence (AI) solutions to improve stroke patient care by addressing two key challenges: (1) detecting severe strokes before patients arrive at the hospital and (2) predicting long-term recovery outcomes more accurately. In the first part of our project, we will create AI models that can quickly identify large-vessel occlusion (LVO) strokes, which require triage to a comprehensive stroke center equipped for specialized surgical intervention (e.g., only Rhode Island Hospital in Rhode Island). These AI tools will analyze brain activity using portable EEG devices in ambulances, helping emergency teams make faster decisions about patient care. This could lead to quicker treatments, improving patient outcomes and reducing the risk of disability. The second part focuses on predicting a patient's recovery using brain scans and other medical data. We will develop AI models that analyze these data to provide more precise recovery predictions, offering better clinical guidance and improving resource allocation.This project combines the strengths of engineering, neurology, and AI research at Brown University to address critical needs in stroke care. By improving early detection and outcome prediction, our work could have a significant impact on how stroke patients are treated, ultimately leading to better recovery and quality of life for many individuals globally.

PI: Benjamin McDonald, Assistant Professor of Chemistry

Co-PI: Michelle Dawson, Associate Professor of Molecular Biology, Cell Biology and Biochemistry, Assistant Professor of Engineering

Breast cancer is the most commonly diagnosed cancer in women in the United States. In 2021, more than 250,000 new breast cancer cases were reported in females in the United States, while in 2022, more than 40,000 women died from breast cancer. Outcomes vary dramatically depending on the stage of the disease, with the five-year survival rate for distant metastasized breast cancer reported as 32.4% vs ~99% for localized breast lesions.   Understanding the biology underlying the progression of this disease and the mechanisms for its spread are therefore critically important. While the extracellular environment is recognized as a critical regulator of breast cancer progression and spread, the interaction between its architectural and biochemical properties remains unclear, a critical hindrance to progress in fundamental understanding of these processes. This proposal will therefore develop a cross-disciplinary high throughput screening platform to enable the development of 3D synthetic extracellular matrix systems that accurately recreate the cellular behaviors of healthy and various cancerous mammary tissue states and further enable the independent manipulation of key physical and biochemical properties. This platform will uniquely bridge the critical gap between material development and biological validation, accelerating the research progress in the etiology and treatment of breast cancer.

PI: Mauro Rodriguez, Assistant Professor of Engineering

Co-PI: Patrick Green, Assistant Professor of Ecology, Evolution, and Organismal Biology

Cavitation–the formation and collapse of bubbles resulting from impacts and other high-speed movements–damages materials from boat propellers to human brains. Many marine animals produce and withstand cavitation in their daily lives; by using principles of “biomimicry”, we can inspire the engineering of cavitation-resistant materials that can advance material durability and human health. Species of mantis shrimp (Stomatopoda) compete over territory by exchanging bullet-like strikes from spring-powered appendages onto each other’s armored tailplates, while other species do not compete. Strikes result in dissipated cavitation across the tailplate, with little evidence of external damage. We hypothesize that mantis shrimp species that receive strikes on their tailplates have evolved broad-scale properties that reduce cavitation damage; that is, evolution has engineered cavitation-resistant materials. In particular, given the basic physics of cavitation damage and the broad evolutionary trends in mantis shrimp telson geometry and material properties, our central hypothesis is that telson geometry, the combination of raised carina with troughs between, and material properties, the alternation of stiff and flexible materials along this geometry, reduce the damage potentially caused by cavitation. The central aim of this collaborative proposed work is to test our hypotheses and predictions by harnessing the worldwide diversity of mantis shrimp, combining analyses of telson geometry and material properties, theoretical modeling of cavitation dynamics, and high-speed video based analyses of cavitation.

PI: Ellie Pavlick, Briger Family Distinguished Associate Professor of Computer Science, Associate Professor of Cognitive and Psychological Sciences, Associate Chair of Computer Science

Co-PI: Roman Feiman, Thomas J. and Alice M. Tisch Assistant Professor of Cognitive and Psychological Sciences, Assistant Professor of Linguistics

Do Large Language Models (LLMs) reason like humans do? Recent research, including our own preliminary findings, show that humans and LLMs learn to classify objects according to logically-structured rules (e.g. blue or circle) similarly well and similarly quickly, when given similar evidence. In this proposal, we further investigate whether they represent these rules using the same logical operators, and whether these match the functional specifications of operators in classical formal logic. To do this, we will use cutting edge techniques that can identify sub-circuits within LLMs that perform specific, well-defined computations. These techniques will test whether these circuits represent the logical elements that different rules have in common, and probe whether the logical operators operate the same way independent of their input operands, following the specification of logical operators in formal logic, or whether alternatively, they are content-specific, following research on content effects in human reasoning. We will compare results from these techniques to the results of new experimental studies on human participants’ ability to learn logically-structured rules within the same task setting. These will allow us to test whether humans and LLMs represent the same rules when temporary equivalent alternatives are available, and whether the logical operators both deploy are similarly content dependent, in similar cases.

PI: Jacob Rosenstein, Associate Professor of Engineering

Co-PI: Ian Wong, Associate Professor of Engineering, Associate Professor of Pathology and Laboratory Medicine

Key Personnel: Ritambhara Singh, John E. Savage Assistant Professor of Computer Science and Data Science, Brown University

Key Personnel: Jonathan S. Reichner, Professor of Surgery (Research), Rhode Island Hospital and Brown University

Coordinated motility of human cells is essential to shape organ formation in the embryo, to repair wounds, and to respond to infection. Historically, cellular morphology and dynamics have been imaged using bulky and expensive optical microscopes using fluorescent labeling, which often limits accessibility in the clinic or low resource settings. Moreover, detecting and tracking features in time-lapse images remains labor-intensive and time-consuming. First, we propose to develop a capacitance imaging platform for label-free sensing of mammalian cell cultures at single-cell resolution, combined with deep learning to virtually stain for cell states, biomarkers, and features of interest. Second, we propose to investigate how directed cell migration can be guided by highly localized electric fields. The effect of electric fields on collective cell migration is currently poorly understood due in part to prior work being limited to coarse spatial control with large electrodes. Third, we propose to explore new architectures for mechanically flexible electrochemical imaging and stimulation chips, based on a robust commercially-available thin film transistor (TFT) technology. As a case study, we will investigate how adherent epithelial cells transition to a more motile mesenchymal phenotype, which is relevant for collective migration and wound healing. The preliminary results enabled by this Research Seed Award will make this collaborative team competitive for future projects at the interfaces of semiconductors, biology, and medicine.