Northeastern University

NIH Research Projects · FY 2025 · 2021-09

PROJECT SUMMARY/ABSTRACT Adolescence is a period of heightened vulnerability for many forms of psychopathology, particularly depression, anxiety, and suicidal behaviors. Disorders that emerge during this time have lasting consequences, including elevated risk of recurrence, and poor psychosocial functioning. This vulnerability comes at a time when peer victimization becomes more common and emotional and physiological responses to peer rejection are elevated, rendering victimization particularly damaging during adolescence. Despite the strong links between peer victimization and internalizing problems during adolescence, the behavioral and neural mechanisms underlying this association remain unclear, limiting our ability to prevent the onset of victimization-related psychopathology in youth. The proposed project will test a novel conceptual model, in which it is proposed that two underlying dimensions of peer victimization, peer threat (e.g., presence of negative social experiences, like rejection) and peer deprivation (e.g., absence of positive social experiences, like ostracism) differentially shape neurocognitive processes and social behaviors that have relevance for psychopathology. Specifically, it is argued that peer threat influences neurocognitive and behavioral processes in ways that enhance threat detection and processing (Aim 1), whereas experiences of peer deprivation may contribute to blunted reward sensitivity and low social motivation (Aim 2). The project will then examine whether neurocognitive and behavioral alterations serve as mechanisms linking peer victimization with internalizing psychopathology (Aim 3). The proposed research will test this conceptual model by using a combination of experimental behavioral and fMRI tasks, as well as an intensive longitudinal design, leveraging advancements in digital phenotyping, computational neuroscience, and predictive modeling approaches. Critically, by implementing advanced statistical machine learning methods for predictive modeling, the proposed research may be able to identify patterns of real-world social behavior that are influenced by victimization and, in turn, predict the emergence of psychopathology. Identifying developmental processes that are disrupted following peer victimization and ultimately lead to psychopathology is a necessary first step in developing targeted intervention approaches. This award will also provide the candidate, who has a strong background in developmental social neuroscience and clinical science, with critical training in the implementation of digital phenotyping, computational modeling, and advanced statistical techniques to promote a successful transition to an independent research career.

NIH Research Projects · FY 2025 · 2021-09

Effective measurement of cognitive abilities is fundamental to effective diagnostics, risk assessment and evaluation of interventions targeted towards older adults (OA) and in particular those with Alzheimer's Disease and related dementias (ADRD). With the ubiquitous availability of smartphone/tablet technology in modern society a proliferation of mobile cognitive assessments from companies, healthcare providers, and researchers are being developed. However, a difficulty in evaluating such interventions, and in particular making comparison between them is the lack of standardization/interoperability of assessment tools. This is especially the case for early stage/mechanistic studies where it is common for investigators to each use their own labs' toolset to evaluate intervention outcomes. Here we address particular needs in the field of cognitive training, as well as for other longitudinal assessments focused on OA, where the limited standardization and accessibility of cognitive outcome measures makes it difficult to evaluate effectiveness of interventions. This R21/R33 infrastructure proposal seeks to develop shared tools to facilitate effective translation and sharing of cognitive assessment and training procedures. We accomplish this by leveraging technologies, existing assessment batteries, and know- how from 3 groups that have each independently developed robust systems for cognitive assessment and training that can run on mobile devices (UCR Brain Game Center, UCSF Neuroscape Center, and UCI Working Memory and Plasticity Lab). We target development of systems that allow for interoperability of assessments, enrollment/participant tracking systems, data visualization, and participant compliance systems. In the R21 phase, we aim to develop such systems and demonstrate that they can be effectively shared across labs, and in the R33 phase, these systems will be both tested for robustness in large scale-research projects that will now be able to share outcome measures, and for developing personalized, precision training approaches for participants based upon these assessments. Further, these systems will be documented and will be shared with other scientists groups to reduce the barrier of entry for other groups. The long-term impact of this work will be an infrastructure that will support better comparison across studies of cognitive training, as well as other interventions that are increasingly being used to ameliorate cognitive declines in older adults such as those related to ADRD. The key value of this system compared to others is that it will simultaneously support the flexibility required for basic research, by facilitating groups to continue to use their own lab's software systems, while at the same time providing them with a powerful infrastructure for sharing that allows them to incorporate assessments, server infrastructure and compliance tools into their own studies. This will facilitate comparisons across studies using common outcome measures as well as the ability to use the same assessments in numerous other domains including risk-assessment and longitudinal testing in older adults at risk for ADRD.

NIH Research Projects · FY 2025 · 2021-09

PROJECT SUMMARY Leveraging the power of the human genome to understand the risks, causes, and treatments of human dis- ease remains a grand challenge for all of biology and medicine. While sequencing costs have plummeted, and clinical implementation has become commonplace, interpreting human genomes remains a highly challenging task. It is our hypothesis that understanding the function of the genome and its products at a molecular, tissue, and phenotypic level using advanced machine learning will help unlock the door to better interpretation for sci- entific discovery and better clinical outcomes based on genomic medicine. To that end, our team has spent the past two decades working to develop computational models of biology, to predict how those models are perturbed through changes in the genome, and to use those perturbations to model phenotype and disease. We have had many research outputs in this area, having developed and published a number of widely used methods that predict biochemical and phenotypic changes caused by genetic variants to infer phenotype and pathogenicity. However, we believe that there is a coming convergence between the variability in clinical inter- pretation, high-throughput biotechnology assays, and modern machine learning methodology that will result in more accurate clinical assessments and improved clinical care. Therefore, in this ambitious proposal, we are addressing important questions in variant and genome interpretation consistent with this view and the mission of the IGVF Consortium. Our major goals include (1) developing advanced semi-supervised approaches to predict variants that disrupt molecular function and/or are capable of altering phenotypes; (2) identifying in- formative assays, variants, and genes to automate experimental design with an emphasis on resource alloca- tion and reduction of ascertainment bias in the Consortium; and (3) developing machine learning approaches to integrate these models into a workflow of the IGVF Consortium and enable the interaction between compu- tation and experiment in order to catalyze advances in both genetic variant interpretation and predictive model development.

