Johns Hopkins University

NSF Awards · FY 2025 · 2025-01

Hard-to-manage outbreaks of infectious disease are emerging more frequently, and the U.S. requires public health systems that protect populations against the adverse social, economic, political, psychological and physical impacts. This project involves planning for a future academic research center that blends the social and biological sciences with community knowledge to understand and to promote epidemic control strategies that affected communities consider appropriate in terms of core values, feasibility, fairness, and health benefits. Public health experts, a range of social scientists, and community stakeholders will together explore the organizing question, “What infrastructure – including financing, organizations, personnel, processes, and practices – supports epidemic interventions that address the social norms, cultural values, language needs, economic realities, local knowledge, and active participation of socially vulnerable populations?” With hard-to-manage outbreaks of infectious disease emerging more frequently, the U.S. requires systems that protect populations against adverse social, economic, political, psychological and physical impacts. This project constitutes planning for an academic research center that partners with community stakeholders to advance the understanding, design, and implementation of human-centered infrastructure at the community/public health interface which supports “socially astute” epidemic management. Socially astute describes those epidemic controls that affected communities consider appropriate in terms of core values, feasibility, evenhandedness, and health benefits. The planning activities posit a synthesis between the social and the biological, between problem-solving experts and those who experience the problem in the contexts of their everyday lives. Combining human-centered systems thinking with a community-based participatory research approach, the project asks, “What infrastructure – including social milieu, financing, organizations, personnel, processes, and practices – underpins epidemic interventions that adequately address the social norms, cultural values, language needs, economic realities, local knowledge, and active participation of socially vulnerable populations?” Objectives for a series of planning workshops and community stakeholder consultations are: (1) develop a systems map for socially astute epidemic management infrastructure, including an over-arching narrative and visual depiction; (2) build a “living document” research agenda to propel the research center; and (3) draft an organizational management plan that addresses the center’s human capital needs, trajectory of fundamental and use-inspired research, operationalization of community collaboration, and the translation of research into curricula, internships, trainings, and briefings. 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 · 2025-01

Project Summary Immune checkpoint inhibitors (ICIs) blocking PD-1, PD-L1, CTLA-4 and LAG-3 are groundbreaking therapies for cancer. Approximately 50% of cancer patients in the US are eligible for ICI therapy. However, ICI treatment can cause various immune-related adverse events (irAEs), which present as inflammation targeting numerous organs including the heart. ICI-associated myocarditis (ICI-myocarditis) occurs in 1-2% of patients treated with ICI therapy and is fatal in up to 50% of cases. More ICI inhibitors are in the testing phase and combining two or more ICI inhibitors has been shown to increase ICI-myocarditis incidence. Thus, there is an exceptionally high need to understand ICI myocarditis pathogenesis in order to aid the development of new targeted therapies, given the high mortality of ICI-myocarditis and its increasing incidence. We generated a clinically relevant mouse model for human ICI-myocarditis by injecting αPD-1 blocking antibody into the immunocompetent A/J mouse strain. Our mouse model showed multi-organ irAEs, with significant inflammation in the heart. Resembling ICI- myocarditis in humans, we observed cardiac troponin elevation, arrhythmia, and massive CD8+T cell infiltration in the heart. We showed that ICI-myocarditis disease was mediated by autoreactive myosin-specific (Myhc+) effector T(eff) cells. We found that Myhc+PD-1+ CD69+ T cells are present in naive mice hearts as well. We further investigated the residency of Myhc+ T cells and found through flow cytometry, intravascular labeling, qPCR, and scRNA sequencing that they upregulated tissue residency markers and resided in the heart as tissue resident memory T (TRM) cells. Upon further characterization, we found that myosin-specific TRM cells were increased in mice with previous cardiac injury, aged mice, and male mice. We observed with 2D imaging that they localize along the perimyocardium and atria. We hypothesize that ICI-myocarditis is driven by Myhc+ PD-1+ autoreactive TRM cells, which increase in number in the heart after cardiac injury, and are also dependent on biological factors such as age and sex. In AIM 1, to investigate how biological variables modulate ICI-myocarditis risk, we will induce ICI-myocarditis in mice that recovered from ischemic reperfusion (I/R) injury or modified experimental autoimmune myocarditis (mEAM). We anticipate that previous cardiac injury, older age or male sex will increase the number of PD-1+ Myhc+ TRM cells in the heart leading to increased vulnerability to ICI-myocarditis. In AIM 2, we will comprehensively characterize the 3D localization and transcriptional profile of cardiac myosin-specific autoreactive TRM cells in naïve mice, mice with previous mEAM and during ICI-myocarditis progression. We will use CODA, a high-dimensional deep learning algorithm that enables the 3D reconstruction of a heart’s cardiac conductive system, immune foci (including the TRM cells), and blood vessels at micron and single-cell resolutions. In AIM 3, we will investigate if cardiac myosin-specific autoreactive TRM cells are necessary for ICI-myocarditis development through a parabiosis surgery mouse model. Lastly, we propose autoreactive TRM cell depletion by neutralizing IL-15, which is essential for TRM maintenance, as a novel method for the prevention of ICI-myocarditis.

NIH Research Projects · FY 2026 · 2025-01

PROJECT SUMMARY Within the stereocilia of hair cells of the inner ear, a variety of molecules including proteins and lipids assemble into a mechanotransduction (MET) complex. This complex converts mechanical stimuli from sound waves into electrical signals that can then be propagated through the auditory pathway. Although many protein elements of the MET complex have been discovered, the lipid elements of this complex are poorly defined. Here, I propose to elucidate the molecular function of PIP2 in hair cell mechanotransduction. PIP2 has been shown to bind to TMIE, an essential subunit of the MET complex of cochlear hair cells that binds to TMC1, the likely pore-forming subunit of the MET channel. In addition, pharmacological depletion of PIP2 in hair cells decreases MET suggesting that the binding of PIP2 to TMIE affects MET channel function. To test this model and to define the mechanisms by which PIP2 affects MET in cochlear hair cells, I propose to combine biochemistry and electrophysiology in order to elucidate how TMIE mutations that perturb PIP2 binding affect MET. I anticipate that my studies will provide insights into the molecular mechanisms by which a phospholipid that is implicated in the regulation of the function of many ion channels contributes to the regulation of the mechanotransduction process in hair cells.

