Yale University

NSF Awards · FY 2024 · 2024-08

Compounds that contain a transition metal connected to a hydrogen atom, referred to as transition metal hydrides, are critical for the synthesis of fine and commodity chemicals, pharmaceuticals, and precursors to plastics. In this project, Professor Nilay Hazari from Yale University is developing a fundamental understanding of how transition metal hydrides transfer their hydrogen atom to organic compounds, which is a key step in many catalytic cycles. This knowledge is crucial for designing improved and new transition metal hydrides, which will transfer their hydrides faster and with more selectivity to organic compounds. Both experimental and theoretical methods are used to understand the reactivity of transition metal hydrides and one hypothesis being tested is that a model currently used to explain electron transfer reactions can also be used to explain hydride transfer reactions. The research is complemented by Professor Hazari’s involvement in a series of outreach activities related to chemistry generally and catalysis specifically, which cater to students from underrepresented minorities in science. These activities include educating the general public about chemistry through demonstrations, organizing a five-lecture summer program on catalysis for high school students in the New Haven area, and hosting undergraduates to perform research in Professor Hazari’s laboratory. With funding from the Chemical Structure, Dynamics & Mechanisms-B of the Chemistry Division, Professor Nilay Hazari of the Department of Chemistry at Yale University is developing understanding of hydride transfer reactions involving transition metal hydrides. This knowledge provides guidance about the design of catalysts and optimization of reaction conditions for the plethora of reactions that involve hydride transfer, such as carbon dioxide reduction to methanol. Specifically, by measuring the rate of hydride transfer from a metal hydride to an organic acceptor, the kinetic hydricity of a metal hydride is determined. Independently, the thermodynamic hydricity of the same metal hydride is measured using equilibrium exchange reactions or electrochemistry. By correlating thermodynamic and kinetic hydricity, linear free energy relationships are developed. These relationships enable the evaluation of hypotheses relating to whether secondary coordination sphere effects such as hydrogen bonding or electrostatic effects stabilize the transition state for hydride transfer and whether hydride transfer reactions can be modelled using Marcus theory. Density Functional Theory calculations are carried out in parallel with the experimental work and probe the structure and energy of transition states and intermediates. Professor Hazari is actively engaged in outreach programs focused on increasing the representation of students from underrepresented minorities in chemistry. He hosts a public lecture that explains fundamental principles of chemistry through demonstrations, runs a short course on catalysis for high school students, and enables undergraduates from diverse backgrounds to perform research in his laboratory. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-08

Modern artificial intelligence (AI) and machine learning (ML) systems are trained using massive datasets and complex models combined with optimization algorithms. Traditional "greedy" methods, which make incremental improvements at each step, often fall short in both efficiency and adaptability when faced with problems at this scale. This project proposes a novel framework for algorithm design based on the Hamiltonian dynamics, a fundamental concept in physics and mathematics that uses the conservation principles to describe the interaction of multiple objects. Such dynamics appear naturally in many branches of computational sciences but are rarely used as a fundamental principle in algorithm design. Motivated by emerging challenges in ML, this project aims to develop a systematic methodology that leverages Hamiltonian conservation to solve problems in optimization, random sampling, and game theory. This project has the potential to revolutionize our understanding of computational and statistical problems by introducing a new class of algorithmic principles for training modern ML systems. This project will also advance the curricula for algorithms in computer science and electrical engineering, with unique training opportunities for undergraduates and graduate students, the development of open-source software, and a dissemination of ideas via joint workshops. This project will explore a framework called “the LCP scheme”, which stands for Lift, Conserve, and Project. This proceeds by taking a parameterized decision space, appropriately lifting the problem to incorporate additional variables, applying the conservation property of the Hamiltonian dynamics to update the problem state in the augmented parameter space, and finally projecting the state back into the original space. This scheme provides a fresh perspective for analyzing several known algorithms, and developing new ones, in the domains of optimization and random sampling, as well as to understand the behavior of players in multi-agent systems. This project will develop a robust algorithmic complexity theory for implementing the continuous-time Hamiltonian dynamics as discrete-time sequential procedures with an emphasis on the large-scale modern applications. By introducing concepts such as invariance, conservation, and the principle of least action, this project will provide a more nuanced view of the state evolution of computational objects that can help overcome many limitations of the standard algorithm design paradigm. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2024 · 2024-08

PROJECT SUMMARY/ABSTRACT White matter hyperintensity (WMH) seen on MRI of the brain is an important biomarker of elevated risk for stroke and Alzheimer's disease and related dementias. There are qualitative scales to measure WMH severity as well as automated techniques that quantify WMH volumetrically. However, these techniques – manual or automated – were developed for conventional high-field MRI and are not optimized for the unique imaging characteristics of low-field portable MRI (pMRI). The pMRI device costs a fraction of a conventional high-field MRI, is FDA approved, does not require magnetic shielding, can be rolled from room to room, and plugged into a standard wall socket. Taking into account the unique imaging attributes of pMRI, we will create both a qualitative low-field WMH scale that can be used widely and a machine learning enabled quantitative measurement of WMH for more sophisticated applications. To ensure the reliability of these WMH measurement systems, we will enroll 100 participants who will receive both a pMRI and high-field 3T MRI at a single study visit for the purpose of comparing WMH measurements against a gold standard (3T MRI). Using the Delphi method, an expert panel of pMRI researchers will develop the low-field WMH grading scale, iteratively refine it, and validate it within this cohort. Parallel to this, advanced machine learning methodologies will be utilized in this cohort, allowing for precise quantification of WMH volume on pMRI. These advances are possible because our multidisciplinary team has expertise that spans translational vascular research to MRI physics and computational medical imaging. However, our vision transcends merely introducing a novel imaging measurement method; we aspire to make brain health assessments more universally accessible and economically feasible compared to the current hospital-based high-field MRI. Upon validation of these WMH quantification methodologies, we will immediately make them available in our popular software package, FreeSurfer (over 60,000 worldwide licenses), and implement them in our ongoing pMRI-based assessments of brain health in real-world settings including a safety net emergency department and at a health center providing primary care to underserved and understudied communities. Combining pMRI innovation with rigorous measurement techniques, we aspire to widen the reach of brain health evaluations, encompassing a more diverse beneficiary group. By permitting early detection of WMH in a range of healthcare or community settings, our project holds the potential to refine intervention strategies, particularly for debilitating conditions like stroke and Alzheimer's disease and related dementias.