NIH Research Projects · FY 2026 · 2021-09

Traditional file-based neuroimaging data management and integration strategies have shown increasing limitations in accommodating the meteoric growth in both the scale and complexity of today’s neuroimaging data. The sophisticated software and hardware pipelines required in many of today’s neuroimaging studies have produced numerous platform-specific data files that are increasingly difficult to parse, exchange, and understand by the broader research community. Modern neuroimaging studies are hampered by not only the challenge of parsing and managing rigid and diverse file types, but also the lack of a unified interface for systematic data validation, query, manipulation and integration, thereby limiting its ability to handle large datasets. Creating a future-proof, highly scalable, and low-maintenance data storage and dissemination platform is highly desirable for the broad and rapidly growing neuroimaging community. The future of neuroimaging data management must be scalable, searchable, verifiable, and capable of accommodating highly complex hierarchical data generated from complex paradigms involving multi-modal inputs. Inspired by the great success of NoSQL database platforms, we envision that a unified database interface for managing complex neuroimaging data and exchanging human-readable hierarchical data records will be highly suitable to address the urgent needs of next-generation neuroimaging data management. In this project, we aim to solidify a series of easy-to-adopt, easy-to-extend, human-readable JSON (http://json.org) based data file specifications to systematically assist the storage, exchange and integration of existing and emerging neuroimaging datasets. These JSON-encoded universal data files readily enable users to utilize highly scalable and high-performance NoSQL databases, such as CouchDB and MongoDB, to rapidly disseminate large, NIH-funded neuroimaging public datasets, and enable validation and automation. Our group has been a major contributor to JSON-based scientific data storage since 2011. We have published open- source specifications (http://openjdata.org) to standardize the exchange of neuroimaging data for common formats such as NIfTI/GIfTI/SNIRF, building a solid foundation for application-specific adoptions. In this project, we seek to further develop, solidify, and disseminate JSON-based data exchange specifications and NoSQL databases. We have built collaborations to major neuroimaging data analysis stakeholders, such as FreeSurfer, SPM, FieldTrip, HOMER, BrainStorm. At the end of this project, we will be able to 1) develop a set of stable universal file formats that greatly modernize data sharing in neuroimaging applications, easing future maintenance and extension, and 2) provide open-source tools for users to build NoSQL database backends to facilitate integration and automation of public/private databases, enabling query, validation, and scale-up. Success in this project will result in a robust data exchange platform to facilitate convenient data sharing, promote reproducible research, and forge efficient collaborations among a broad neuroimaging community.

NIH Research Projects · FY 2025 · 2021-07

PROJECT SUMMARY Acinetobacter baumannii is among the most antibiotic-resistant pathogens known, and the emergence of isolates with enhanced virulence poses an urgent public health challenge. Understanding how the microorganism thwarts antibiotic and immune attack via its protective cell envelope is essential to developing new strategies for controlling this threat. Envelope synthesis and integrity in bacteria are typically maintained by a large number of response systems that control specific aspects of the envelope. A. baumannii, however, has diverged substantially from this paradigm. The pathogen lacks orthologs of many canonical envelope response proteins and instead relies on a single two-protein regulatory system to globally modulate every layer of the envelope and control both antibiotic resistance and ability to cause disease. This unique system, known as BfmRS, lowers susceptibility to a wide range of drugs, antagonizes innate immune killing, and facilitates development of lethal disease in mice. Intriguingly, a clinical isolate showing enhanced virulence requires the system for growth. BfmRS is therefore tightly linked to the intractability of infections with the pathogen and represents a key potential therapeutic target. Despite its fundamental importance, we lack an understanding of how the large BfmRS regulon controls broad- range drug resistance and pathogenicity, and what signals the system senses. The objective of the proposed studies is to understand how A. baumannii uses a single control circuit to simultaneously modulate resistance and virulence. Our central hypothesis is that BfmRS jointly controls the barrier to both drug penetration and innate immune attack by modulating the level of key outer membrane (OM) structures in response to disruptions in envelope protein production. We will test this hypothesis by pursuing three Aims, which build on our extensive preliminary data defining the BfmRS regulon and its chemical-genetic profile, as well as the phosphorylation cascade it uses for signaling. In Aim 1 we will test the model that BfmRS controls the bacterial interface with both antibiotics and innate immune effectors by modulating the OM barrier. In Aim 2, we will identify the antibiotic-induced and intrinsic stress signals that are sensed by BfmRS. In Aim 3, we will define the relationship between variability in BfmRS activity, growth-dependence, and virulence across diverse patient isolates as a test of the model that variation in BfmRS signaling level is a driver of enhanced virulence in invasive strains. This work will elucidate the mechanisms by which a unique regulatory system controls both resistance and pathogenicity in a critically important nosocomial microbe. These results will inform strategies for potentiating antibiotic and immune action for killing extensively drug-resistant bacteria.

NIH Research Projects · FY 2025 · 2021-04

The Antimicrobial Resistance Crisis (AMR) has been recognized for years, and the significance of this global human health problem sets it apart from other types of diseases, because it affects not only individuals, but has a potential to disrupt the life of society. We have a stark reminder of the ability of a pathogen to bring normal life to a halt, as we experience the COVID-19 pandemic. In the case of a virus, we can usually count on a reasonably rapid development of a vaccine. For multidrug-resistant bacteria, we do not have a similarly reliable approach, and the pipeline of novel compounds against the most problematic pathogens, MDR Gram- negative bacteria, is very thin (Lewis, Cell 2020). We recently discovered a novel class of compounds acting against important Gram-negative pathogens, the darobactins (Imai et al., Nature 2019). Darobactin A is a 7- mer modified peptide containing two unusual fused rings. This creates a rigid β-strain from the peptide backbone. The target is BamA, an essential chaperone that inserts proteins such as porins into the outer membrane. BamA recognizes a signal sequence of incoming peptides that bind to one of its β-strands. Darobactin, which has a unique preformed β-strand, is a better binder and prevents substrates from interacting with BamA. Importantly, BamA mutants resistant to darobactin A lose virulence. Darobactin A has no cytotoxicity and shows good efficacy in mouse septicemia and thigh models against such pathogens as polymyxin-resistant E. coli and KPC K. pneumoniae. Darobactin A is ribosomally translated and coded by a RiPP operon. Bioinformatics search of the NCBI genomes database resulted in identifying 8 analogs with the same scaffold, and 6 darobactin-like analogs. The goal of this project is to evaluate the darobactins and identify the best leads. Additional analogs will be identified by searching through the raw data of the metagenomics database. We will synthesize the dar operons, clone them into E. coli and optimize production. For this, we will use an approach we recently developed, screening mutagenized producers in agarose microdroplets containing a YFP-labeled test pathogen. FACS analysis allows to sort droplets in which the test strain is inhibited. Spectrum of action will be determined, resistant mutants will either confirm BamA as a target, or point to a new one. We will analyze virulence of resistant mutants in detail. Compounds will be tested for cytotoxicity and animal safety and efficacy with target pathogens. This project will result in leads ready to enter into development to treat pathogens of critical priority.