NIH Research Projects · FY 2026 · 2025-01

Adolescent heavy cannabis use increases the risk of impaired cognitive and executive functions. Clinical evidence indicates an adverse effect of chronic cannabis use on impulsivity, which is critically regulated by inhibitory control in executive function. The adverse effects of cannabis are mainly attributed to delta-9- tetrahydrocannabinol (THC), which primarily targets cannabinoid receptor type 1 (CNR1). The potency of THC in cannabis has increased in recent decades, posing a growing concern in society. Importantly, recent studies indicate CNR1 expression not only in neurons and astrocytes but also in microglia actively participating in adolescent maturation of the prefrontal cortex (PFC). Nevertheless, whether and how microglia mediate THC- induced aberrant adolescent PFC maturation, leading to impaired cognitive and executive functions remains poorly understood. This proposal aims to address these knowledge gaps by leveraging our recent work. We have reported that adolescent THC exposure induces amoeboid-like reactive morphological changes in microglia, specifically in the medial PFC (mPFC), which are exacerbated in a mouse model with a genetic risk factor for neuropsychiatric disorders. These effects are mediated by microglial CNR1, reducing the excitability of a particular type of excitatory neurons in mPFC called pyramidal-tract (PT) neurons. However, it is not yet known how adolescent THC exposure affects microglia function at the molecular level, what intercellular mechanisms direct the effects of THC on microglia-mPFC neuron interaction in a neuron subtype-specific manner, and whether hypoexcitability of PT neurons contributes to deficits in long-term cognitive and executive functions induced by adolescent THC exposure. Previous studies suggest that up-regulation of the cyclooxygenase-1 (COX-1)-prostaglandin E2 (PGE2) pathway in microglia is involved in reduced neuronal excitability. The mPFC PT neurons mediate response inhibition, an essential element of inhibitory control. Our preliminary data reveal that adolescent THC exposure induces up-regulation of the COX-1-PGE2 pathway in mPFC microglia, reactive morphological changes of mPFC microglia, reduced neuronal excitability of PT neurons, increased impulsivity, and deficits in response inhibition. Based on these findings, we hypothesize that adolescent THC exposure up-regulates the COX-1-PGE2 pathway in microglia via CNR1-mediated mechanisms, and that PGE2 secreted from microglia acts on neuronal PGE2 receptors and alters PT neuron maturation, thereby reducing their neuronal excitability. These mechanisms may be adolescence- and neuron subtype-specific, leading to deficits in response inhibition. To address these hypotheses, we will identify the time course of CNR1-mediated molecular and morphological changes in microglia induced by adolescent THC exposure. We will also identify the microglia-mediated mechanisms underlying adolescent THC-induced aberrant mPFC neuronal maturation. Finally, we will identify the specific neuronal circuits of mPFC PT neurons contributing to increased impulsivity and deficits in response inhibition induced by adolescent THC exposure.

NIH Research Projects · FY 2026 · 2025-01

Project Summary Hair cells are mechanosensory receptors used in the lateral line systems of aquatic vertebrates and in the auditory and vestibular organs of all vertebrates. Only hair cells in zebrafish and in other non-mammalian vertebrates can regenerate completely after being destroyed or damaged by acoustic or chemical exposure. While in mammals, destroying or damaging hair cells results in permanent impairments to hearing and/or balance. To understand how to trigger a regenerative response in humans to repair hearing loss due to the loss of hair cells in the inner ear, it is essential to understand how genes respond to injury and how those responses are controlled in the genome. My research program is focused on building a complete understanding of hair cell regeneration by integrating genomics, high throughput genetics, and phenotypic analysis to investigate gene regulatory networks for hair cell regeneration using the zebrafish as a model system. We previously performed a targeted ablation of the hair cells in adult zebrafish auditory and vestibular organs and interrogated the epigenomic and transcriptomic landscape of regenerating adult inner ear sensory epithelia. Using single-cell analyses, we showed that the support cell population transitions to an intermediate, “progenitor” cell state that become new hair cells and demonstrate that the cell fate decisions may be driven by the coordinate regulation and spatial co-binding of Sox and Six transcription factors. By functionally validating a predicted regeneration responsive enhancer upstream of sox2, we show that precise timing of sox2 expression is critical for hair cell regeneration in zebrafish. We aim to use these findings to gain insight into the molecular mechanisms that regulate hair cell regeneration in zebrafish. To do this, we propose to elucidate how Sox and Six transcription factors control hair cell regeneration. We will examine their spatial and temporal patterns of expression in the regenerating adult inner ear and conduct loss of function studies to assess their roles in hair cell regeneration. Based on chromatin accessibility data on regenerating inner ear tissues at single cell resolution, we have predictions where Sox and Six transcription factors bind. We propose to identify and characterize their direct targets during regeneration using transgenic reporters, CUT&RUN, and CRISPR/Cas9 disruption. We also have a comprehensive dataset of all the enhancer loci activated during hair cell regeneration at cellular resolution. A major component of our research program is identifying new regulators of hair cell regeneration using a mass mutagenesis approach to identify essential enhancers with roles in hair cell regeneration in vivo. These studies will identify a mechanistic understanding of Sox and Six transcription factors in regenerating auditory and vestibular epithelia. We predict that these studies will illustrate the gene regulatory networks involved in hair cell regeneration, which has major implications in discovering therapies to trigger a regenerative response in humans.

NIH Research Projects · FY 2026 · 2025-01

The development and progression of calcific aortic valve disease (CAVD) is a complex and multifactorial process principally characterized by aortic valve calcification (AVC) that ultimately progresses to severe aortic stenosis (AS). The prevalence of AVC and severe AS has the highest prevalence among persons ≥75 years old. Currently, the only treatment for severe AS is aortic valve replacement (AVR) and as the US life expectancy continues to increase, the number of persons needing AVR is expected to double between 2025 and 2050 to 1.4 million. Recent provocative subgroup analyses from the Simvastatin and Ezetimibe in Aortic Stenosis (SEAS) trial and FOURIER trial testing the PCSK9 inhibitor evolocumab showed a 35-60% lower rate of severe AS with randomization to active treatment. These results suggest that it may be possible to prevent severe AS in appropriately selected patients and/or using novel treatment strategies. However, there is no risk- score/algorithm to identify older persons at high risk for severe AS. Our pioneering work in the MESA cohort of primarily middle-aged patients has demonstrated an extremely strong association between AVC measured using non-contrast cardiac CT and severe AS, which is even stronger than the association between coronary artery calcium (CAC) and coronary heart disease. Additionally, we have shown that persons with a high hsCRP and lipoprotein(a) have a significantly higher risk of developing incident AVC. The 2011 NIH Working Group on CAVD highlighted crucial areas for research including: 1) identifying risk factors for persons at high-risk for severe AS, 2) determining if AVC measurement may identify patients most likely to benefit from earlier intervention, & 3) determining if there are interactions between risk factors for AS. However, 13 years later, little is known about the potential synergy between acute phase reactants, inflammatory markers, lipoprotein(a), and other key traditional risk factors in the risk for AVC and severe AS. Furthermore, the distribution of AVC among older persons and its associated 5-10 year risk for severe AS among older persons is incompletely described. The population percentiles for AVC among older persons, a crucial reference for the clinical interpretation of AVC, is also incompletely described. We propose to use the previously quantified AVC and meticulously phenotyped risk factor data from the ARIC and MESA cohorts in which cardiac CT was performed among 3,242 participants age ≥75 years old to examine the middle-age risk factors for the long-term development of AVC, create AVC population percentiles, and adjudicate incident cases of severe AS to identify older persons at greatest risk for progression to severe AS. These results will have a direct impact on: 1) elucidating the long-term risk factors for the development of AVC, 2) clinical reporting of AVC results in this older age group with the highest prevalence of AVC, and 3) to help efficiently plan clinical trial enrollment of persons most likely to benefit from new treatment therapies. 1