NSF Awards · FY 2024 · 2024-08

Generative machine learning models are currently revolutionizing the artificial intelligence (AI) community with their significant capabilities in creating innovative images and text. At its heart, generative AI fundamentally addresses a high dimensional density estimation problem. Alternatively, it can be perceived as a transport problem, transforming a simple and known distribution/noise into a complex and unknown distribution. Despite the development of numerous successful algorithms, the literature lacks statistical guarantees to theoretically underpin these algorithms, and concerns about the environmental impact due to extensive computations continue to persist. Among these models, score-based diffusion models are currently replacing the generative adversarial neural nets and at the forefront in terms of popularity and efficacy. However, the score training process can be exceedingly slow and energy intensive. To address this, the investigator will study the more computationally and energetically efficient rectified flow algorithm and its variants which turn the high-dimensional density estimation to an iterative regression problem, and this iterative regression leads to an optimal transport. The research will advance the understanding of the success of models in generative AI. The intrinsic connections to be explored among those models will help convert statistical guarantees from one generative model to another, and lead to novel and improved algorithms, which would eventually advance the state of art of generative AI. The investigator aims to study the statistical and computational assurances of rectified flow and diffusion models and explore two connections among those models: score matching and solving ordinary/stochastic differential equation, with an intriguing linkage to nonparametric empirical Bayes. The following questions will be addressed: 1) can we show that the iterative rectified flow obtains density and transport estimation optimally in just one step of regression? 2) how fast does the iterative regression of rectified flow converge to the optimal transport? 3) can we propose an improved algorithm over the rectified flow for a better statistical and computational guarantee? 4) what are the statistical and computational guarantees of diffusion models? 5) can we improve the denoising diffusion probabilistic models by an iterative algorithm to obtain the optimal transport? In addition, the project will explore applications of generative models and optimal transport to neuroscience and autism spectrum disorder. Research results from this proposal will be disseminated through articles, workshops, and interdisciplinary seminar series. It will integrate research and education by teaching monograph courses and organizing workshops and seminars to enhance the career development of the next generations of statisticians and data scientists, including a particular focus on the underrepresented groups in mathematical sciences. 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 · 2024-08

ABSTRACT Major depressive disorder (MDD) and anxiety disorders (ANX) are genetically related traits that cause very substantial morbidity and mortality worldwide. The large, diverse, and actively expanding Veterans Affairs (VA) Million Veteran Program (MVP) is an ideal setting for study of these problems. We have identified risk loci for MDD and ANX in large genomewide association studies (GWAS), but the size and power of the MVP dataset, and other datasets worldwide, continue to grow, and considerable work still needs to be done to develop and maximize the scientific value of the MVP for these traits, and in relation to other datasets available worldwide. The MVP sample (currently at >900,000 participants) is continuing to expand, and with more subjects there will be increased power to map relevant traits, including constituent traits (i.e., social phobia). Extensive phenotype data are available. MVP data releases for the coming year will include whole genome sequence (WGS) data that will be available for novel analyses; and array data increasing marker density for non-European populations. Together, availability of these data presents an enormous opportunity. We will make our GWAS data available via dbGAP and make further use of it for the present project We will conduct gene-discovery analysis and leverage polygenicity to dissect the predisposition to MDD, ANX (and specific anxiety and anxiety -related disorders and subphenotypes which may have distinct as well as overlapping genetic risk). We will also study medication response genetics and traits phenotypically associated with MDD and ANX in the MVP sample, e.g. cardiovascular disease, using approaches that permit testing of likely causality among these correlated traits (e.g., Mendelian Randomization [MR]). We will use genomic structural equation modeling (gSEM) to investigate the underlying genetic relationships of ANX, MDD, and other related psychiatric and medical traits. We will complete polygenic risk score (PRS) analysis within MVP and with respect to external samples using MVP data. WGS data will facilitate study of rare and structural variants. As the VA has electronic health record (EHR) data going back to the 1990’s and the MVP has EHR linkage, longitudinal and repeated measures can be constructed. We will replicate findings in collaboration with other consortia such as the Psychiatric Genomics Consortium MDD (PGC-MDD) and anxiety (PGC-ANX) working groups (we participate presently). Considering the MVP and other samples we will be able to access, we predict that case numbers of both traits will increase 2-3 fold over the next five years, greatly increasing power for locus discovery and thus for post-GWAS analyses. Finally, we will investigate these traits in non-European populations including African Americans and Latin Americans – both particularly well-represented in the MVP sample -- and other populations. Increased understanding of the underlying genetics of these traits should ultimately lead to repositioning of currently-approved medications and identification of novel treatment targets.

NSF Awards · FY 2024 · 2024-08

The PI studies problems of geometry using methods imported from particle physics. This project is divided into two major parts. In the first part, the PI aims to prove new formulas for solutions of an equation first studied by Einstein, describing the geomery and curvature of space-time. The second part concerns the physics of two-dimensional systems which have "scale invariance," meaning they look the same both at short and long distances. The PI aims to study these systems by reducing them to simpler ones which can be solved exactly. The results of this work will be disseminated broadly both in the mathematics and high-energy physics communities, helping to bring these two areas closer together. The PI will continue outreach through expository lectures and articles. The project will also contribute to the training of graduate students in mathematics. In joint work with Davide Gaiotto and Greg Moore, the PI introduced a conjectural picture of the hyperkahler geometry of moduli spaces of Higgs bundles. Parts of the conjecture have been verified over the last several years, in the work of various authors, including the PI. Building on this recent progress, the PI will make a direct attack on proving the conjecture, as well as a detailed numerical study of the hyperkahler metric in a particular example. The PI will also employ a new approach to the construction of conformal blocks for the Virasoro vertex algebra and more generally the W(gl(N)) vertex algebras. Conformal blocks are much-studied functions arising in two-dimensional conformal field theory, which are notoriously difficult to describe in explicit terms. The new approach the PI will use involves a new technique of abelianization, which relates complicated vertex algebras to simpler ones. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2024-08

Project Summary/Abstract HIV remains a major global health issue; strict adherence to combined antiretroviral therapy can reduce a patient's active virus to undetectable levels, but the presence of latent viral reservoirs necessitates lifelong adherence to antiretrovirals. HIV-1 reverse transcriptase (RT) is a major target in the treatment of HIV as it is responsible for producing a double stranded DNA copy of the viral genome, which can be integrated into the host. RT inhibitors fall into two major classes: nucleoside RT inhibitors (NRTIs) which are incorporated into the growing DNA chain but usually lack the 3' hydroxyl group needed to continue reverse transcription, and non- nucleoside RT inhibitors (NNRTIs) which bind to an allosteric pocket 10Å away from the active site, causing a conformational change which alters the rate of chemical catalysis. These two classes of inhibitors are typically combined when treating patients to provide more protection against inevitable drug-resistant mutants. In cells, NNRTIs have shown synergy with NRTIs, but some in vitro experiments suggest that NRTI incorporation and NNRTI binding are mutually exclusive. The molecular mechanism which underlies these drug interactions remains unclear. Additionally, recent studies have revealed that a subset of NNRTIs can enhance RT homodimerization and induce HIV-specific pyroptosis by prematurely activating the viral protease. This subset of compounds may be promising components of new “shock and kill” therapies, which aim to cure HIV by activating viral reservoirs and killing HIV-infected cells. While these dual function NNRTIs would represent an exciting advancement in HIV treatment, the mechanistic and structural features of this process are not yet understood. The proposed project will provide insights into the allosteric interactions underlying these two key processes using complementary kinetic, structural, and dynamic methodologies. In Aim 1, I will investigate the mechanism for crosstalk between the active site and the NNRTI-binding site. To do this, I will use transient kinetic analyses to determine whether incoming nucleotides can be incorporated in the presence of a covalent NNRTI, which cannot be displaced. Additionally, I will identify long-range interactions responsible for allosteric inhibition with HDX MS experiments. In Aim 2, I will identify key structural characteristics of dimerizing NNRTIs and the underlying mechanism by which they mediate this dimerization effect. To do this, I will use complementary X-ray crystallography and cryoEM techniques to determine the structures of NNRTIs bound to the p66/p66 homodimer and p66 monomer and identify the interactions between dimerizing NNRTIs and their binding site(s) which are responsible for this effect. I will also monitor the progression from p66 monomer to p66 homodimer in the presence of dimerizing NNRTIs by HDX MS to determine the mechanism by which they enhance dimerization. The insights gleaned from the proposed experiments will lay the foundation for the development of novel NNRTIs which have better synergy with NRTIs and are more potent enhancers of RT homodimerization.