NIH Research Projects · FY 2025 · 2021-01

Project Summary My long-term career goal is to become an independent pharmacist researcher with expertise in psychiatric multimorbidity research in cardiovascular disease. Through my research program I aim to identify factors that influence the health and medication taking behaviors of older adults with psychiatric multimorbidity using “real world” observational data and direct patient and stakeholder engagement. I have clinical (PharmD) and graduate (PhD) training in pharmacy and I am an assistant professor at Northeastern University in the Department of Pharmacy and Health Systems Sciences. Training facilitated by this K01 award will help me achieve my long- term career goal by building on my clinical and technical skills as a pharmacist and researcher and providing me with the protected time necessary to expand my expertise in longitudinal data analysis, mixed methods and patient-centered research in cardiovascular disease and multimorbidity. I have assembled an outstanding, committed, and interdisciplinary team of nationally renowned experts in fields relevant to the proposed research and my long-term goals. My training plan complements my prior training and experience and incorporates interaction with mentors, formal coursework, hands-on training, workshops, conferences and research activities. Using a mixed methods design, my proposal seeks to build and expand on my preliminary work completed in partnership with the ongoing NHLBI funded study, Systematic Assessment of Geriatric Elements in Atrial Fibrillation (SAGE-AF). I will leverage observational data, currently being collected in SAGE-AF, to evaluate the longitudinal impact of psychiatric multimorbidity (i.e., dyads and the triad of depression, anxiety and cognitive impairment) on oral anticoagulant (OAC) use and outcomes in patients with atrial fibrillation (AF). I will examine relationships between psychiatric multimorbidity and clinical outcomes such as bleeding and medication adherence. I will also examine relationships between psychiatric multimorbidity and patient reported outcomes that are not readily available in the medical record but are increasingly important to patients, their families and clinicians such as patient satisfaction and health-related quality of life. Then, I will conduct focus groups of SAGE- AF participants (92% report willingness to continue in SAGE-AF) and stakeholders to understand protective factors that influence anticoagulation success among patients with psychiatric multimorbidity, including factors that are not evaluated in the SAGE-AF data. The specific aims are to: 1) Examine the relationship between psychiatric multimorbidity and OAC prescribing in AF; 2) Among OAC users, examine longitudinal associations between psychiatric multimorbidity and OAC success indicators and patient reported outcomes over 2 years; and 3) Conduct 6-8 qualitative focus groups including SAGE-AF participants, their caregivers and their clinicians, to identify factors that influence OAC success. This proposal aligns with NHLBI’s overarching objective 3 by investigating factors that account for differences in health among populations and has high potential to identify factors that can be useful targets for patient-centered interventions.

NIH Research Projects · FY 2025 · 2020-09

Project Summary Despite the existence of promising osteoarthritis (OA) drugs, its treatment remains a challenge due to ineffective drug delivery systems. Intra-articular (IA) delivery is inadequate as drugs rapidly clear out from joint space and are unable to penetrate through the dense, negatively charged cartilage and reach their cell and matrix target sites at optimal concentrations. As a result, no disease modifying OA drugs (DMOADs) have passed clinical trials due to concerns of systemic toxicity and lack of cartilage targeting. For effective treatment, it is critical to stimulate a disease modifying biological response within multiple joint tissues, including cartilage, synovium and subchondral bone. Interleukin (IL)-1 receptor antagonist (IL-1RA) is proven to be a promising DMOAD for modulating both synovium inflammation and cartilage catabolism in preclinical models of post-traumatic (PT)OA; however, it has failed to show sustained clinical effect owing to lack of cartilage targeting and short joint residence time. The high negative charge density of cartilage provides a unique opportunity to use electrostatic interactions for enhancing uptake, depth of penetration, and retention of cationic drugs or drug carriers. We have shown that the cationic glycoprotein Avidin, owing to its optimal size and charge, was effective for intra-cartilage delivery as it rapidly penetrated through full thickness of cartilage in rats and rabbits following IA injection, resulting in 400- fold higher intra-cartilage uptake compared to its neutral counterpart and was retained inside cartilage for 3-4 weeks. Based on Avidin’s structure, we have designed a Cationic Peptide Carrier (CPC) that displayed similarly high uptake in both normal and glycosaminoglycan-depleted cartilage. This project will develop electrically charged IL-1RA by conjugating it with Avidin and CPC to make it cartilage penetrating and binding, thus increasing its tissue specificity and residence time. This way, cartilage can be converted from a barrier to drug entry into a drug depot, such that the anti-catabolic effects of charged IL-1RA in both cartilage and nearby synovium are significantly enhanced compared to unmodified IL-1RA. In Aim 1, Avidin-IL-1RA and CPC-IL-1RA will be characterized and their key transport properties (diffusivities, equilibrium uptakes, partitioning, binding constants) will be compared with unmodified IL-1RA in normal and arthritic cartilage. Aim 2 will evaluate the biological efficacy of a single dose of charged IL-1RA for inhibiting cytokine induced catabolism in a cartilage- synovium co-culture OA model, comparing Avidin/CPC-IL-1RA conjugates with single and continuous dose of unmodified IL-1RA. Aim 3 will determine the therapeutic potential of a single IA injection of charged IL-1RA relative to unmodified IL-1RA using a rabbit PTOA model. This work will advance the field of charge based drug delivery in targeting multiple joint tissues for effective, holistic OA treatment by applying fundamental concepts of bio-electrostatics and bio-transport. This charge-based platform can be used for delivering a wide range of drugs to other tissues with similar properties, such as meniscus, intervertebral disc and fracture callus, and also enable clinical translation of various OA drugs that have failed clinical trials due to lack of tissue targeting.

NIH Research Projects · FY 2024 · 2020-09

ABSTRACT   In response to RFA-AA-20-007, which calls for the development of medication to treat Alcohol Use Disorders (AUD), this U01 proposal describes research to advance the CB1 neutral antagonist AM6527 towards IND-enabling studies for treating AUD. The current therapies for AUD are either behavioral or are limited to drugs such as disulfiram, acamprosate and naltrexone, which are restricted to specific patient populations in terms of their therapeutic effects.  Given these epidemiological and economic issues related with AUD, there is an urgent need for novel pharmacological interventions that are more acceptable and selective towards treating AUD. Rimonabant, a CB1 antagonist, has proven to have unacceptable adverse effects, possibly resulting from its CB1 inverse agonist actions. In contrast, AM6527 is a novel, selective, CB1 cannabinoid-receptor neutral antagonist that is devoid of inverse agonist activity. The preclinical pharmacology of AM6527 in rodents and other prototypes within this class of compounds in nonhuman primates strongly supports their potential utility as therapy for AUD. AM6527 has a favorable safety profile, without evidence of adverse side effects (e.g., nausea, depression) that have been reported with CB1-receptor inverse agonists such as rimonabant and taranabant in preclinical and clinical studies. Based upon these positive preclinical data indicating its effectiveness as a CB1 antagonist, we plan to move AM6527 toward IND-enabling studies in preparation for first-in-man studies to treat AUD.  Specifically, our work in this U01 proposal is designed to meet the following aims: Chemistry, manufacturing, and controls of drug substance, non-GLP preclinical safety studies including hERG, genotoxicity, ADME and dose escalation studies in two species. The subsequent steps will comprise: Chemistry, manufacturing, and controls of drug product including formulation development, cGMP scale-up, stability, and analytical methods; toxicokinetics with single and repeat 28-day dosing safety pharmacology, followed by advisement on product development planning, IND preparation and preclinical regulatory strategy.