NIH Research Projects · FY 2025 · 2025-01

PROJECT SUMMARY This research is motivated by the goal of improving the lives of individuals with sensorimotor deficits through somatosensory feedback. Recent neuroprostheses development has demonstrated providing sensory feedback improves the control of neuroprosthetic devices. The technological and clinical development is approaching long-term investigations for ultimate translation to independent home use of a neuroprosthesis. As such, a growing number of studies suggest that the projected fields (PFs), the locations of perceived sensations through stimulation, are relatively consistent over the study period. However, there has been little focus on determining the spatial stability of the PFs. Spatial stability is important because knowing where percepts occur will be critical for designing somatosensory neuroprostheses that can be used in daily life without the need for regular recalibration of the stimulation. My goal is to develop a quantitative approach to determine the spatial stability of sensory stimulation PFs. By establishing an understanding of how PFs are spatially preserved over time, I envisage the quantitative framework to bridge the intuition of PF stability and the experimental data collected from sensory stimulation experiments. I will leverage PF data obtained from two different electrical stimulation techniques and determine the co-occurrence between PFs, their spatial contiguity, and investigate functionally relevant PFs. Aim 1 will determine the co-occurrence and spatial contiguity of PFs. I will develop new algorithms that take into consideration the overall activation frequency of each PF, its co-occurrence with other PFs, and the spatial contiguity. To determine the statistical significance, I will use a statistical null model that sets a benchmark of “by-chance” level stability. The statistical null model will be determined through binomial process simulations. Aim 2 will determine the spatial stability of functionally relevant PFs. Functional relevance is important because a critical goal for somatosensory neuroprostheses is to provide reliable somatosensory feedback and enable independent home use for daily tasks. In this aim, I will determine regions of the hand that are important for different prosthesis grips used in functional tasks. I will then compute the co-occurrence of these functionally relevant regions and determine their spatial stability. Fundamentally, Aim 1 focuses on PF spatial stability from empirical data regardless of sensory stimulation techniques, while Aim 2 focuses on PF spatial stability for functional relevant regions. When completed, this work will be applicable to all somatosensory neuroprostheses development in which ensuring PF spatial stability is crucial to establish long-term somatosensory feedback for independent home use.

NIH Research Projects · FY 2024 · 2025-01

PROJECT SUMMARY Background: In the United States transgender and gender diverse (TGD) communities experience a higher burden of mental health disorders compared to their cis-gender counterparts. Stigma has been shown to be a key factor in mental health disparities in the TGD population, though comparatively little is known about how stigma on the structural level, including laws, policies, social attitudes, and norms, affect the mental health of TGD individuals, especially among those living in rural areas. Given the recent increase in laws and policies that restrict or ban access to gender-affirming care and the NIMH's Strategic Plan Goal 2 highlighting the need for greater understanding of social and environmental risk factors for poor mental health, it is a critical moment to further understand this topic. Study Goals and Aims: The proposed study explores the formation of structural stigma and its relationship with mental health of TGD individuals living in the rural Untied States. Study aims include: 1) Analyze the formative process of state-level policies and laws pertaining to transgender healthcare in a subset of US states; 2) Explore the experiences of structural stigma and its effects on mental health among TGD adults living in the rural US; and 3) Assess relationship between state-level structural stigma and individual-level depressive symptoms. Approach: An explanatory, sequential mixed-methods design will use a combination of publicly available data and survey data from the Rural Engagement and Approaches for LGBTQ+ Mental health (REALM) study, an ongoing NIMH-funded R01 cohort study of LGBTQ+ adults living in the rural US led by scholars at Johns Hopkins University and Emory University. Aim 1 will include a content analysis of legislation to understand the rhetoric and scientific evidence used to support recent state-level laws and policies focused on transgender health. Aim 2 will collect primary qualitative data through repeated in-depth interviews with 25-35 TGD individuals to understand experiences of structural stigma and its effects on mental health. Aim 3 will use structural equation modeling to explore the relationship between state-level structural stigma and individual-level depressive symptoms. Triangulating findings from Aims 1-3 will bring a holistic understanding of the research topic and can be used to inform future mental health policy specifically for TGD individuals. Fellowship Information: The proposed research will serve as doctoral dissertation of Kirsten F. Siebach, PhD student at Johns Hopkins Bloomberg School of Public Health. The training and research will be supported by a tailored mentorship team who, combined, offer expertise in mental health, policy, and relevant methodological techniques. The training plan will prepare Kirsten to become a leading independent researcher in the relationship between the structural environment and mental health of LGBTQ+ populations.

NIH Research Projects · FY 2026 · 2025-01

Project Summary Individuals with deafness and hearing loss face significant communication challenges that impact their social, educational, and professional lives, leading to potential mental health and cognitive issues and increased physical safety risks. In the United States, around 48 million people report experiencing hearing loss, exceeding the combined cases of diabetes and cancer. Cochlear implants have allowed for the restoration of hearing in those suffering from debilitating hearing-related impairments; however, children show higher post- implantation success than adults due to their increased neural plasticity. Prior research from our laboratories suggests adult auditory plasticity can be enhanced by visual deprivation. Circuits involved in sensory gating control likely facilitate this restoration of adult plasticity, as we have found that these changes coincide with the weakening of inhibitory inputs from the thalamic reticular nucleus (TRN) to the primary auditory thalamus (the ventromedial geniculate bodies: MGBv) and the strengthening of inhibitory synaptic transmission originating from cortical parvalbumin interneurons (PV-IN) onto A1 layer 4 (L4) principal neurons. Key circuits involved in gain control originate in layer 6 (L6) and involve corticothalamic projection neurons (L6-CTNs), which receive sensory information from other cortical areas (e.g. vision), in addition to auditory information from MGBv, and send excitatory signals onto the MGBv, TRN, and PV-INs within A1. Thus, I hypothesize that layer 6 corticothalamic projection neurons (L6-CTNs) in A1 play a critical role in coordinating the adult thalamocortical plasticity observed in A1 following visual deprivation and that dark-exposure (DE) will lead to plasticity of circuits involving L6-CTNs in adult mice. I propose to investigate whether DE leads to changes in the strength of L6-CTN outputs to the TRN, L4 principal neurons, and L6 PV-INs (Aim1) as well as changes in the strength of inputs onto L6-CTNs from the MGBv and within A1 (Aim2) to determine how DE alters layer 6 sensory gating circuits in adults. Additionally, I will investigate how acute L6-CTN activity effects feed-forward thalamocortical auditory processing from the MGBv to A1 L4 and whether DE alters these interactions (Aim3).