NIH Research Projects · FY 2025 · 2024-08

ABSTRACT Approximately 60% of patients with obsessive-compulsive disorder (OCD) respond to the psychotherapy of choice, exposure and response prevention (ERP). Unfortunately, troubling symptoms persist for most patients, even treatment responders. ERP is believed to depend on safety learning and engagement of related brain circuitry. OCD patients have repeatedly demonstrated deficits in safety learning and memory processes, but important gaps exist in this literature. In particular, there are limited data on neural functions related to safety learning and memory deficits in OCD. This is important as multiple studies have shown that OCD is associated with dysconnectivity in large-scale networks that are implicated in fear signaling and to safety learning, namely, the default mode network (DMN) and salience network (SN). Preliminary research by the PI suggests that that anodal (excitatory) frontopolar (over the anterior medial prefrontal cortex [mPFC]) multifocal transcranial direct current stimulation (tDCS) reduced resting functional connectivity between these networks and accelerates safety learning in community volunteers. Preliminary research by the PI also suggests that frontopolar tDCS accelerates safety learning during a single session of in vivo ERP with patients diagnosed with OCD. The proposed R01 project would be the first to examine if OCD is associated with deficits in inhibitory safety learning – fear extinction learning that is acquired after original fear conditioning has been consolidated into long-term memory – and would be among the first to probe neural functions associated with extinction deficits in OCD. The proposed R01 project would also replicate and extend the PI’s exciting preliminary findings to test if frontopolar tDCS can normalize dysconnectivity and functional activity within and between the DMN and SN and recover safety learning deficits in patients with OCD. Results from this project would pave the way for additional experimental therapeutics target engagement research; research that examine the effects of frontopolar tDCS on neural abnormalities and safety learning deficits in other anxious psychopathology (e.g., posttraumatic stress disorder) and clinical trials examining the interactive effects of frontopolar tDCS and ERP on brain dysconnectivity and symptoms of OCD.

NIH Research Projects · FY 2025 · 2024-08

PROJECT SUMMARY Acanthamoeba castellanii is a free-living amoeba (FLA) that causes usually fatal central nervous system infections like granulomatous amebic encephalitis. FLAs like A. castellanii are bacterivorous and co-exist with Pseudomonas aeruginosa. P. aeruginosa are ubiquitous environmental Gram-negative bacteria. Given that A. castellanii creates an evolutionary pressure for P. aeruginosa to defend itself against amoebal grazing, we hypothesized that P. aeruginosa may produce one or more secreted amoebicidal compounds that can be harnessed to treat amoeba infections in humans. We have since demonstrated that the cell-free supernatants of P. aeruginosa strain PA14 are lethal to A. castellanii trophozoites. Our preliminary data has established that trophocidal activity resides in a <3 kDa (“small molecule”) fraction of the cell-free supernatant, which is not toxic to mammalian cells. Importantly, the small molecule fraction prepared from ᐃrhlRᐃrhlI mutant bacteria (PA14ᐃrhlRᐃrhlI), which do not secrete rhamnolipids, retains trophocidal activity. Using a bioactivity guided fractionation pipeline developed by the Crawford lab, I have begun to narrow down fractions which retain activity. I will determine the identity and structure of small molecules present in these active fractions through mass spectrometry and nuclear magnetic resonance spectroscopy. The isolated compounds will be tested for efficacy against FLAs and toxicity against human cells. The PA14ᐃrhlRᐃrhlI small molecule fraction causes rapid A. castellanii cell rounding and detachment (within 40 minutes) and complete amoeba death between 8 and 10 hours. In the second aim of this proposal, I will determine the mechanism of cell death of small molecules produced by P. aeruginosa by assessing essential cellular processes such as cell membrane permeability, changes in mitochondrial activity, and DNA damage. My long-term goal is to understand the key chemical weapons that P. aeruginosa is secreting to kill A. castellanii trophozoites. Knowledge of these natural compounds and their mechanism of action will help us develop novel therapeutic agents for devastating human infections by A. castellanii and other free-living amoebae. This project encompasses a broad interdisciplinary training in not only microbiology and chemistry, but also molecular biology. My training plan entails immersing myself in two distinct but intertwined fields, rigorous integrative coursework, multiple opportunities for mentoring and developing scientific communication skills. This training plan will provide me with a supportive, collaborative, and gratifying environment that will allow me to develop and succeed as an emerging scientist in field.

NIH Research Projects · FY 2025 · 2024-08

SUMMARY Undiagnosed disease remains a major burden for millions of Americans and their families. Here, we propose the Yale Diagnostic Center of Excellence (YDCoE) as a new site for the Undiagnosed Diseases Network (UDN). The YDCoE builds on the outstanding track record of human genetics, next generation sequencing (NGS) and rare disease discovery at Yale over many decades. To expand access to the UDN, we will partner with multiple community healthcare organizations that provide services for poor individuals who are under/uninsured, all of whom see >90% patients belonging to racial and ethnic minority groups. These partnerships are mutually beneficial, as the YDCoE will provide training and education opportunities in undiagnosed and rare diseases for these organizations. We will work with the DMCC to develop an enrollment plan that utilizes a tiered evaluation strategy. Our highly efficient clinical evaluation approaches and infrastructure will allow us scale clinical capacity to engage more participants over time, a significant portion of which will be covered by community and third-party payer support. For patients who remain without a diagnosis, we will leverage our extensive expertise in NGS approaches, cutting-edge bioinformatics and functional genomics – including in vivo and MPRA approaches - to increase diagnostic yield. Finally, we will conduct additional site-specific studies to enhance clinical evaluation efficiency, enrollment and patient experience. These include novel machine learning approaches to extract clinical information from EHR for more efficient evaluation and diagnoses, an augmented enrollment approach for uniquely vulnerable individuals, and a novel tool to measure patient-reported measures of discrimination in care. We have further secured additional institutional and industry support, which will allow us to prioritize NIH funds for under/uninsured patients. Altogether, our in-house clinical genetics operations together with our extensive network of community and industry partnerships, our expertise in NGS and bioinformatics innovation, our innovative track record in rare disease discovery and the vast clinical knowledge base across multiple specialties at YSM to form a DCoE that is fully equipped to engage a diverse cohort of underserved patients at an unprecedented clinical scale.