NIH Research Projects · FY 2024 · 2020-09

PROJECT SUMMARY In this proposal we plan to contribute addressing the above foundational and operational challenges by advancing the science of influenza modeling and contributing novel methods and data sources that will increase the accuracy and availability of seasonal and pandemic influenza models. To address these challenges, we plan to build on the unique mechanistic spatially structured modeling approaches developed by our consortium, that includes stochastic metapopulation models and fully developed agent-based models nested together in our global epidemic and mobility modeling (GLEAM) approach. The objective of this project is to generate novel and actionable scientific insights from dynamic transmission models of influenza transmission that effectively integrate key socio-demographic indicators of the focus population, as well as a wide spectrum of pharmaceutical and non-pharmaceutical interventions. Our proposed work in specific aim 1 (A1) will leverage our global modeling (from the global to local scale) framework that can be used to explore the multi-year impact of influenza vaccination, antiviral prophylaxis/treatment, and community mitigation during influenza seasons and pandemics. Our specific aim 2 (A2) will focus on using high quality data to model heterogeneous transmission drivers and novel contact pattern stratifications that will allow us to guide mitigation strategies and prioritization for interventions. In our Aim 3 (A3) we will use artificial intelligence approaches to identify interventions that are particularly synergistic and well-suited to particular epidemic scenarios, for seasonal and pandemic influenza. Our overarching goal is to provide a modeling portfolio with flexible and innovative mathematical and computational approaches. We aim to address several questions commonly asked about seasonal and pandemic influenza and match these with analytical methods and outbreak projections. The modeling and data developed in this project can help facilitate and justify transparent public health decisions, while contributing to the definition of standard methods for model selection and validation. Finally, our influenza modeling platform can also benefit the broader network of modeling teams and can be used to improve result sharing and harmonization of modeling approaches.

NIH Research Projects · FY 2025 · 2020-09

Project Summary Acquired cognitive impairment associated with aging, neurocognitive disorders, like Alzheimer’s disease, and traumatic brain injury (TBI) pose major challenges to healthcare systems throughout the world. To facilitate successful aging, effective methods are needed for monitoring cognition in order to detect early signs of impaired cognition and dementia. However, the effectiveness of existing neuropsychological assessments is thwarted by their sporadicity, difficulty in accounting for the context-dependent nature of patients’ health (e.g., having a “good” or a “bad” day), and reliance on frequently inaccurate patient and caregiver reports. Thus, new approaches are needed for objective and ecologically valid assessment of cognitive function and for early detection of impaired cognition associated with neurocognitive disorders like Alzheimer’s disease and TBI. Current mHealth and AI approaches enable continuous context inference from smartphone use and location data. Patient reports of Instrumental Activities of Daily Living (IADLs) can be greatly enhanced with continuous, objective context inferences from passively collected smartphone-based data that can contribute to earlier and more accurate diagnoses of neurocognitive disorder including Alzheimer’s disease and related dementia. Cross-sectional studies involving younger and older adults suggest that characteristics of an individual’s mobile application use and typing speed correlate with working memory, attention, and psychomotor function. Therefore, the proposed augmentation of clinical assessments with continuous and objective estimates of cognitive changes derived computationally and unobtrusively from mobile application use characteristics and motor interactions has the potential to inform both the early detection and diagnosis of impaired cognition associated with neurocognitive disorders, such as Alzheimer’s disease and TBI, and potentially treatment selection for their amelioration. The central hypothesis to be tested in this longitudinal study is that changes in cognition, including impaired cognition associated with Alzheimer’s disease and other neurocognitive disorders, can be inferred from smartphone data. In the F99 phase, Mr. Kos will a) augment existing measures derived from IADLs and lifespace mobility with analogous measures inferred from mobile location-finding systems (e.g., GPS) and application use data, and b) propose a digital biomarker for tracking cognitive changes and impaired cognition based on temporal patterns in application use and motor interactions with smartphones. Validation will be conducted on 80 middle-aged individuals; 20 with subjective cognitive impairment, 20 with diagnosed Mild Cognitive Impairment, 20 with TBI, and 20 cognitively intact controls. In the K00 phase, Mr. Kos will select the subset of these measures determined to be most applicable to tracking cognitive changes and impaired cognition and, thus, prime for detecting early signs of Alzheimer’s disease and other types of neurocognitive disorder in older adults. Their refinement and validation in clinical trials using biomarkers and brain imaging data will enable developing a mechanistic understanding of a) the relationships between impaired cognition, Alzheimer’s disease, and neurocognitive changes including TBI and b) how these relationships are reflected in interactions with mobile technologies.

NIH Research Projects · FY 2023 · 2020-09

PROJECT SUMMARY Substance use disorders (SUDs) are heritable psychiatric disorders with a significant genetic component. Opioid dependence, one of the most heritable SUDS, has reached epidemic proportions in the United States. Human genome-wide association studies (GWAS) are statistically underpowered to detect the majority of common genetic variation that contributes to opioid dependence. Discovery-based genetics in mammalian model organisms is a powerful complement to human GWAS and can uncover novel genetic factors, biological pathways, and gene networks underlying addiction traits. Mouse models are advantageous because they enable collection of the relevant brain tissue at the appropriate time points under controlled opioid dosing. Furthermore, gene editing permits the validation of functional variants in vivo within the same species on a controlled, genetic background. Reduced Complexity Crosses (RCCs) are genetic crosses between inbred mouse substrains that are nearly genetically identical and can vastly improve the speed at which causal genetic factors can be identified. Our primary objective is to use an RCC between BALB/c substrains to discover the genetic and molecular basis of opioid addiction-relevant traits at two stages of opioid dependence following repeated administration of the mu opioid receptor agonist oxycodone (OXY; the active ingredient of Oxycontin®). We found robust differences between BALB/c substrains in opioid adaptive behaviors, including state-dependent learning of OXY-induced locomotor stimulation and reward following limited, low-dose administration (1.25 mg/kg, IP) as well as the emotional-affective component of opioid withdrawal and weight loss following repeated high-dose administration (40 mg/kg, IP). In Aim 1, we will map quantitative trait loci (QTLs) underlying these OXY phenotypes in an RCC F2 cross. In Aim 2, we will map QTLs controlling gene expression (eQTLs) in the relevant brain tissues of control F2 mice and in OXY-trained F2 mice. We will then nominate candidate causal genes and nucleotides underlying behavior by integrating eQTL with behavioral QTL analysis. To increase precision in assigning candidate variants with the regulation of gene expression and behavior and to identify biological pathways and opioid-adaptive gene networks in specific cell types, we will use single nucleus RNA- seq (snRNA-seq) of brain tissue following limited, low-dose OXY and repeated high-dose OXY. In Aim 3, we will validate candidate functional variants underlying OXY phenotypes using CRISPR/Cas9 gene editing of each of the two alternate alleles onto each reciprocal substrain background. This approach will allow us to demonstrate both necessity and sufficiency of the quantitative trait nucleotides. The proposed studies will identify the genetic basis of unique opioid phenotypes across two stages of opioid dependence. Independent from gene discovery, these studies have broader application in revealing novel, actionable insight toward cellular adaptations at progressive stages of the opioid addiction process and potentially improving behavioral outcomes.