NIH Research Projects · FY 2026 · 2025-01

Our LONG-TERM GOAL is to test patterned micro-stimulation of intermediate and higher-level visual cortex as a prosthetic strategy for restoring 3D object vision. We think that compositional (parts-based) coding of shape in the ventral visual pathway can be hijacked as an unusually efficient way to generate unlimited shape percepts with a practical number of stimulation sites. Moreover, targeting intermediate/higher-level neurons that explicitly encode 3D shape fragments may be the only way to evoke 3D shape percepts, which normally depend on subtle shading, specularity, texture flow, and stereoscopic cues that would be difficult or impossible to duplicate with the phosphenes produced by micro-stimulation of retina or V1. We will combine array recording, stimulation, and behavior in macaque monkeys, to test the HYPOTHESIS that micro-stimulation-generated activity in recently demonstrated V4 clusters tuned for specific 3D geometric fragments causes perception of those fragments. The causal role of specific part signals in ventral pathway has never been tested. We will use linear probes to measure local V4 multiunit activity at 32 sites during passive fixation. We will analyze tuning for part geometry and spatial position at these V4 sites with our previously published methods. We will then start a 4-alternative 3D shape matching task, in which, for some trials, correct perception of the sample stimulus can partly depend on stimulation of a V4 site that adds signal strength for the geometric part and spatial location that the V4 site encodes. Visibility of the sample stimulus at this location will be variably obscured by dynamic pixel noise. Preliminary behavioral data with no stimulation show that match choice accuracy is a roughly sigmoidal function of signal to noise ratio of the obscured object part, near chance up to ~20% signal and at maximum above ~80% signal. Our EXPECTED RESULT is that stimulation of the V4 site during sample presentation will bias behavioral choices toward stimuli with the encoded geometry at the encoded spatial location. The sensitivity of this approach in dorsal pathway motion studies argues that we will be able to detect even weak causal V4 influences. We also expect to demonstrate that micro-stimulation adds specifically 3D perceptual evidence, by showing the same choice bias when 3D information is needed to choose between stimuli with identical 2D shape. If we do not observe micro-stimulation effects in V4, we will perform experiments in inferotemporal cortex, where multi-part 2D and 3D tuning have been observed, and where micro-stimulation is known to produce observable effects, though not yet for complex, specific shape information. The SIGNIFICANCE of the expected results would be (a) The first causal test of perception produced by compositional, 3D parts-based shape coding, which would have a major impact in resolving the nature of object representation in the brain, and (b) Justification for further exploration of 3D parts-based prosthetic strategies in intermediate/higher-level visual cortex, which could impact clinical approaches to blindness.

NIH Research Projects · FY 2026 · 2025-01

Project Summary Endothelia and epithelia throughout the body rely on macromolecular complexes known as tight junctions to modulate tissue barrier selectivity, proliferation, and polarity. Tight junctions are intercellular adhesion complexes that form strands around the apical side of cell-cell junctions. Tight junctions act as barriers that protect organs from pathogens while simultaneously allowing for the selective passage of nutrients, which establishes the chemical microenvironments of all tissue compartments. Thus, tight junction dysregulation is a hallmark of many human diseases. According to prevailing models, tight junctions assemble around oligomeric strands of claudin integral membrane proteins. The properties and stoichiometry of these strands are then augmented by the integral membrane protein occludin and the peripheral membrane protein ZO-1. To date, no direct observation of the interaction of tight junction subunits has been observed to support these models. This is a major gap in our understanding of how the chemical microenvironments in all tissue systems are established, and a roadblock in therapeutic intervention. In this proposal, using an approach integrating structural biology, cell biology, and physiology, we will uncover the structural determinants of tight junction function through elucidating the mechanisms of claudin oligomerization, claudin barrier function, and how tight junction subunits assemble with claudins to tune tight junction function.

NSF Awards · FY 2025 · 2025-01

Prior research and applications of voice conversion models have raised challenging problems that are both theoretical and use-inspired. Notable challenges include processing emotional speech and speech in noisy environments and generating speech that represents the characteristics and expressiveness of specific speakers such as personality traits, mood, prosody, and emotional state. These challenges are exacerbated by a limited availability of data. Improving such capabilities will have a wide range of social impacts ranging from giving natural voice to patients who have lost it to rendering comprehensible and speaker faithful renderings of old poor quality recordings that have become hard to understand to generating seamless speech translations in real time communications while staying faithful to the voice characteristics of the speaker. To address these challenges, the project proposes to explore and expand theories about speaker identity, emotion, and expressiveness in challenging conditions. Practically, this means studying how factors like background noise, emotions such as stress, cultural differences or other idiosyncratic ways of speaking affect a system’s ability to recognize and render faithfully the speech of a specific individual. This work will enable a second aim of this project which is to create voice technology that can be used for safeguarding ethical and responsible use of voice generation. Sophisticated voice conversion techniques can be used to detect and prevent spoofing and other fraudulent activities and make it challenging for unauthorized users to mimic or imitate target speakers. Besides security and defense other areas that will benefit from this project include security and defense, accessibility and healthcare assistive technologies, medical voice preservation, speech therapy and rehabilitations as well as entertainment and gaming. This award aims to develop novel algorithms utilizing deep learning techniques to advance voice conversion models with the ability to represent faithfully the characteristics and emotional states of individual speech. The project includes the following key areas of research. The first research target is to explore learning speaker identity and emotion representations for robust voice conversion with self-supervision. By investigating joint representations, this project seeks to develop a deeper understanding of how speaker characteristics and emotions can be effectively transformed. The second research target is to investigate voice conversion solutions for challenging conditions such as noisy environment, emotional speakers, and limited training to enhance the expressiveness and naturalness of the converted speech. The third research target is to investigate novel deep learning techniques for the detection of synthetic voices and joint training strategies to further improve voice conversion performance and evaluation. By exploring the synergies between transformation and detection of synthetic voices, this project has the potential to significantly impact society with a) accurate and expressive voice-based applications and b) applying the same techniques to detect when speech is naturally occurring or synthetic for the prevention of spoofing. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-01

Project Summary: The direction-selective (DS) circuits in the retina detect the direction of motion in the visual field. Starburst Amacrine Cells (SACs) are specialized interneurons in DS circuits that respond to increases (ON) or decreases (OFF) in light and asymmetrically synapse onto direction-selective ganglion cells (DSGCs). Very little is known about the development and specification of SACs and their downstream connections with DSGCs and other SACs in the human retina. In this project, I will explore mechanisms of SAC fate specification and characterize the development of SAC connectivity in human retinal organoids. The mechanisms controlling when and how SACs develop in the human retina are poorly understood. In my preliminary studies, I found that SACs arise early during retinal organoid development. In Aim 1, I will establish a timeline of SAC birth using IHC, live imaging, and EdU birthdating in organoids. The transcription factor ISL1 is required for SAC specification in rodents8. To assess the role of ISL1 in human SAC generation, I will differentiate organoids from an ISL1∆ null mutant stem cell line9 and assess the density and timing of SAC generation. SACs are classified as ON or OFF subtypes based on their responses to light intensity, soma positions, and gene expression. In Aim 2, I will establish a timeline of SAC subtype generation in human fetal retinas and organoids using IHC, RNA FISH and live imaging. The transcription factor FEZF1 regulates ON SAC cell fate in mice. Preliminary data shows that FEZF1 is expressed in a subset of SACs in the human fetal retina. I will test the role of FEZF1 in the specification of ON SACs by generating and analyzing a FEZF1∆ null mutant stem cell line. SACs synapse onto other SACs and DSGCs in DS circuits. While the identification of SAC synaptic partners has not been established in humans, my preliminary data show that SACs form close associations with retinal ganglion cells. In Aim 3, I will differentiate organoids carrying a VACHT-Cre/tdTomato reporter and infect with a virus harboring a WGA-GFP monosynaptic tracer. In this method, the presynaptic SAC will express tdTomato and GFP, and the postsynaptic cell will express GFP only. To examine homotypic SAC-SAC connections, I will use a chimeric organoid strategy. Successful completion of this project will identify mechanisms of SAC specification and DS circuitry in the human retina, providing insights will inform the development of therapies to treat vision disorders.