NIH Research Projects · FY 2024 · 2024-08

Research Summary: The May 2018 edition of CDC Morbidity and Mortality Weekly Report (MMWR) highlighted the alarming increase in the number of reported vector-borne disease cases in the United States. Between 2004 and 2016, 491,000 (76%) vector-borne diseases were caused by tick-transmitted agents. One such disease is human babesiosis, a potentially fatal and rapidly emerging tick-borne illness reported worldwide and endemic in the United States. Nine species of Babesia distributed into 2 major groups (small and large Babesia) have been linked to infection in humans. Small Babesia species encompass B. microti, which is responsible for most human cases worldwide and B. duncani, which causes fulminant infection that leads to severe and often fatal outcome. Despite the importance of babesiosis in public health, little is known about the biology, pathogenesis, and mechanism of virulence of Babesia parasites. We found that, unlike other Apicomplexa, B. microti and B. duncani employ a novel mode of communication with the host involving two mechanisms of protein export, one vesicular-mediated and another non-vesicular-dependent. The molecular mechanisms underlying vesicular-mediated secretion in Babesia and the role Babesia-derived vesicles (BDVs) play in parasite-host interactions remain to be elucidated. We found that BDVs trigger strong immune modulation of host macrophages. Interestingly, our immunization studies revealed that these vesicles also confer complete protection from subsequent Babesia challenge. The overarching goal of this application is to unravel the molecular mechanisms by which Babesia spp communicate with their host and leverage this knowledge to engineer an effective and safe vaccine for human babesiosis. This goal will be achieved through three specific aims. In Aim 1, we will pursue further cell biological analyses to elucidate the basic properties of the BDVs that confer immune protection. We will employ sub-fractionation analyses to isolate vesicle populations based on size and origin, characterize their ability to confer immune protection, and determine their structural constituents. In Aim 2, we will examine how BDVs elicit cellular and humoral immune responses to gain insights into the mechanism of BDV-mediated immune protection. In Aim 3, we will translate the knowledge about the structure, composition and function of BDVs to develop an effective vaccine for human babesiosis. These studies are novel and are designed to help us better understand how Babesia parasites that infect humans interact with the host, and to guide the development of new diagnostic assays and therapies.

NIH Research Projects · FY 2025 · 2024-07

PROJECT SUMMARY/ABSTRACT Tuberculosis (TB) remains a major public health concern worldwide with more than 1.4 million estimated deaths in 2022. Despite recent declines in global TB incidence, the emergence and spread of drug-resistant Mycobacterium tuberculosis have complicated the control of TB in many settings. Drug-resistant TB is associated with higher mortality and morbidity and requires longer duration of treatment with multiple second- line antibiotics that often have severe side effects. With the widespread adoption of Xpert MTB/RIF (a molecular test for the rapid detection of TB and resistance to rifampicin) over the last 10 years, a growing number of individuals with rifampicin-resistant TB (RR-TB) are being detected and notified in many high- burden settings. To determine an effective treatment regimen for a patient with RR-TB, the selection of antibiotics would ideally be made based on the results of drug susceptibility tests (DSTs). However, because of limited access to DSTs and lengthy delays in receiving DST results, the treatment of RR-TB in most settings remains empiric (i.e., without the results of DSTs) and according to standardized second-line regimens, which are endorsed at the global level. This results in many patients with RR-TB receiving suboptimal treatments, which exposes them to a higher risk of treatment failure, increased toxicity, and the emergence of additional resistance. To mitigate these issues, this project develops a clinical decision support (CDS) tool to optimize medications for individuals with RR-TB, at the point of care, and based on the patient’s basic demographic and clinical information (e.g., age, residence in urban or rural area, and history of TB treatment). The proposed tool combines spatiotemporal machine learning and decision models to synthesize data from clinical trials of anti- TB drugs, local surveillance systems of drug-resistant TB, and studies of cost and loss in quality of life due to illness, treatment toxicity, treatment failure, and emergence of additional resistance. Employing a user- centered design approach with direct input from stakeholders (e.g., TB practicing physicians, health services researchers, laboratory specialist, and policymakers), this project develops a prototype of a user interface for the proposed CDS tool with the potential to be implemented in routine clinical care and a follow-up randomized clinical trials. This project also evaluates the effectiveness and cost-effectiveness of treatment recommendations that are customized according to the local epidemiology of drug-resistant TB and/or according to patients’ basic demographic and clinical information compared with the standardized treatment regimens, which are determined at the global level. The proposed research is significant because it provides TB clinicians in low-resource, high-burden settings with essential evidence and tools to improve the treatment outcomes of patients with RR-TB while containing cost and slowing the spread of drug-resistant TB.

NIH Research Projects · FY 2025 · 2024-07

Our proposal aims to co-design and pilot trial the implementation of a package of brief interventions for suicide prevention among youth (12-24 years of age) delivered by Peers (YPFSuPP). YPF-SuPP will contain two evidenced-based interventions tested in South Asia (safety planning and contact follow up) with two core innovations: safety plans delivered through a personalized youth-designed jewelry and protocolized strategies to culturally and safely engage trusted family members to support youth. We will leverage a nationally renowned nongovernmental organization (Social Changemakers and Innovators – SOCHAI) in Nepal and their well-supported Peer Volunteer workforce to deliver YPF-SuPP alongside the existing health system. We establish an implementation system that provides supportive supervision to Peers. Using mixed methods we will establish two advisory boards, a Youth Advisory Board and a Community Advisory Board of local adolescent mental health clinicians, suicide experts, teachers, family members of youth with lived experience of suicide, and youth advocates. We focus our work on a rural municipality of Makwanpur District, focusing on providing care to a broad population in Nepal. We will identify successes and improvement based on operationalized benchmarks to move on to a fully powered hybrid type 2 clinical trial. Youth clinical outcomes include suicide-related coping, suicide ideation severity, and mental health service use, among other secondary and mechanistic domains. Implementation outcomes include reach, adoption, acceptability, appropriateness, fidelity, and potential for sustainment. This study will lay the groundwork for a future cluster randomized trial of the YPF-SuPP intervention to evaluate its effectiveness in delivering robust, community level, support for youth at risk for suicide.