NIH Research Projects · FY 2025 · 2020-08

How can a basketball player reliably land his free throw in the basket, and yet still miss occasionally under nominally identical circumstances? While such skills are a paragon of motor expertise, even seemingly mundane actions also require surprising dexterity. When carrying a full cup of coffee, we exhibit motor skill that is far beyond what is typically studied in the laboratory. Specifically, when interacting with objects - the essence of any tool use -, successful actions require fine-grained control of interaction forces that have been beyond the purview of neuroscience to date. The proposed research examines the neural basis of motor expertise by bringing rich interactive tasks into the laboratory. The two PIs combine their long-standing experience in computational motor control and neurophysiology to study novel behavioral paradigms both in humans and non-human primates. Building on conceptual and computational overlap in their respective research, where skill is associated with low-dimensional structure in high-dimensional neural and behavioral redundant spaces, they will test the overall hypothesis that patterns of neural activity exhibit many of the characteristics of the behavior. Two aims will study two examples of motor skill: throwing an object and transporting an object with internal dynamics, both rendered in virtual environments. Parallel experiments in humans and primates will generate rich behavioral data that will be matched with intracortical recordings in the cerebral cortex of non-human primates. To date, non-human primate studies have necessitated that animals perform near-identical repetitions of simple behaviors to facilitate the analysis of neural activity. Now, modern multi-neuronal recording techniques make it possible to embrace more sophisticated real-world behaviors and address core principles of movement discovered in human motor control: high dimensionality, redundancy, and the ever-present variability. This research will develop a suite of computational tools that afford the analysis of behavioral and neural data with commensurate techniques and sophistication. This research will be transformative as it advances the motor challenges examined and brings insights from intracortical neurophysiology closer to understanding of human motor expertise. These scientific insights will channel into a large range of outreach activities to achieve broader impacts for the general public. RELEVANCE (See instructions): Patients with neurological disorders such as stroke face challenges in their daily activities, grasping a cup to bring to their mouths to drink; these actions are essentially interactive tool use. This research seeks insights into neural activation during such skilled actions and interactions to get closer to understand neural activity in tasks relevant in real life. Extending from PI Batista’s experience, neuroprosthetics and brain-machine interfaces are direct clinical application that may benefit from our findings and recovery

NIH Research Projects · FY 2024 · 2020-08

Up to 2/3 of youth with autism spectrum disorder (ASD) exhibit challenging and often dangerous behaviors, including aggression and self-injury, that profoundly limits their access to community, educational, and therapeutic resources. Risk for challenging behavior and its consequences (e.g., psychiatric hospitalization; polypharmacy) increases during adolescence, particularly for severely affected youth with ASD (SA-ASD; characterized by intellectual disability or minimal verbal ability). Identifying contributors to challenging behaviors that can be modified using non-pharmacological and non-invasive approaches is a high priority for clinicians, researchers, and family caregivers. Yet, SA-ASD youth are under-represented in research, which restricts the potential for improvements for those who need it most. Sleep and the circadian regulation of sleep are modifiable through behavioral and chronotherapeutic interventions. In typically developing youth, the homeostatic and circadian regulation of sleep exhibits marked changes during adolescence (e.g., circadian phase delay). These changes confer risk for sleep problems and challenging behaviors mediated in part by changes in alertness, mood, and cognition. The proposed cross-sectional mentored study (K99) will evaluate if circadian phase delay is associated with sleep problems and challenging behaviors in SA-ASD youth. The study will be conducted in an inpatient psychiatric unit to facilitate training with, and enrollment of, SA-ASD youth, a population that is difficult to study in the community. The study will serve as a platform for training in (1) supportive techniques and non-invasive sleep measurement methods to increase participation of SA-ASD youth in objective sleep and circadian science, (2) systematic coding of observed challenging behavior and ecological momentary assessment methodology, (3) advanced analytic methods to model prospective associations and individual differences, and (4) the process of intervention development to support future intervention research. Training and lessons learned during the mentored phase will inform and support an independent prospective study (R00) of associations between circadian phase, sleep, and challenging behaviors in an independent sample of SA-ASD adolescents. The proposed study will follow SA-ASD youth as they are discharged from a psychiatric inpatient facility and return home. Following a pre-discharge baseline assessment in the hospital, in-home assessments of circadian phase and sleep and ecological momentary assessments of challenging behavior will be conducted at 6 and 12 months post-discharge. The objective of this study is to determine if circadian phase delay precedes the emergence/escalation of sleep problems and challenging behavior. This work will set the foundation for future studies to compare prospective associations between circadian phase, sleep, and challenging behavior across multiple comparison groups (e.g., higher- functioning ASD and typical development) and evaluate the clinical utility of intervening upon circadian phase to improve sleep and challenging behavior in ASD youth.

NIH Research Projects · FY 2024 · 2020-07

Project Summary/Abstract Mutant forms of KRAS are a key driver in human tumors but remains refractory to therapeutic intervention despite over three decades of research. Clinical attempts to directly or indirectly inhibit KRAS function have both yielded unsatisfactory results. The difficulty for developing small molecule KRAS inhibitors has heightened the importance of alternative methods targeting the oncogene. One such strategy involves therapeutic nucleic acids, which make it possible to deplete target proteins that are intractable to conventional drug modalities. We have developed a novel form of nucleic acid therapeutics, termed pacDNA, that substantially enhances the antitumor activity of nucleic acid drugs by elevating in vivo stability, accelerating cellular uptake, and improving plasma pharmacokinetics and tumor accumulation, allowing a much lower dosage to be used compared to conventional methods. The pacDNA also suppresses nearly all side effects associated with traditional nucleic acid drugs by reducing unwanted nucleic acid-protein interactions. In this proposal, we aim to build upon our promising preliminary results, and gain deeper insights into the cell biology of the pacDNA with respect to cell uptake mechanism, intracellular trafficking, KRAS depletion and subsequent cell signaling, and demonstrate efficacy in KRAS-dependent non-small cell lung cancer cell lines and 3D models. In addition, we will study the primary pharmacology and antitumor activity of pacDNA in advanced preclinical lung cancer models including an orthotopic tumor model, a patient derived tumor model, and a syngeneic genetically engineered mouse model (GEMM), and perform initial in vivo safety and tolerability studies. The outcome of this project will be a safe and potent anti-KRAS agent that can be readily translated into clinical studies for non-small cell lung cancer and potentially additional cancer classes.