NIH Research Projects · FY 2025 · 2025-01

NIH grant application: Project summary/Abstract (30lines of text) Physicians are always out of time. Burnout is a key risk in medical practice today. For medieval physicians, this fact was as old as the profession: Hippocrates famously encapsulated this complaint in his Aphorisms: “The art is long, and life is short.” There was truth to this aphorism. Studying medicine was a long and arduous process; prospects of financial gain were uncertain, and practice meant seeing many patients in little time. This grant will support writing a book under contract with Johns Hopkins University Press. The book investigates the history of time as a medical category: what it meant in medical thought and practice, in understanding bodies and diseases, in professional development, and in medical ethics, and how it interacted with the religious, cultural, and social meanings of time. Medieval physicians and patients contended with acute conditions and emergencies, chronic conditions, and disabilities, and disease progress, which was thought to be influenced by astrological conditions. Time was also key to understanding the human body, which was influenced by seasons and changed by age. The proposed book investigates the production of time as a conceptual and practical category in medieval Islamic medicine. “Time” here is seen here as a cultural category understood at the intersection of the theoretical and practical and as a locus of meaning-making. It also acquires additional texture within non- medical cultural constructions. In this context, “medical time” encountered the five daily prayers, the religious feasts, and the lunar and solar calendars. In short, this book aims to investigate the thick web of meanings that made “time” a sensible and comprehensible, if not always coherent, category. The book starts with an analysis of how time, manifesting in the passage of seasons and other astronomical phenomena, was seen as a medical category. I then investigate how time manifests on the body, considering age and age-related conditions. Time was also key in understanding sex and gender. Female bodies were thought to develop in utero and reach puberty later than male bodies, and female bodies were medically understood in relation to fertility and menstruation. I then investigate time in the patient-physician encounter, diagnosis, and treatment, in the development of diseases and disease categories, and in epidemic conditions. Finally, time is an important ethical category: from the beginning of life to death, to elder care, to the management of a physician’s time. I will look at the place of time in the making of Islamic medical ethics. The book will significantly enhance our understanding of medieval Islamic medicine, have important implications on how we understand medical practice, issues of time management, burnout and other related questions. Moreover, the book will address how these conceptions of time played an important role in the making of medical ethics especially in relation to Muslim physicians and Muslim communities. The book will be an important addition to our understanding of Islam and medicine in the past and how it impacts the present.

NIH Research Projects · FY 2026 · 2025-01

PROJECT SUMMARY/ABSTRACT Dysautonomia, or autonomic nervous system dysfunction, is a common and disabling post-infectious syndrome that can occur following COVID-19 and Lyme disease. Dysautonomia accounts for many of the symptoms in Post-Acute Sequelae of COVID-19 (PASC, also called Long COVID) and Post-Treatment Lyme Disease (PTLD, also called Chronic Lyme). Dysautonomia has a wide variety of manifestations, including POTS (Postural orthostatic tachycardia syndrome), gastrointestinal dysmotility, interstitial cystitis, and neuropathic pain. A small- fiber neuropathy is also often present. The mechanisms of dysautonomia in patients with PASC and PTLD are not well understood. A subset of patients with dysautonomia have ganglionic acetylcholine receptor (gAchR) autoantibodies and often respond to immunomodulatory therapy with intravenous immunoglobulin (IVIG), implicating autoimmune destruction of small nerve fibers as a potential mechanism of dysautonomia. Some patients without gAchR antibodies still respond to IVIG, suggesting that some autoantibodies remain to be discovered. This project will leverage the clinical resources of the Johns Hopkins post-Acute COVID Clinic, the Lyme Disease Research Center, and the POTS Clinic to identify patients with post-infectious dysautonomia. Patients with confirmed PASC and PTLD dysautonomia will prospectively undergo objective autonomic testing in the Autonomic Lab, histopathological examination of small-fiber nerve density on skin biopsy, and clinical phenotyping using patient-reported outcome measures. In Aim 1, we will identify distinct clinical subgroups using unbiased latent variable cluster analysis. In Aim 2, we will determine the clinical significance of small-fiber neuropathy in post-infectious dysautonomia by investigating the association with disease severity, and will correlate clinical outcomes with changes in nerve fiber density over time. In Aim 3, we will perform immunoprecipitation and mass spectrometry to identify novel autoantibodies targeting the sympathetic ganglia in post-infectious dysautonomia. This Award will help the candidate, who is currently an Assistant Professor at Johns Hopkins University, develop her career as an independent physician-scientist with a focus on dysautonomia. Throughout the Award period, she will enhance her clinical research and biostatistical skills through hands-on experience and formal coursework. A key focus of the proposal is for Dr. Adler to refine her skills in autonomic testing and learn how to perform transcranial doppler ultrasound which is currently being integrated into the Autonomic Lab and will be a key skill that she will utilize throughout her research career. She has assembled an exceptional mentorship team that each provides complementary skills to ensure the success of this project, and includes experts in autonomic neuroscience and peripheral neuropathies, PASC and PTLD, immunology and autoantibody discovery, and biostatistics. With the guidance of her mentorship team, the candidate will develop an independent translational research program and a track-record that will lead to a successful R01 application.

NSF Awards · FY 2025 · 2025-01

We are in the middle of the “second quantum revolution”, with quantum technology poised to deliver unprecedented advances in computing, communication, and sensing. The power of quantum technology in biomedical sciences, so far untapped, will be developed in this study. By leveraging the quantum property, namely electron spins, in engineered proteins proposed in this study, biochemical reactions, and subsequent cell behaviors, can be controlled using magnetic fields. Magnetic fields can alter electron spin states in the engineered proteins. The design of the engineered proteins is inspired by a family of naturally occurring proteins which are sensitive to magnetic fields. This family of magneto-sensitive proteins is known to control biological clocks, DNA damage repair, neuronal activities, and seasonal navigation of migratory birds. The engineered proteins will exhibit graded sensitivity to magnetic fields, where their activities are correlated to the strength of the externally applied magnetic field. As such, biochemical reactions can be dialed down or up by adjusting the field strength. Therefore, this new class of engineered proteins is named “quantum-enabled dial (QED)”. QED will have high impacts in many fields, since QED activity can be tuned on a continuous scale. QED can be excited by light or by chemical energy, is modular and can be designed to control numerous enzymatic activities, and the QED activities can be spatially addressed by patterned magnetic fields. QED activity can be varied over time by varying field strengths over time. Lastly, QED can potentially be useful for quantum computing. Therefore, QEDs will open new avenues in synthetic biology, medicine, and quantum computing. The project team will create immersive research experiences that engage a diverse group of participants and organize regular discussions that address the challenges and solutions to increasing the participation and success of the full spectrum of engineering and science talent in the nation. A new class of synthetic proteins, “quantum-enabled dial (QED)”, will be developed with the aim to control prescribed biochemical reactions thus actuating cellular behaviors. The backbone of QED is a flavoprotein, where the flavin and a neighbor amino acid form a spin-correlated radical pair upon chemical or photonic excitation. Magnetic fields can change the energy landscape within the QED molecule. As a result, the rate of the radical pair decaying to the ground state can be controlled by applying external magnetic fields. The decay process releases energy that can be utilized to drive other biochemical reactions. Two types of QED, QED-up and QED-down, will be developed. Increasing the field strength will increase the rate of the reaction coupled to QED-up in a graded manner, but it will decrease the rate of the reaction coupled to QED-down. The cDNA constructs encoding QED-up and -down will be first designed and then expressed in cells or fabricated tissue samples. QED activities in cells when subjected to different magnetic field strengths will be calibrated to guide future applications. QED can be used to solve grand challenges in synthetic biology, tissue engineering, medicine, and quantum computing. For example, QED-based bespoke molecules can be built in a straightforward fashion. QED can be used as smart drugs. When and where the drug is active can be controlled precisely. QED can also be the basis of hot Q-bits, so that ambient temperature quantum computing can be realized. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-01