NIH Research Projects · FY 2025 · 2024-07

This proposed research aims to investigate the emergence and transmission of multidrug resistant tuberculosis (MDR-TB) in the Republic of Moldova, which has one of the most severe MDR-TB epidemics in the world. The study will focus on the transmission of two specific lineages of MDR-TB, the Beijing lineage and the Ural lineage, which we have identified as major drivers of the MDR-TB epidemic in Moldova. Our recent genomic epidemiology research in Moldova has revealed that several large transmission clusters of Beijing-lineage and one large cluster of Ural-lineage MDR Mycobacterium tuberculosis (Mtb) are currently spreading in the country. These lineages have different geographic distributions, with the transmission of multiple Beijing clades focused predominantly within the eastern region of Transnistria, while the Ural clade is more diffusely spread throughout the rest of the country. At this time, we have neither a sufficiently precise estimate of the speed by which each of these clades has grown and spread nor a clear understanding of the distinct genetic mechanisms favoring emergence and spread of these two MDR lineages. The first aim of this proposal is to compare the evolution and expansion of Beijing and Ural Mtb lineages in Moldova since 2010, using isolates and data systematically collected from all culture-positive patients prior to TB treatment. This will allow us to estimate the speed at which these two lineages have become prevalent in the population and to investigate how specific mutations accumulated within these lineages may contribute to their apparent evolutionary success. The second aim is to quantify and compare the transmission of these lineages using data collected during investigations of known contacts of TB patients in Moldova. This will allow us to estimate the relative reproductive potential of these lineages. The final aim is to develop transmission dynamic models to estimate the health impact and costs associated with interventions to interrupt MDR-TB transmission. These models will help to inform public health policy and guide resource allocation for MDR-TB control efforts in Moldova. This research addresses a significant public health crisis of transmitted MDR-TB using innovative phylogenomic, phylogeographic, regression and transmission modeling approaches. Our long-term partnership with investigators in Moldova ensures access to a unique longitudinal collection of Mtb isolates and associated clinical, laboratory and epidemiological data. The ~2000 Mtb whole genome sequences, extensions to analytical tools for inferring transmission and estimating the relationship between bacterial and host factors and transmission, and transmission dynamic models developed during this project will be made freely available to other researchers to ensure maximum use of the data and methods generated during this project.

NIH Research Projects · FY 2025 · 2024-07

Project Summary Engineered T cells are emerging as promising therapeutic agents against a wide variety of cancers. However, despite remarkable success against blood cancers, these cells remain largely ineffective against solid cancers, due to their inability to sustain antitumor activity in response to chronic tumor stimulation, a process termed exhaustion. While antigen stimulation is essential for driving the acquisition of effector functions in T cells, strong and continued stimulation can cause cells to lose effector capabilities and enter an exhausted, dysfunctional state. Limiting the intensity or duration of signaling could enable T cells to sustain effector functions without concomitant exhaustion; indeed recent studies have shown that inhibiting signaling to rest cells from chronic stimulation can enhance T cell persistence and tumor control. However, as some stimulation is needed for effector function, it will be critical to have therapeutic strategies that can reduce signaling activity to appropriate intensities or durations for optimal function. In this proposal, we seek to design, build and test a synthetic feedback controller of T cell signaling, to maintain optimal signaling for prolonging anti-tumor effector functions and mitigating exhaustion. Feedback loops are widely used in engineering to maintain systems at defined set points, and could constitute an effective strategy for tempering excessive signaling in T cells due to chronic stimulation in the tumor microenvironment. To enable feedback circuit design, we will first define the relationships between input signals, pathway output, and downstream gene regulatory and functional responses. These input/output relationships are critical for feedback circuit design, as they reveal how much signaling activity is elicited by different inputs and how much is optimal for sustaining desired function, thereby defining the optimal system set points and feedback strengths. We will measure these input/output relationships in T cells (Aims 1-2), utilizing a dual-pathway reporter system we have developed that enables concurrent live-cell measurements of the activity of two key signaling nodes, Erk and NFAT, in primary mouse T cells. Next, informed by these quantified input/ouput relationships, we will identify genetic components for the actuation in this feedback controller, then proceed to build prototype circuits and test their ability to boost antitumor T cell functions within in a mouse tumor model (Aim 3). If successful, our work will define a new strategy to counter exhaustion for engineered T cell therapies and establish a new paradigm for engineering self-regulating cell therapies that can maintain optimal function through environmental sensing and internal adaptation.

NIH Research Projects · FY 2025 · 2024-07

Abstract/Summary People living with HIV (PLWH) continue to harbor virus-infected cells despite suppression by antiretroviral therapy. These reservoir cells are the source of virus rebound that typically occurs in 1-8 weeks if treatment is discontinued. Eliminating the provirus integrated in the genome of infected cells or the infected cells themselves can cure PLWH. The recently discovered genome editing technique called CRISPR is being developed as an approach to target and excise HIV DNA sequences present in infected cells. Large challenges to using the CRISPR approach for HIV elimination are the wide diversity of viral isolates in PLWH and the risk of genotoxicity as the Cas9 enzyme causes double-stranded breaks (DSBs) in the host chromosome. The absence of a targeted delivery system for cells that make up the HIV reservoir increases these risks due to off-target effects. This R01 application in response to RFA-AI-20-076: “New Technologies for the In vivo Delivery of Gene Therapeutics for an HIV Cure” addresses the challenges posed by CRISPR-based gene therapy. We will use a designer recombinase, Brec1, that has been developed to solely target HIV-1 provirus sequences and remove these from the genome. 90% of the major HIV-1 subtype groups A, B, and C are expected to harbor the precise Brec1 target sequence. Brec1 shows no immunogenic potential, does not induce cellular toxicity or display off- target activity. We will in adition re-engineer Brec1 to generate a super-repressor called Brec1-Off that can durably silence HIV transcription by epigenetically modifying HIV promoter sequences. Key to the success of this approach is our development of capsid-engineered adeno-associated viral (AAV) vectors to deliver Brec1 and Brec1-Off specifically to human T cells and myeloid cells in peripheral tissues and the central nervous system. AAV vectors are one of safest clinical gene therapy vectors that are approved for human use, albeit one obstacle is their unselective delivery. We therefore redirect AAV tropism to the human immune cells in tissue compartments which can harbor HIV-1 proviruses by deleting their natural receptor binding and engineering specific target cell binders to their surface. Thus we pave the way to the development of a viable, non-invasive approach for a gene therapy-based cure for HIV/AIDS. The proposal has two specific aims- In Specific Aim 1, we will develop AAV6 and AAV9 vectors with capsids modified to target human CD4 T cells and myeloid cells and optimize their design for expression and delivery of Brec1/Brec1-Off to these human immune cells as well as thoroughly assay AAV-Brec1 interventions in cell models of HIV-1 infection. In Specific Aim 2, we will test the capsid-engineered AAV vectors packaging Brec1 and Brec1-Off in three different physiologically relevant humanized mouse models for HIV infection to assess their ability to inactivate peripheral and CNS reservoirs of HIV-1 that also contain the diverse HIV isolates found in PLWH.