NIH Research Projects · FY 2024 · 2020-07

Project Summary/Abstract Accurate measurement of human behavior using devices could significantly advance current knowledge on the dose-response relationships between chronic diseases and behaviors such as physical activity, sedentary behavior, and sleep. The primary objective of this proposal is to develop valid approaches to measure 24-hour physical behavior, as well as to demonstrate a procedure via which those approaches can be compared to others. We aim to help the research community to converge on methods that use devices to accurately measure physical activity type and intensity, sedentary behavior and posture, and sleep in adults. Many promising methods have been proposed to measure behavior from activity monitors. Unfortunately, these methods – which are now being proposed in large numbers – are typically validated on small amounts of data. Thus, they may perform well on lab data, but fail when used in the field on large-scale epidemiological or intervention studies. Moreover, the performance of different methods is rarely compared head-to-head, creating uncertainty for public health researchers about which are the best to use. Quantifying the relative performance of methods that produce similar outcome measures but use different devices or on-body device locations is even more unusual. We will make it easy for researchers interested in physical activity measurement to meaningfully compare performance between new methods and confidently apply those methods to both large-scale surveillance studies and longitudinal interventions. The project has four specific aims: (1) Collect well-annotated data of physical activity, sedentary behavior, and sleep, (2) Use the data from Aim 1 to develop and validate approaches that yield 24-hour estimates of free-living physical activity (type, intensity), sedentary behavior (type, posture), and sleep (wake/sleep, stages), (3) Develop and incrementally refine a suite of tools that researchers can use to easily deploy advanced approaches to measure physical activity, sedentary behavior, and sleep, even for large data, and (4) Use the data and new approaches (Aims 1 and 3) to host four competitions evaluating models, where all entries submitted by other researchers, will be directly compared, ranked, and improved. The goal is to help researchers converge on “gold standard” methods to robustly measure physical activity using common monitor configurations, as well as those devices and configurations likely to be used soon.

NIH Research Projects · FY 2025 · 2020-07

PROJECT SUMMARY The majority of highly diverse biological processes are enabled through proteins, protein post-translational modifications (e.g., glycosylation), proteoforms, protein interactions, as well as other biological molecules (e.g., lipids, RNAs, etc.). Aberrations of abundance, activity, function, integrity, and localization of these biological molecules and their diverse interactions can lead to severe diseases. Furthermore, disruption of molecular profiles and structural characteristics by novel targeted therapies can be an important biomarker for the response to these drugs in personalized medicine approaches. Clinical and biological specimens are often available in limited amounts, which greatly hampers the progress in diagnostics, therapy development, and biomedical research. Microbiopsy and liquid biopsies may contain small populations of rare cells, macromolecular complexes, or other biomolecular structures and species. Traditional analytical techniques cannot be readily used for the analysis of small cell populations, microscopic clinical samples, and individual cells, mainly due to limitations in sensitivity. Therefore, many biological and clinically relevant studies have not been undertaken because of the lack of technology for such low-level samples. Here, we propose to develop analytical platforms that will enable high sensitivity spatial multiomic analysis of scarce biomedical samples. This task will demand the development of novel approaches in sample preparation, ultra-low flow liquid phase separations interfaced with MS, MS data acquisition, and data analysis. Developing such novel methods for thorough profiling of microscale samples and integrating them in innovative, robust, and reproducible automated platforms capable of efficient and high sensitivity quantitative and structural characterization of released glycans, peptides, intact proteoforms, protein complexes, and PTMs by MS will be highly desirable for gaining biological insights into molecular mechanisms of the disease and discovery of therapeutic targets and biomarkers for diagnostic and prognostic purposes. The developed platforms will be evaluated using well-controlled model systems and applied in the most clinically relevant settings to examine model line-based systems and primary samples.

NIH Research Projects · FY 2026 · 2018-08

Project Summary/Abstract Title: New Computational Systems Biology Methods for Modeling Gene Regulatory Circuits One of the biggest challenges in biology is to elucidate complex gene interactions and networks for the purpose of developing interventions in human disease. Particularly important are those gene networks that control cellular state transitions (e.g., replicative to quiescent, epithelial to mesenchymal, etc.). Thanks to the emergence of next-generation sequencing technology, rich data resources are available for mapping gene regulatory interactions. However, the field still lacks a systems-level understanding of how genes in a network collectively perform their functions and control cellular state transitions, information that is critical for informed clinical intervention. The PI’s long-term goal is to design effective computer-aided strategies for predicting therapeutic interventions by integrating knowledge of gene regulatory networks, genomics data from patients, and systems- biology model simulations. So far, numerous computational methods have been developed to infer and model gene regulatory networks. However, they typically suffer from the following issues. First, current approaches are still ineffective to choose an appropriate set of genes and regulatory interactions in a network to model. Current approaches infer regulatory relationships based on association of gene expression signals, but generally don't also consider whether an inferred gene regulatory network can operate as a functional dynamical system driving expected transitions between the network states. Second, traditional mathematical modeling is hard to be applied systematically to large systems, because many kinetic parameters are unmeasurable directly from experiments, especially in vivo. The parameter uncertainty and the potential risk of overfitting in large systems have limited the predictive power of systems biology. To address these issues, the PI’s research program will develop a suite of computational systems biology algorithms to construct and model high-quality core regulatory circuits driving cellular state transitions. We have recently developed enhanced ensemble-based mathematical modeling algorithms for simulating network behaviors without the need of detailed kinetic parameters. This advance has allowed an integrated top-down and bottom-up systems-biology modeling, as evident from the PI’s recently developed network reconstruction and modeling method NetAct and network coarse-graining algorithm SacoGraci. The PI’s research program will further advance novel technologies of ensemble-based modeling and their applications to optimize high-quality systems-biology models that capture cellular state transitions. The algorithms will be benchmarked and refined using in-silico simulated data, publicly available omics data sets, and data from collaborations, with a focus on cell differentiation in developmental processes and state transitions in oncogenesis. Success of the research program will result in a comprehensive toolkit that will unveil the gene regulatory mechanism of cellular state transitions. The algorithmic development is expected to have a broad impact on not only basic research in systems biology but also shed light on therapeutic intervention in genomic medicine.

NIH Research Projects · FY 2026 · 2017-09

Abstract Alzheimer's disease (AD) poses a triple threat to public health, as its prevalence is on the rise, its costs are immense, and there is no effective therapy. However, drug development attempts for the treatment of AD have met with minimal success. The failure is largely attributable to a reductionist concept of "one drug, one gene, one disease." As AD is a multigenic heterogeneous illness, a new therapeutic strategy is urgently required to concurrently target the numerous pathogenic processes involved for the genesis and progression of AD in each individual patient. Many translational bioinformatics strategies for AD drug repurposing have been developed in recent years. Existing target-based, phenotype-based, network-based, and patient-based drug repurposing strategies are unable to fully address the challenges of AD drug repurposing due to the lack of thoroughly validated drug targets, potent lead compounds, and high-throughput phenotype readouts that can characterize the molecular complexity of AD. Over the past decade, we have built an artificial intelligence- based quantitative systems pharmacology (AI-QSP) platform that attempts to predict and characterize genome-wide chemical-protein interactions and functional activities, as well as correlate molecular interactions with phenotypic responses. Our AI-QSP platform integrates diverse omics data synergistically and incorporates machine learning, biophysics, and systems biology methodologies. The AI-QSP platform has been effectively applied to drug repurposing including AD, polypharmacology, side effect prediction, and precision medicine. Established our proof-of-concept studies, we propose to develop and thoroughly evaluate a unique computational methodology that combines target-based and mechanism-driven phenotypic chemical screening for AD individualized drug repurposing. Using a novel domain adaptation strategy, we will expand our context- independent phenotypic compound screening methodologies to AD patient-specific, cell type-specific, transcriptome-based drug repurposing. In addition, we will analyze the ADME features of repurposed pharmaceuticals in the human brain utilizing cutting-edge physiologically based pharmacokinetics (PBPK) techniques. We will improve state-of-the-art drug-gene-disease network models for Alzheimer's disease drug repurposing by incorporating understudied dark proteins that are abundant in the target list suggested by AD omics studies and their inhibitory or activatory effects, and by applying graph mining techniques for drug-gene- disease link predictions. Using cell-based disease models and RNA-seq studies, we will combine complementary phenotype-based and target-based techniques to rank drug candidates and confirm their efficacy and toxicity on AD treatment. In conclusion, the successful completion of this project could provide the scientific community with a novel translational bioinformatics resource for identifying potential therapeutics for effective personalized AD treatments and advancing drug repurposing to a new phase of lead optimizations and clinical trials.