The transition to 5G is expected to witness not only an emergence of new applications such as mobile augmented and virtual reality, but also opens up the attack surface to both known, and previously unknown threats. Thus, wireless networks of the future will need better control and management at different temporal and traffic aggregation granularities (e.g., how to allocate spectrum, how to quarantine distributed attacks etc.). This project aims to develop scalable, machine learning based analytics on the data from a large set of geographically distributed wireless core network entities such as base stations. The research will enable new approaches for: (a) compressing the raw data via novel summaries and sketches, that reduce overhead while simultaneously enabling highly accurate scalable analytics (b) scalable yet highly flexible distributed learning approaches that are built upon the emerging federated learning paradigm and (c) flexible allocation of bandwidth to support the control plane analytics that minimizes the impact on the data plane. The proposed research outcomes will be systems, algorithms, and data analytics workflows that will inform the design and management of next generation critical wireless infrastructures. The approaches developed will enable ISPs to better apportion resources and enable better performance for emerging augmented reality applications for societal benefit (e.g., disaster response and management). In addition, the approaches can enable the discovery and profiling of new threats, which will have significant implications on national security. The proposed education activities are expected to provide students with a comprehensive training in networking, security, system building, and data science. Thus, there is significant potential for broader impact in terms of contributions to workforce development in an area of national need. 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 · 2025-01

SUMMARY Our project will develop powerful and user-friendly computational methods to characterize alternative splicing variation in human physiology and disease, at the level of individuals and population. Alternative splicing (AS) is a complex and widespread gene regulatory mechanism in eukaryotic species, with important implications for development and disease. As efforts to determine the role of AS in disease are accelerating, genetic determinants of splicing and splicing related disease are starting to be broadly investigated. Two emerging developments are driving our research goals: increasingly large patient cohorts and population studies generating complex sequencing and clinical data, and the rise in use of third generation sequencing technologies, including Oxford Nanopore and PacBio long reads, and technologies for single cell transcriptomics. Research on these fronts demands new specialized models and tools, however, such tools are scarce at the moment. We will develop a general framework and specialized tools to detect and characterize splicing variation from RNA sequencing data produced by diverse sequencing technologies, to model AS and its regulatory ‘code’ from heterogeneous omics data, and to predict genetic determinants of splicing in disease and in the population. Our first broad research effort will develop a statistical framework along with novel technology-specific models and tools for differential splicing analysis from short RNA-seq, long read and single cell RNA sequencing data, accounting for confounding factors such as age, sex, and clinical and metabolic measurements. Our second research effort will develop complex and accurate deep learning models of AS regulation, in the context of RNA processing pathways and by combining RNA sequencing and other types of omics data. Cross-cutting, we will equip our methods to detect the effects of genetic variation on AS and disease. We will take an innovative approach that focuses on introns rather than full length or locally reconstructed transcripts, which capture alternative splicing extensively, allow discovery of novel variants, and drastically reduce ‘noise’ from biological and sequencing artifacts and from assembly errors. We will validate our methods in silico and in the lab with minigene experiments with aid from collaborators, disseminate them to the community at large through the third party repositories GitHub and Anaconda, and popularize them through a new online Coursera course on practical transcriptomic methods. Our tools will allow biomedical investigators to characterize changes in AS associated with disease, along with the effects of genotype, to identify potential biomarkers or treatment targets, and thus will help advance research to elucidate the role of AS in human health and disease and its translation to precision medicine.

NSF Awards · FY 2025 · 2025-01

Non-Technical Summary: Quantum materials can be used to address profound challenges in information technology, health, and sustainability. With support from the Solid State and Materials Chemistry program in NSF’s Division of Materials Research, Prof. Thomas Kempa and his group at Johns Hopkins University search for new quantum platforms and focus on using few-atom-thick molecular crystal lattices to manipulate the optical emission properties of two-dimensional (2D) semiconductor layers. An important and unique feature of the quantum materials platform is that the molecular crystal lattice can be chemically modified so that the optical properties of the 2D semiconductor can also be tuned. To build this platform, which could potentially function as a quantum light source, the researchers use chemical design strategies and synthesis of molecular crystal lattices called metal-organic frameworks. They then show how to merge them with 2D semiconductors to yield layered heterostructures. These heterostructures represent a large departure from conventional approaches and could provide a route towards unprecedented materials integration, interface design, and property discovery. These efforts contribute to future quantum technologies that enable more sustainable high-performance computing, more powerful biological and chemical sensors, and more secure communications, all of which serve US national interests. Aside from this, the principal investigator spearheads two outreach efforts intended to educate diverse communities on the practical and ethical impacts of quantum information science, and to educate undergraduate and graduate students how to be more effective communicators of quantitative information through visual media (e.g., slides, figures, simulations). Technical Summary: Quantum materials display a host of intriguing phenomena including superconductivity, giant magnetoresistance, spin liquid states, and single photon emission, to name a few. However, despite great progress, much work remains to develop quantum materials that can address profound and growing societal challenges in information technology, health, and sustainability. Prof. Thomas Kempa and his group at Johns Hopkins University present a new material platform on which to elicit and manipulate quantum emission through atomically-precise control of excitons. The material platform is comprised of two-dimensional (2D) metal-organic frameworks (MOFs) layered with 2D transition-metal dichalcogenides (TMDs) creating so-called van der Waals heterostructures. A key feature of the platform is that the 2D MOFs can be chemically tailored to provide periodic potentials with bespoke symmetries, sizes, and strengths. In turn, the interaction of these potentials with a 2D TMD, or other 2D material, can elicit new excitonic phenomena. The research addresses two core objectives focused on (a) chemical design of 2D MOFs with bespoke topology and charge order and (b) assembly of 2D MOFs and 2D TMDs into vdW heterostructures. The proposed vdW heterostructure platform likely represents the first attempt to alter explicitly the exciton landscape within a 2D TMD through a tailored 2D molecular lattice. It also differs markedly from twistronic approaches that elicit moiré potentials, because the strategy Prof. Thomas Kempa and his group use provides for independent, chemically selective control over the potential thereby offering a vast new parameter space for materials integration, interface design, and property discovery. Besides its many intellectual merits, the research has broader impacts by contributing improvements to the performance and sustainability of computing hardware, and by contributing to the education and greater engagement of society with science. The latter impact occurs through two specific outreach efforts: a biannual JHU–Morgan State University “Quantum Computing and Security Workshop” to broaden community awareness of quantum information science, and a freely available training module called “From Data to Impact” that teaches students how to create salient and impactful figures, tables, and 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 2026 · 2025-01