NIH Research Projects · FY 2025 · 2024-07

Abstract Malfunction of the brain’s waste clearing glymphatic system has been shown to play a key role in cerebral small vessel disease the most common cause of vascular cognitive impairment prevalent with chronic hypertension in the aging population. Cerebral small vessel disease (cSVD) affects 750 million people worldwide, causes up to 45% of all dementias, and accounts for ~20% of all stroke cases, yet the underlying pathophysiology - even in the setting of hypertension - remains incompletely understood. Remarkably, the status of the lymphatic system (LS) has not been investigated in cSVD. We have novel data showing functional impairment of the LS a cSVD rat model with chronic hypertension (spontaneously hypertensive stroke prone (SHRSP) rat strain). Here we posit the entirely novel idea that alterations of the lymphatic system including meningeal lymphatics and lymph nodes draining brain waste in the setting of chronic hypertension contribute to cSVD. Specifically, we will test the main hypotheses that, cSVD with chronic hypertension is associated with uncoupling of the glymphatic and lymphatic systems and that systemic impairment of lymphatic drainage contributes to the pathogenesis of cSVD. We will apply our novel imaging approaches and computational analysis tools to characterize lymphatic drainage concurrently with glymphatic transport in a comprehensive series of experiments involving SHRSP rats with normotensive Wistar Kyoto (WKY) rats serving as controls. In Aim 1, we will determine the role of hypertension on glymphatic-lymphatic coupling. Lymphatic system function will be measured concurrently with glymphatic transport in untreated SHRSP rats and in SHRSP rats treated with antihypertensive medication (amlodipine) to reduce small vessel pathology at the whole-body level. We hypothesize that decreased glymphatic transport in SHRSP rats with chronic hypertension will be associated with glymphatic-lymphatic uncoupling. Furthermore, we expect amlodipine treatment will restore LS function in SHRSP rats. In Aim 2, we hypothesize that the pathogenesis of cSVD with chronic hypertension may be caused by a defect in lymphatic drainage that impairs glymphatic waste clearance over time. This assumption will be tested by gain-of-function (increasing the lymphatic vasculature with Vascular Endothelial Growth Factor C (VEGF-C)) and loss-of-lymphatic drainage (blocking drainage to the dcLN) experiments. In Aim 3 the goal is to generate a molecular and biochemical platform which will drive the understanding of which changes in the lymphatic system are associated with chronic hypertension. We will first conduct a comprehensive “omic” analysis of the CSF collected from WKY and SHRSP rats in Aims 1 & 2, since our scientific premises indicate that CSF is a reliable mirror of the ongoing brain activities. Next, we will move towards the anatomical, cellular and biochemical characterization of the lymphatics draining to the deep cervical node, since preliminary studies indicate that collagen deposits are associated with increased blood pressure. Molecular pathways that we identify can later be probed for therapeutic benefit.

NIH Research Projects · FY 2025 · 2024-07

PROJECT SUMMARY This proposal brings together experts in the genomics of neuropsychiatric disorders and translational neuroscience to investigate the molecular pathology of alcohol use disorder (AUD). Alcohol use is the fourth leading cause of preventable death in the United States that affects over 16 million adults. Understanding the molecular basis for individual differences in susceptibility to alcohol addiction will facilitate the development of diagnostics and personalized therapeutics. Brain molecular phenotypes, such as gene expression and DNA methylation can be translated into neurobiological mechanisms by identifying causal disease genes and pathways. The central goal of this research is to identify novel lasting transcriptomic and epigenetic changes in biological pathways harboring potential biomarkers and therapeutic targets for AUD. Our central hypothesis is that genetic and epigenetic variations lead to brain region-specific dysregulation of transcriptomic networks and disturbances in reward-related genes during AUD. This proposal will perform a postmortem, multi-omic study of brains from 341 total subjects AUD (n=100), a psychiatric control (MDD; N = 132) and neurotypical controls (n=109). To test our hypothesis, we will determine differential DNA methylation and RNA expression across 8 total brain areas of the prefrontal cortex, amygdala, nucleus accumbens and hippocampus. We will analyze RNA-sequencing of 8 discrete regions to identify transcriptional changes in etiologically relevant AUD brain regions (Aim 1). Next, we will identify gene regulatory mechanisms of AUD by integrating whole-genome bisulfite sequencing of matched donors and regions (Aim 2). Finally, we will prioritize AUD genes and test whether genome-wide significant variants identified from GWAS of AUD and alcohol-related phenotypes are associated with gene expression or DNA methylation across brain regions (Aim 3). The expected outcome of this research will be a model of genetic, epigenetic, and transcriptomic profiles of brain regions from cases versus controls that will allow us to understand the region-specific molecular effects of AUD. By identifying new molecular mechanisms of AUD using a large number of donors across many regions, we will be able to explicate the neurobiological mechanisms affected by AUD and identify novel targets for intervention in the addiction process.

NIH Research Projects · FY 2025 · 2024-07

Project Summary/Abstract My career goal is to be a principal investigator at a top-ranked research institution. My scientific goal is to understand how perceptual experience is transformed into adaptive memories, and how these memories impact immediate perception. In this application, I outline an innovative research program to study how the hippocampus, a brain region typically associated with memory function, supports visual perception and attention. This research is motivated by an emerging literature suggesting that the hippocampus is involved in central aspects of vision, such as eye movements, imagery, scene perception, and visual search. However, the underlying principles that govern this contribution are debated. Recent functional magnetic resonance imaging (fMRI) studies have shown that the hippocampus distinguishes visual stimuli through a process known as differentiation. This process leads to the unexpected result that related stimuli are represented less similarly than unrelated stimuli in the hippocampus. This form of relational coding stands in stark contrast to the tuning properties of visual cortex, which represents related visual stimuli more similarly than unrelated stimuli. My research strategy uses behavioral and fMRI methods to test the novel hypothesis that hippocampal differentiation biases visual perception and behavior through the repulsion of attention away from related visual stimuli. Specifically, we test the impact of hippocampal differentiation on the repulsion of object-based attention (Specific Aim 1) and feature-based attention (Specific Aim 2). We implement proven learning protocols to measure hippocampal differentiation with high-resolution fMRI and examine how these unique neural representations relate to behavior in validated attention tasks. In Aim 1, we adapt a shape-based visual statistical learning task from the Sponsor’s lab to drive differentiation, followed by a cueing task that measures object-based attention. We predict that the degree of hippocampal differentiation between related shapes will be associated with the amount of object-based attentional repulsion. In Aim 2, we use a color-based visual associative learning task to drive hippocampal differentiation, followed by a contingent capture task that measures feature-based attention. We predict that the degree hippocampal differentiation between related colors will be associated with the amount of feature-based attentional repulsion. By linking memory systems to visual functions in the human brain, this proposed research program may reveal new sources of attentional control, with significant implications for advancing knowledge of visual deficits and for developing new approaches for visual rehabilitation.