NIH Research Projects · FY 2026 · 2017-09

Project Summary/Abstract Oligonucleotides face several biopharmaceutical difficulties, including stability and delivery issues as well as non-hybridization activities such as coagulopathy and unwanted activation of the immune system. We have developed a unique oligonucleotide delivery system, termed pacDNA, which uses a high-density bottlebrush polymer to provide oligonucleotides with binding selectivity. The polymer amounts to an entropic barrier, reducing access to the oligonucleotide by various proteins (and thus side effects) but still allows for unhindered hybridization. This novel strategy not only improves nuclease stability, preserves target-binding capability, and minimizes off-target side effects, but also massively enhances plasma pharmacokinetics, tissue retention, and antisense potency in vivo. Our current studies also reveal that the pacDNA’s pharmacological properties are intimately related to the bottlebrush backbone. In addition, the pacDNA appears to be uniquely capable of evading anti-carrier adaptive immunity, which is useful for therapies that requires long-term/frequent dosing. Finally, the pacDNA deposits into tissues and organs that lack mature delivery technologies for, such as the skin, the skeletal muscle, and the heart. These surprising and enabling discoveries will be the basis for investigations in the next funding period, in which we will 1) explore the property space of the pacDNA structure using a combinatorial polymer library with specific backbone compositions and monomer sequences; 2) probe in vivo properties of the pacDNA in mouse and non-human primate models and how it is able to evade adaptive immunity; and 3) explore the potential of pacDNA to create first/best-in-class therapies that take advantage of its unique strengths using a relevant preclinical disease model (progeria). We anticipate that accomplishment of these objectives will yield significant fundamental understanding of this class of materials and bring us much closer to clinical evaluation of pacDNA.

NIH Research Projects · FY 2025 · 2017-08

Abstract Genome-Wide Association Studies, whole-genome sequencing, and high-throughput techniques have generated vast amounts of diverse omics data. However, these sets of data have not yet been fully explored to improve the effectiveness and efficiency of drug discovery. Only 5-10% of druggable proteins are targeted by pharmaceuticals. The undrugged proteins are potential targets of yet-incurable diseases but they remain dark, i.e., their endogenous and exogeneous small molecule ligands (chemicals) are unknown. Furthermore, there is a knowledge gap to link drug-target binding affinities to clinical outcomes. We know little if the target is activated or inhibited by the binder (i.e., function activity: agonist vs. antagonist). To date, few experimental and computational tools can determine genome-wide protein-chemical interactions (PCIs) for dark proteins and ligand-induced functional activities (LIFAs) for majority of proteins including both dark and well-studied proteins. Existing machine learning techniques is unsuccessful in exploring the dark protein space due to an out-of- distribution (OOD) problem, i.e., they cannot reliably predict the function of an unseen protein or chemical if it is significantly different from proteins or chemicals in the training data, respectively. Commonly used computational tools for structure-based drug design, such as protein-ligand docking/scoring and Molecular Dynamics simulations, are neither scalable nor particularly reliable. As a result, we only have a limited capability of compound screening and lead optimization between novel chemicals and dark proteins. This proposal seeks to develop and experimentally validate innovative methods for illuminating the molecular function of dark proteins and apply them to drug discovery for presently incurable diseases. Building on our successful proof-of-concept studies and our close multidisciplinary collaborations between experimental and computational laboratories, we will develop a novel computational framework to model drug actions on a multi-scale by integrating big data from chemical genomics, functional genomics, and structural genomics and developing innovative deep learning algorithms. Specifically, we will develop a structure-enhanced deep learning framework to reliably and accurately predict the binding affinity of novel small molecule ligands to dark proteins on a genome-scale. We will integrate functional genomics with chemical genomics to predict ligand-induced functional activity. We will apply the methods developed to design and experimentally test inhibitors of dark anti-cancer target AVIL and selective dual antagonists of dopamine receptors for opioid use disorder (OUD). The proposed research offers an innovative concept, methodology, and translational applications. Completing this research will fill a critical knowledge gap in understanding drug actions in a biological system and significantly impact drug discovery for complex diseases, many of which lack effective and safe treatments. The developed methodology and platform will not only immediately impact the NIH’s “Illuminating the Druggable Genome” Program but also has potentially broad applications in other areas of biomedical research.

NIH Research Projects · FY 2025 · 2016-09

PROJECT SUMMARY This proposal aims to improve the rigor and reproducibility of research on plasticity in human working memory (WM), and related executive functions (EFs) in adolescent youth. We address a critical gap between research and practice that is characterized by a growing commercial space marketing cognitive training approaches (with WM being one of the most common targets), which are particularly catering to typically developing children and those diagnosed with ADHD to improve mental health and scholastic performance. However, despite expansive literature, there exists limited basic research on WM and EF training in adolescents, and both methods and findings are mixed across studies. Here, we address these significant gaps that pose obstacles to understanding interventions’ reliability and validity by collecting a large-scale open dataset that compares different training approaches on a common set of outcome measures. This is addressed through 4 aims, all based upon a large- scale study involving 720 adolescents, stratified across age, sex and ADHD status randomized across interventions: Aim 1 – What ingredients of cognitive training mediate training outcomes? We compare training conditions that represent some of the most common ingredients used across cognitive training studies (gamification, multiple domain training, and coaching) in comparison to basic non-gamified n-back training, using a common set of outcomes measures. Aim 2 – How do training ingredients differentially impact youth with ADHD? With ADHD being one of the most common targets of cognitive training, we examine whether and how youth with ADHD are differentially impacted by the training conditions. Aim 3 – What individual characteristics moderate training outcomes? We examine how cognitive training outcomes may differ across individuals, and whether this variability may explain differential effects across studies. Aim 4 - Promote open science through sharing of software tools and data. We will develop and share a cross-platform training and assessment app and a research portal that will promote data sharing, replication, and access to other groups using common individual difference variables and outcome measures. As a renewal, targeting adolescents with and without ADHD is a natural next step in this research program that started in typically functioning adults. This research is significant and timely as it addresses the limited basic research of cognitive training in adolescents with and without ADHD, whom are understudied, but at the same time, are common targets of commercial products in this space. This will lead to a more robust and representative understanding of factors that moderate and mediate cognitive training that will be impactful whether or not one hypothesizes near or far transfer from cognitive training. Further, findings could inform future research to understand broader age groups (younger children and older adults) as well as to address lifespan and developmental factors more broadly, mental health, disease, and brain damage, that interact with cognitive training to give rise to individualized outcomes.