PROJECT SUMMARY Critically ill children undergoing invasive mechanical ventilation (IMV) experience elevated rates of hospital- acquired venous thromboembolism (HA-VTE) but are also at increased bleeding risk. The absence of trial- derived safety data for thromboprophylaxis and prospectively validated pediatric HA-VTE risk models precludes the application of universal thromboprophylaxis in critically ill children. Therefore, the development of risk stratification and predictive models to identify children with the greatest risk of HA-VTE balanced by bleeding risk is vital to inform risk-stratified thromboprophylaxis trials. This proposal is based on a conceptual model that greater IMV exposure intensity impairs functional fibrinolysis and results in a thromboinflammatory each of which confers an increased risk of HA-VTE. Accordingly, the scientific objectives are to determine the impact of IMV intensity on functional fibrinolysis (Aim 1); identify plasma biomarkers of HA-VTE risk (including select thrombo- inflammatory cytokines [Aims 2a] and plasma proteomic phenotypes [Aim 2b]); and derive an enhanced risk prediction model for HA-VTE for children undergoing acute IMV combining fibrinolytic and coagulative function, IMV parameters, thromboinflammatory cytokines, and prothrombotic clinical features (Aim 3). The three independent aims will inform risk-stratified thromboprophylaxis trial design and identify pathophysiological mechanisms that will lead to alternative/adjunctive interventions for this vulnerable population. The K23 training aims are directly aligned with the complementary expertise of the mentors and the scientific objectives, and include: achieving a deeper understanding of the role of the coagulative and fibrinolytic system in acquired prothrombotic states and thrombogenesis (Training Aim 1); acquiring proficiency in the evaluation of thromboinflammatory pathways and biomarker discovery and validation (Training Aim 2); and obtaining certificate and practical training in clinical trials to gather the essential skills and knowledge in risk prediction modeling and the design, conduct and analysis of risk-stratified, biomarker-informed multicenter randomized clinical trials in children (Training Aim 3). The candidate, Dr. Anthony Sochet, Assistant Professor of Anesthesiology and Critical Care Medicine at the Johns Hopkins University (JHU) based on the Johns Hopkins All Children’s Hospital (JHACH) campus, has assembled a team of mentors with internationally recognized clinical and research expertise in thrombosis, biomarker discovery, advanced risk prediction modeling, and clinical trials. His career development will include a Certificate in Clinical Trials through JHU and thrombosis curriculum through the International Society on Thrombosis and Haemostasis. He will be supported by the extensive resources of the Institutes for Clinical and Translational Research at JHACH and JHU. Through this K23 mentored career development award, Dr. Sochet will become an independent physician scientist with expertise in pediatric thrombosis and fibrinolysis, prepared to design, conduct, and analyze multicenter clinical trials focused on risk-stratified HA-VTE prevention strategies in critically ill children.

NSF Awards · FY 2025 · 2025-01

This project investigates how supercoiled DNA impacts transcription, a core process of life that converts genetic information encoded in DNA to RNA. Classical models of transcription regulation depict that proteins such as transcription factors bind specific DNA sequences to turn a gene on or off. Recent studies suggest that the mechanics of DNA itself, such as supercoiling, the over- or under-twisting of a double-stranded DNA, can influence how RNA polymerase (RNAP) transcribes RNA. As such, supercoiling could act as a transcription regulator and play a vital role in modulating gene expression in cells. We will examine this hypothesis by performing imaging experiments both inside and outside cells and developing computational models of the process. The work will provide significant knowledge about a new transcription regulation mechanism independent of transcription factor activity. The interdisciplinary nature of the research will also provide excellent research training and teaching opportunities for undergraduate and graduate students. The project will also support participation of Baltimore city public school children in science camps. The research will (1) characterize how supercoiling impacts transcription kinetics and correlation in vitro using single-molecule imaging of purified protein and DNA constructs; (2) characterize how chromosomal topological domains impact gene expression in vivo using single-molecule fluorescence in situ hybridization (smFISH) and single-molecule gene expression reporters; and (3) model and simulate transcriptional regulation by supercoiling at all molecular and time scales of transcription by combining coarse-grained modeling of DNA base-pairing under torque and data obtained from the experiments. The research is anticipated to yield sensitive transcription kinetics measurements that reveal the temporal dynamics of supercoil-mediated interplay between neighboring RNAP molecules and genes and the impact of chromosomal topological domains on these kinetics and dynamics. The overarching goal is to develop a comprehensive quantitative theoretical framework describing supercoil-mediated transcription regulation, from DNA base-pair energetics to kilobase propagation of supercoils. This collaborative US/France project is jointly supported by the Genetic Mechanisms program in the Division of Molecular and Cellular Biosciences and the Physics of Living Systems program in the Division of Physics at the US National Science Foundation and the French Agence Nationale de la Recherche, where NSF funds the US investigator and ANR funds the partners in France.” This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-01

In many applications, data is expressed in terms of graphs (e.g., molecules or social networks), and those graphs are typically expressed as matrices or lists of edges and nodes for processing purposes. However, the representation of a graph as a matrix or list is not unique, and the non-uniqueness can be expressed as symmetries that machine-learning models on graphs should respect. Namely, the predictions of the model on two matrices that represent the same graph should be consistent. This is a classical idea and the fundamental concept behind the field of geometric deep learning. This project claims that one key factor that determines the computational capabilities of current graph neural networks (GNNs) is the choice of the data representation and its symmetries. Hence, in order to explore new computational capabilities, new ways to express graphs are proposed (called non-canonical representations), with new forms of approximate symmetries (non-canonical symmetries), and corresponding approximately symmetry-preserving GNNs (non-canonical GNNs). The design of these non-canonical GNNs is inspired by applications in social networks, spatiotemporal graphs such as traffic networks, and graphs arising from complex material design. The project includes collaborations with chemical engineers in the application of these models to protein and macromolecule design. In addition, this project will have an impact on many diversity outreach and educational activities, including mentoring Baltimore City public high-school students, and outreach and mentoring efforts targeting groups historically underrepresented in applied mathematics and computer science. The project will explore the development of novel non-canonical representations, symmetries, and GNNs, as well as their applications. The research activities are divided into three main tasks: (1) The design of a non-canonical representation that expresses any graph as a combination of finitely many intersecting cliques or communities. The goal is to define a scalable representation and efficient data-processing algorithms. Applications to very large graphs, like social networks, will be considered. (2) The development of a procedure for fitting coarse piecewise-constant templates to graphs, using the structure of the templates to design non-canonical GNNs that are sensitive to the large-scale structure of the graph. This method is expected to shine in domains where sensitivity to large scale is important, such as classification of proteins or shapes, molecular dynamics, and predictions on traffic and power networks. (3) The design of methods that exploit symmetries of the graph in the spectral domain. The main applications will be molecular dynamics and spatiotemporal graphs. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-01