NIH Research Projects · FY 2025 · 2024-07

PROJECT SUMMARY Despite the significant clinical success of AR-targeted therapies in the management of prostate cancer (PCa), resistance to these treatments remains a significant challenge. This resistance often manifests as persistent or even elevated levels of AR and AR signaling. Emerging evidence indicates that the dysregulation of the ubiquitination-based protein degradation process is pivotal in the accumulation of oncogenic proteins like AR, contributing to therapeutic resistance. While the role of various E3 ligases in the degradation of AR and the tumorigenesis of PCa has been extensively studied, there is a crucial yet unaddressed gap in our understanding regarding the specific E2 enzyme responsible for AR degradation. Our research identifies UBE2J1 as the authentic E2 ubiquitin conjugating enzyme accountable for AR ubiquitination in PCa. Preliminary results reveal that the frequent loss of UBE2J1 in 10-15% of PCa patients leads to dysregulated AR ubiquitination and degradation, contributing to its accumulation and resistance to AR-targeted therapy. By utilizing ubiquitination- based AR degraders, we have successfully reinstated AR degradation and impeded the growth of therapy- resistant PCa tumors. In light of these compelling findings, we propose three aims to validate the central hypothesis that UBE2J1 serves as the E2 ubiquitin conjugating enzyme in regulating AR degradation and resistance to AR-targeted therapies. The overarching goal of this study is to unravel the molecular functions of UBE2J1 in governing AR ubiquitination, degradation, and various facets of PCa tumorigenesis and therapy resistance. In Aim 1, we will undertake a comprehensive analysis of the role of UBE2J1 in regulating AR functions across multiple stages of PCa, employing both in vitro and in vivo models. We will utilize single-cell RNA sequencing and spatial transcriptomics to assess the effect of UBE2J1-loss on tumor heterogeneity and resistance at single-cell resolution. In Aim 2, we will comprehensively dissect the molecular mechanisms through which UBE2J1 regulates AR degradation. We will identify the critical domains of UBE2J1, pinpoint the specific interaction and ubiquitination sites on AR, and reveal the E3 ligase responsible for UBE2J1-mediated AR degradation. We will also explore alterations in the AR cistrome and transcriptome landscape in the context of UBE2J1-loss. In Aim 3, we will initially evaluate the in vitro and in vivo effectiveness of restoring AR degradation using PROTAC-based AR degraders AC67 and AC176 in various UBE2J1-KO PCa models. We will subsequently assess the predictive value of UBE2J1 expression as a biomarker for responses to both AR antagonists and AR degraders. Completion of this project will not only substantially refine our comprehension of the molecular mechanisms governing AR degradation and therapy resistance in advanced PCa but pave the way for the development of an impactful predictive biomarker, as well as a new therapeutic approach to combat this lethal complication of modern targeted therapies, thereby improving the clinical outcomes for patients.

NIH Research Projects · FY 2025 · 2024-07

Project Summary Use of Electronic Nicotine Delivery Systems (ENDS) can have positive and negative effects on public health. Ideally, people who smoke would switch to ENDS, and vulnerable populations like youth would avoid ENDS altogether. Yet, most adults who use ENDS also smoke, and ENDS have been the most popular nicotine product among youth since 2014. Effective informational labels for the nicotine strength of ENDS could help to address these problems. For smokers who want to switch to ENDS, easily selecting a product with sufficient nicotine to facilitate a complete switch is critical. For youth, informative labels may deter ENDS uptake, use of high-nicotine ENDS, and development of nicotine dependence. The FDA requires nicotine concentration labels on ENDS, typically in mg/mL or percent, but neither youth nor adults understand these metrics. We previously pilot tested new nicotine concentration labels that improved ENDS users' ability to interpret the strength and addictiveness of nicotine in ENDS. Here, we propose to extend our work to develop two labels: 1) a nicotine concentration label that could replace all current FDA-required ENDS nicotine concentration labels, and 2) a novel, potentially more informative label for nicotine emissions (“flux”) from popular closed-system devices like Vuse and JUUL. In Aim 1, we will refine our piloted nicotine concentration label by re-anchoring the nicotine strength categories on the label (e.g., medium, high) to the nicotine content in cigarettes. We also will develop nicotine flux labels for closed-system ENDS, which have fixed, calculatable nicotine emissions; are the most popular device type; and are the only device type to receive FDA approval for sales in the United States. Flux, the rate of nicotine emissions from ENDS, better approximates nicotine delivery than concentration alone. We will employ qualitative methods (i.e., expert feedback, focus groups) to identify the optimal scaling for our concentration label and to determine key features for inclusion on our flux labels. In Aim 2, we will compare the mg/mL market label (the top-performing market label from our prior work), optimized nicotine concentration label, and flux labels. We will survey 2,000 youth and adults who use ENDS, cigarettes, or no tobacco products to compare the ability of our new labels to accurately convey ENDS nicotine strength relative to current market labeling and to each other. We also will assess perceived addictiveness, perceived label utility, and product use intentions. We expect the new concentration and flux labels to more accurately convey ENDS nicotine strength than mg/mL alone. We also expect the new labels to better convey information about addictiveness, to be rated as having more utility than mg/mL, to decrease use intentions among never users, and to increase intentions to use ENDS in a quit smoking attempt. Project findings will provide the FDA a superior alternative to current nicotine concentration labels for all ENDS and will indicate if nicotine flux labeling has added benefit for the most popular ENDS device type: closed-system ENDS. In sum, the proposed project will increase public understanding of nicotine concentrations of all ENDS and nicotine emissions from closed-system ENDS.

NSF Awards · FY 2024 · 2024-07

Oceans play an important role in the climate system, having already taken up around one-third of anthropogenic carbon released into the atmosphere since the Industrial Revolution. However, as temperatures continue to climb, the extent to which oceans will continue to mitigate rising atmospheric carbon remains to be fully constrained. Yet quantifying this atmospheric carbon sink is critical to projecting the future response of the climate system. To narrow this gap, researchers in this study are investigating changes in carbon uptake and storage in the Pacific Ocean during the Pliocene, an interval of warmth around 3 million years ago that is commonly used as an analog to investigate the response of the climate system to modern warming. Using both data and models, the aim of this study is to quantify the marine carbon response to two specific temperature-sensitive mechanisms within the ocean with the goal of better predicting carbon storage during future warmth. This collaborative project is also advancing public understanding of climate science through the development of a new exhibit for the Central Gallery of the Yale Peabody Museum showcasing how climate signals are measured from fossil plankton in ancient oceans. Additionally, the project is supporting participation in George Mason’s Summer Undergraduate Research Experience (S.U.R.E) Program, doctoral students at Yale and Mason, and engaging high school and undergraduate students in the translation of core science into a publicly accessible display During warm climate conditions, such as the Pliocene, marine carbon cycling was likely affected by changes in circulation and temperature-dependent rates of biological processes. Changes in these levers are predicted to have cascading effects on the relative amount of short- and long-term marine carbon storage, and through subsequent feedbacks, the climate system as a whole. Although both circulation and temperature-dependent biology have been argued to dominate carbon cycle changes in warm climate states, they have yet to be directly compared in state-of-the-art climate models and model-data comparisons. This study addresses this gap using a series of Community Earth System Model experiments designed to examine each lever individually, and in combination, to quantify the associated model-predicted changes in carbon storage. These predictions are also being tested in the Pliocene using geochemical proxy data for ocean pH, dissolved inorganic carbon, and temperature at four Pacific Ocean sites. This study provides a valuable assessment of the potential strength and interaction of circulation and temperature-dependent remineralization on marine carbon cycling and serves as a testbed for how well climate models simulate carbon cycling and other key elements of ocean biogeochemistry. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-07