NIH Research Projects · FY 2025 · 2016-09

The effect of early environmental exposures on child health and development is an important area of public health that no single cohort, or even a few, can answer alone. Determining this effect becomes more challenging when considering that these exposures influence multiple interrelated health outcomes such as obesity, neurodevelopment and reproductive development. The ECHO Cohort Site in Puerto Rico (PR) will focus on these health outcomes, taking advantage of the rich and large sample size and longitudinal nature of the national ECHO Cohort and expertise across the ECHO consortium. The project will contribute innovative research to (Aim 1) examine the influence of environmental factors on maternal diet and obesity during pregnancy, child diet and obesity during early to middle childhood, and the relationship between maternal-child diet and onset of puberty and (Aim 2) determine the relationship between in utero and early childhood exposure to environmental chemicals, individually and as mixtures, and neurodevelopment and reproductive development, across early childhood, middle childhood, and adolescence. Most importantly, the project will build upon the ongoing ECHO cohort in PR (Aim 3) to follow up and collect data and biospecimens from the 800 children that are participating in the ECHO UG3 cohort, as well as recruit an additional 1,200 pregnant participants, yielding 1,000 more children, for a total of 1800 children contributing data and biospecimens to the ECHO Cohort. The proposed 1,200 new pregnancies will include at least 80 completed pregnancies with preconception data (Aim 4), tracked from a cohort of 470 potential preconception participants and, if available, their conceiving partner. This research will contribute meaningfully to ECHO’s mission, offering (a) an important cohort that will enrich consortium wide data, (b) significant, innovative science on the influence of the environment and exposure to multiple chemicals on multiple interrelated health outcomes such as obesity, neurodevelopment, and reproductive development, and (c) expertise to lead and participate in additional new scientific directions involving biomarkers of exposure or biologic response, child brain or reproductive development, and statistical methods in collaboration with the ECHO consortium, to answer impactful and cutting-edge research questions. Results from our study will inform future clinical intervention, risk assessment and policy-setting efforts, with direct relevance to the U.S. population.

NIH Research Projects · FY 2025 · 2016-05

Project Summary/Abstract While the steady transition from 2-D mammography to 3-D digital breast tomosynthesis (DBT) in recent years has resulted in improved breast cancer diagnosis, x-ray based breast imaging techniques are inherently limited by their inability to provide physiologically relevant functional information. The persistent high percentages of unnecessary biopsies and poor sensitivity to malignant tumors among dense breasts and early-stage cancers have motivated the research community to seek additional functional assessment. Diffuse optical tomography (DOT) – a non-invasive imaging modality using only non-ionizing near-infrared light – has shown promise in fulfilling such a role by exploiting cancer’s high endogenous contrasts in angiogenesis and metabolism. However, the low spatial sampling density due to use of optical fibers and the ill-posedness of the DOT inverse problem have resulted in low image resolution in DOT, hampering its clinical translation. For nearly two decades, our group has been a key contributor in advancing DOT for breast clinical diagnosis via multi-modal imaging. We have developed translational imaging systems combining DOT with DBT as well as prior-guided algorithms that “fuse” x-ray contrasts with DOT reconstructions, and tested these systems in clinical studies with over 450 subjects. Under the initial funding period, we have developed a new DOT imaging architecture that utilizes fiber-less widefield pattern-illumination and camera detection to achieve ultra-high-density and uniform spatial sampling. We have also developed highly efficient compressive-sensing strategies to take advantage of such unprecedented dense datasets without long acquisition time. With these advances, our optical mammography co-imager (OMCI) system prototype is capable of providing a raw measurement density several orders of magnitude higher than previously reported high-density DOT systems at only a fraction of the cost and instrument size. Built on this strong momentum, in the renewed period, we aim to further enhance optical measurement density by developing the first-in-the-field widefield frequency-domain (FD) DOT system combining single-pixel imaging with our compressive-sensing approaches. Moreover, we will also implement real-time adaptive illumination/detection patterns optimized for individual breasts. Furthermore, through our decade-long research in DOT, we have developed a comprehensive suite of toolboxes that have been rigorously validated using clinical data, offering unique capabilities for widefield DOT forward modeling, pattern compression/optimization, and prior-guided image reconstructions. With a strong track record for developing and maintaining high-quality open-source tools, we are excited to share these packages with the biophotonics research community as open-source software. Both our next-generation widefield FD breast DOT hardware and algorithm innovations will be tested by an experienced clinical team with healthy (N=20) and lesion cases (N=60). The success of this project will not only accelerate the clinical translation of breast DOT, but also introduce powerful new DOT architectures to assist biophotonics researchers in a wide range of applications.

NIH Research Projects · FY 2026 · 2015-09

PROJECT SUMMARY Manipulation of complex objects or tool use is a hallmark of daily living, and loss of manual dexterity due to motor impairments lead to loss of independence. Manipulating objects is particularly challenging when the object has internal dynamics that is not directly controlled. Even the seemingly simple task of transporting a cup of coffee has intrinsic dynamics that humans need to predict, preempt, and compensate for to avoid spilling. Control of such complex nonlinear systems with online error corrections based on precise internal models appears daunting, given the slow neural processes and the ubiquitous noise in the sensorimotor system. Hence, this research tests the hypothesis that humans learn to simplify the object interactions, i.e., make the interactions predictable. The task of carrying a cup of coffee is modeled with a cart-and-pendulum system that is rendered in a virtual environment and subjects interact with the virtual cup via a robotic manipulandum. To gain insight into human control strategies, this proposal develops a task-dynamic approach that affords principled hypothesis-testing by parsing the complex dynamics into execution and result variables, with minimal assumptions about the human controller. Eight experiments test the overall hypothesis that humans seek solutions that are predictable, by correlating hand-object motions, and making the behavior stable and tolerant to error and risk to obviate error corrections and prevent failure. Aim-1 tests control of internal dynamics in linear movements and examines how humans choose initial conditions to mitigate perturbations, how they preempt undesired ball oscillations, how they exploit intermittent contact to develop a stable rhythm, and how they modify the object properties to facilitate stable contact behavior. To examine learning, Aim-2 scales up the dimensionality of the task by introducing more real-life planar cup movements, which creates an exponential increase in complexity. Four experiments test task goals that introduce new dynamic challenges, such as combination of rhythmic and discrete movements, complex ball dynamics when changing movement directions, adaptation and modification of object properties, all to show how humans either exploit or override internal dynamics to achieve predictability. Aim-3 introduces a real version of the task with a custom-designed device, the MAGIC Table. Following a comparison of the real and virtual set-ups, the MAGIC Table is used to leverage the theoretical framework to create novel sensitive metrics to quantify motor function for clinical applications. Specifically, we assess severity and recovery of motor impairment in a cohort of patients after stroke. As manual dexterity is compromised in many individuals with neurological disorders, the experimental paradigm and its quantitative analyses promise to become a useful platform to gain insights into neurological diseases.