No-notice and short-notice natural hazards, such as earthquakes and hurricanes, often cause cascading impacts. Timely and accurate modeling and assessment of hazard impacts are critical for rapid disaster response and long-term recovery and for enhancing disaster resilience. Recent developments in remote sensing technologies continue to bring massive data on disaster impacts. However, it remains challenging for end users like emergency management agencies and researchers to directly and accurately decipher the location, severity, and uncertainty of disaster impacts from these datasets, mainly due to the highly complex and uncertain causal dependencies in the cascading physical-digital disaster processes. This project aims to overcome this barrier by pioneering the development of next-generation scalable, efficient, and interpretable probabilistic AI infrastructure to enable joint modeling and assessment of cascading disaster impacts and to advance STEM education and emergency management practice through integrated research, education, training, and outreach framework. It aligns with NSF's commitment to promoting the progress of science and facilitating breakthroughs in AI and resilience. The infrastructure can potentially improve emergency management and enhance community resilience across the U.S. and worldwide by providing near real-time spatial probability estimates of multiple disaster impacts. It can advance the knowledge of complex causal dependencies in compound hazard scenarios. It also may enable the advancement of multi-modal, multi-resolution learning, causal Bayesian networks, causal dependency modeling, and probabilistic inference. Multiple education and outreach activities are integrated into this research, including co-creating computation- and art-infused high school teaching modules with high school art and STEM teachers, engaging diverse students in research activities, improving curriculum design, and convening professional training sessions. These activities will foster interest in computational and interdisciplinary science, raise awareness of natural hazard impacts, promote diversity and inclusion, and facilitate workforce training. The technical objective of this project is to develop a next-generation scalable, efficient, and transparent causality-informed probabilistic AI infrastructure that supports the automatic construction and inference of various complex disaster-type Bayesian networks and thus enables joint modeling and assessment of cascading disaster impacts. This cyberinfrastructure will advance modeling, integration, and inference of complex causal Bayesian networks for dynamic cascading disaster impacts assessment and enhancing disaster resilience by enabling the convergence of geophysical hazard/infrastructure damage models, causal representation, multi-resolution data assimilation, deep probabilistic graphical modeling and inference, and domain-specific probabilistic programming. Specifically, the project introduces the following cyberinfrastructure innovations. First, a novel causal Bayesian network formulation and modeling framework will be developed to allow flexible, interpretable, and structured probabilistic representations of various disasters, environment factors, multi-resolution multi-modal sensing observation, and complex causal dependencies. A novel variational inference framework will be developed to enable fast inference over various causal Bayesian network structures for large-scale cascading disaster impact assessment from multi-modal multi-resolution observations. The framework will be further extended to enable online updating with newly arrived sparse ground truth disaster impact reports. The developed modeling, inference, and preprocessing modules will be evaluated and integrated into an efficient GPU-accelerated probabilistic programming software infrastructure with user instructions made available to the research and emergency response 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 · 2024-12

For those with symptomatic bradyarrhythmia, implantable pacemakers are the only treatment. While effective, the device therapy is burdened with complications directly related to the indwelling hardware such as device malfunction and infection. As an alternative to device pacing, biological pacemakers offer an original approach to cardiac pacing. In this proposed project, we will test the idea that mRNA-based transgene expression suffices to reprogram ventricular myocytes to pacemaker cells in vitro and in situ, thereby creating ventricular pacing in vivo. The concept of this approach is novel in that an inherently short-lived gene delivery method using mRNA can deliver therapeutic modalities that require long-term function. The scientific premise for this idea is supported by our preliminary data that the transgene is an inherently short-lived protein, that ventricular-to-pacemaker cellular reprogramming occurs within the time window of the transgene expression, and that mRNA-based transgene expression creates de novo pacing in vitro and in vivo. The proposed project will translate this concept as a preclinical study by employing complexing synthetic mRNA with lipid nanoparticle (LNP) in a formulation that has already been approved and tested successfully in clinical trials. The objective is to evaluate clinical feasibility and safety of this approach with a goal to identify target clinical populations toward the first-in-human trial.

NIH Research Projects · FY 2026 · 2024-12

PROJECT SUMMARY Premenstrual dysphoric disorder (PMDD) is a reproductive affective disorder with impairing mood symptoms that emerge monthly in the premenstrual (luteal) phase of the menstrual cycle. Excitingly, in recent exploratory analyses by our team, epigenetic biomarkers of postpartum depression also distinguished PMDD cases from controls in the luteal phase of the menstrual cycle. Reproductive affective disorders, including PMDD and postpartum depression, can be conceptualized as disorders of hormone sensitivity - an abnormal brain response to ovarian hormone fluctuations. Epigenetic variations at the TTC9B and HP1BP3 loci identified by Co-Is Payne and Kaminsky were prospectively predictive of postpartum depression risk across multiple studies, with over 80% accuracy. Recently, in a cross-sectional cohort of 50 women with and without PMDD, this postpartum depression epigenetic biomarker linear model distinguished PMDD cases from controls in the luteal phase, suggesting our biomarkers may be markers of sensitivity to reproductive hormone change. The primary aim of this R01 is to explore whether the epigenetic biomarkers may represent a broad marker for hormone sensitivity, by assessing women (controls, PMDD) in both the follicular and luteal phases of their menstrual cycles, using a repeated measures approach instead of cross-sectional. A secondary aim is to examine whether our epigenetic biomarkers differ between women with PMDD who have responded to selective serotonin reuptake inhibitor (SSRI) treatment versus those who have failed SSRIs. SSRIs are the first-line treatment for PMDD, yet 40-50% of PMDD patients do not respond to SSRIs. Preliminary data from PI Hantsoo’s laboratory suggests that this epigenetic biomarker may distinguish between SSRI responders and nonresponders in PMDD. Importantly, SSRI treatment response prediction was validated in a prospective postpartum depression cohort, differentiating SSRI responders from non-responders with an AUC of 0.86 (95% CI: 0.63-1). These data suggest that these epigenetic biomarkers may represent a biology important for SSRI response in reproductive affective disorders, which are understudied and have limited effective treatment options. The proposed study will apply our published postpartum depression biomarker linear model to DNA methylation generated by targeted pyrosequencing at TTC9B and HP1BP3 in a cohort of women with PMDD and controls. We will compare the epigenetic biomarker between controls and PMDD in the follicular and luteal phases. Within the PMDD group, we will compare the biomarker between those who have responded to SSRI treatment and those who have not. Blood will be collected at home by participants using a dried blood spot collection system. DNA will be extracted from blood spots and subjected to sodium bisulfite modification and hybridized to a CLIA certified Illumina MethyEPIC Beadchip. Methylation patterns analyzed by Co-I Kaminsky who will use the published algorithm for postpartum depression to determine if the biomarkers predict PMDD.