This award aims to advance our understanding of how soft, deformable particles, such as oil droplets and cells, move through complex environments consisting of narrow constrictions and obstacles. Unlike previous research on rigid particles like sand, this study will explore the behavior of extremely deformable particles that can significantly change their shapes—by more than 50% in many cases. This research has important implications for numerous industrial processes and applications, from improving diagnostic techniques in healthcare (such as cell sorting), to developing better filtration and wastewater treatment devices. By combining carefully designed experiments with advanced computer simulations, the researchers aim to develop predictive models that can accurately describe how deformable particles flow and interact with their surroundings under different driving conditions. This research not only pushes the boundaries of science, but also has the potential to drive innovations leading to improved design and operation of systems that rely on the controlled flow of soft particles, ultimately benefiting national interests such as public health, economic competitiveness, and environmental sustainability. Furthermore, the project includes educational outreach initiatives to mentor students, promote diversity and inclusion in STEM fields, and foster the next generation of scientists and engineers. The project will systematically investigate the flow of bubbles and droplets through obstacle arrays, studying the effect of particle surface tension, as well as the obstacle density, size, and placement, on the particle trajectories. Specifically, this project will investigate whether the flow rate of deformable particles through a narrow orifice can be described by the Beverloo equation and if flows through obstacle arrays can be characterized by an effective permeability, similar to fluid flows through porous media. It will also explore how surface tension of the droplets and shear forces due to the surrounding fluid influence particle breakup and coalescence, culminating in a quantitative framework capable of predicting the size distribution of deformable particles generated by emulsification devices employing obstacle array geometries. Experimental studies will be complemented by efficient computer simulations of bubbles and droplets flowing through constrictions and obstacle arrays in both two and three dimensions, enabled by the development of a new deformable particle model (DPM) that incorporates surface tension. The project's potential contributions include a predictive framework for deformable particle transport through obstacle arrays, which could lead to improved design and operation of microfluidic and lab-on-a-chip devices. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2024 · 2024-07

Inside eukaryotic cells, the genome is compacted into the nucleus as chromatin - very long DNA-protein polymers. Both chromatin structure and dynamics underlie all genomic processes, from gene regulation, to DNA repair, to the timing of replication. While there has been tremendous recent progress in revealing the time-average chromatin organization in cell populations, there is limited knowledge concerning the dynamics of the intrachromosomal contacts, that are required for many key cellular processes, for example, during gene activation or inactivation. This project will address this knowledge deficit by characterizing experimentally and theoretically the mesoscale dynamics of intrachromosomal contacts in live yeast cells. The project will systematically investigate how long it takes for two distal locations on a chromosome to come into biologically-meaningful contact -- the so-called “first passage time” -- by deploying a recently-developed “recombinase state machine” assay, that provides a fluorescent read-out of contacts between specific pairs of gene loci in individual cells. This approach will enable accurate measurements of first passage time distributions for multiple pairs of gene loci that collectively realize a dense set of different genomic separations and multiple genomic contexts. The project will also exploit the availability of multiple yeast strains and yeast’s genetic malleability to explore the roles of various chromatin-associated proteins, as well as such factors as chromatin density and mobility, in order to comprehensively interrogate how chromatin organization affects first passage time. Further impetus for the project comes from longstanding theoretical interest in first passage times for contact between different locations on a simple linear polymer. However, first passage times for the chromatin polymer, organized via hierarchy of loops, remains largely unexplored theoretically. Therefore, this project will also carry out simulations of the first passage time for intrachromosomal contacts, by combining models of polymer dynamics and chromatin configuration. The chromatin configuration will be modeled by a newly-developed version of loop-extrusion theory, that accurately reproduces measurements of time-averaged chromatin organization in yeast. Comparisons between experimental and simulated first passage times will identify where current theoretical understanding requires further development, ultimately leading to a predictive theory for biologically-meaningful contacts within the genome. Such knowledge is paramount to advance our understanding and control of how rare and stochastic chromatin contacts collectively lead to robust expression patterns, that define a cellular state and identity, including cells in a diseased or cancer state. Being able to control or engineer desired expression patterns or attenuate natural patterns would have a major positive impact on societal health and would advance a new biotechnology. To quantitatively predict and control gene expression patterns, we need to know the fundamental rules that govern chromatin conformations and their rearrangements, and the time scales involved. This project will provide essential knowledge for understanding these rules and developing relevant technologies in the future. The graduate student trained in this project will eventually join the nation’s STEM workforce. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2024-07

Project Summary There is no change from the parent award. Mental health research faces significant challenges, including the heterogeneity of diagnostic groups and the lack of precise characterization of individual patients, hindering effective clinical decision-making. However, data-driven approaches, such as machine learning and computational analyses, have emerged as crucial tools to address these challenges. By integrating data from behavioral assessments, clinical records, and biological markers, these approaches can generate more precise and objective clinical phenotypes, leading to improved diagnostic accuracy, personalized treatment selection, mechanistic insights, and enhanced monitoring and prognosis. The use of data-driven approaches holds immense importance in revolutionizing mental health research, enabling tailored interventions, and advancing our understanding and management of mental disorders. By integrating diverse data sources and leveraging advanced computational techniques, the Individually Measured Phenotypes to Advance Computational Translation in Mental Health (IMPACT-MH) initiative was formed, with the goal to harness the power of big data to address the complexity and heterogeneity of mental disorders, ultimately improving patient care and outcomes. Research gap: Mental disorders exhibit complex characteristics, making it difficult to represent, collect, and analyze heterogeneous data effectively. One major challenge is the absence of a unified representation for mental health data. While the Research Domain Criteria framework (RDoC) aims to provide a systematic framework, a formal representation using existing biomedical standards, such as ontologies, is still lacking. Developing such standards is crucial to generate computational phenotypes. Additionally, once data standards and normalization methods are established, disseminating them to researchers is essential for promoting the generation of interoperable and reusable data. Moreover, ensuring the generalizability of phenotyping algorithms beyond their original development institute and minimizing issues associated with potential errors are critical factors for enabling broad applications of such algorithms in mental health research. Addressing these challenges is pivotal for advancing computational phenotyping in mental health and facilitating its broader utilization. Method: To tackle these challenges, we will establish three cores within our three Aims: Aim 1. Project Coordination and Data Management Core. This core will facilitate effective coordination and communication across IMPACT-MH projects. Additionally, it will build a robust data management system that encompasses the necessary infrastructure and pipelines to efficiently gather, integrate, store, and manipulate de-identified multi-modal data from multiple IMPACT-MH projects and submit them to NIH data repositories. Aim 2. Data Standards Core. This core will work on defining comprehensive data standards by leveraging the RDoC framework and existing ontologies. It will also develop a consensus process and data harmonization methods aimed at maximizing the clinical, administrative, and scientific value of the various ascertainment and assessment practices used across the IMPACT-MH projects. Aim 3. Data Analytics Core. This core will focus on conducting rigorous analyses on the aggregated data from the IMPACT-MH projects. It will develop methods to address potential errors associated with the datasets, algorithms, and applications used. By implementing sound analytical approaches, we aim to ensure the validity and reliability of the findings generated from the data.