Carnegie-Mellon University

NIH Research Projects · FY 2025 · 2024-08

Project Summary There is a critical need to develop therapies targeting Streptococcus pneumoniae (Spn): a major human pathogen contributing to global morbidity and mortality. We propose to use Specifically Targeted Anti-Microbial Peptides (STAMPs), where a broad spectrum anti-microbial peptide (AMP) is fused to a Spn secreted peptide. The former will provide anti-microbial activity, while the latter will impart selectivity of this activity at the taxonomic level or increase efficacy of the AMP. This project builds on preliminary data where we show that human beta- defensin 3 is ineffective in killing Spn. In contrast, a fusion of this AMP to a Spn cell-cell communication peptide resulted in a peptide that eliminated >95% of Spn in culture. Further, we propose an innovative way to produce STAMP peptides, where we will deploy a “molecular velcro” technology based on complementary peptide nucleic acid (PNA) adapters. These PNA adapters are connected to the ends of independent targeting and AMP domains, allowing facile assembly and screening of PNA-STAMP libraries. Conclusion of the proposed work will introduce species-specific antimicrobials for Spn targeting and develop a conceptual and technical platform for the development of Gram-positive antimicrobials.

NIH Research Projects · FY 2026 · 2024-08

Under-treatment of mental health problems remains a major issue in the US, especially for youth, people of color, and individuals with low incomes. Technology may help reduce disparities in and expand the reach of mental health services. However, the newest technologies, such as generative AI, remain fraught with perils such as hallucinations. Therefore, rather than using AI to directly interact with clients, we will harness generative AI to provide training to mental health support providers, with the ultimate goal of increasing accessibility of mental health services and improving mental health outcomes for those receiving care. The goal of this research project is to develop and evaluate an automated, scalable system for delivering experiential training to mental healthcare providers. Specifically, we propose to develop a multi-agent training environment to provide interactive and experiential training on the micro- skills and underlying common factors for both lay counselors and paraprofessionals. Our training environment consists of three components: Virtual Patient, Assessor Agents, and Trainer Agents. Our proposal has four aims. During Aim 1, we will develop and evaluate a set of LLM-based Virtual Patients (VPs) that (a) realistically depict a wide range of common clinical problems (e.g., depression, job-related stress, ADHD, and suicidality), (b) engage in coherent conversations with trainees, and (c) present typical counseling challenges, such as addressing resistance to sharing problems in depth. During Aim 2, we will develop an Assessor Agent capable of automatically assessing the micro-skills used by the trainee, as well as how the trainee accomplishes higher-level segment goals (common factors). During Aim 3, we will develop a Trainer Agent capable of interacting with trainees, the Assessor Agent module, and the Virtual Patient module to achieve optimal training goals. During Aim 4, we will recruit 7Cups supporters to use our multi-agent training environment to evaluate its impacts on the training outcomes.

NIH Research Projects · FY 2026 · 2024-08

In 2010, Pittsburgh Supercomputing Center (PSC) and D. E. Shaw Research (DESRES) partnered to make the Anton special-purpose molecular dynamics (MD) supercomputer available to the national biomedical research community for the first time. Anton enabled researchers to simulate biomolecular systems two orders of magnitude faster than any conventional supercomputer, allowing researchers access to the critical multi-microsecond and longer timescales on which most biologically-significant molecular processes take place. In 2016, with operational support from NIH, DESRES made available a next generation Anton 2 system at PSC at no cost. This multimillion-dollar gift from DESRES, with NIH support, provided a unique opportunity for researchers to tackle even more groundbreaking biological questions. The primary goal of this project is to provide the national biomedical research community with access to the next-generation Anton system — Anton 3. Anton 3 will provide an order of magnitude increase in simulation speed with respect to Anton 2, as well as a ten-fold increase in the maximum chemical system size, up to seven million atoms. Since demand for Anton 3 will be high, our goals are focused not only on providing access but also on maximizing the value of this limited resource to the community — recognizing that the creativity of the broad biomedical research community is one of our most powerful resources for innovation. Anton 3 will be integrated as an NSF ACCESS resource provider, and the accessibility and impact of Anton 3 will significantly increase by engaging in their outreach and training programs. Through these outreach efforts, researchers at many more institutions will have the opportunity to use Anton 3. In addition, the hundreds of long-timescale MD trajectories researchers generate on Anton systems constitute a unique and valuable data collection that can be used to gain additional biomedical knowledge and advance the application of machine learning protocols to augment molecular dynamics simulations. These important datasets will be available to researchers and educators worldwide through a web portal and co-located on the PSC’s Bridges-2 and Neocortex systems for classroom instruction and research, including reanalysis, machine learning, and data mining.

NSF Awards · FY 2024 · 2024-07

The broader impact of this Partnerships for Innovation - Technology Translation (PFI-TT) project is to provide an innovative solution to the heat dissipation challenge at the interfaces between electronics and cooling systems. With continually increasing power density, the performance of heat dissipation technologies has become a dominant limiting factor in maintaining the safe and stable operation of electronics. The technology being developed in this PFI project enables electronics to operate at lower temperatures, thereby enhancing their performance and reliability. Additionally, in data centers where cooling demands consume vast amounts of water and electricity, this technology can significantly improve thermal management and energy efficiency, contributing to industrial decarbonization and climate change mitigation. The team investigates scalable and cost-effective manufacturing strategies for mass producing high-performance, nanostructured, thermal pads and exploring their integration with existing products to realize their full commercialization potential. The high-performance, nanostructured, thermal pads developed in this project could resolve the critical challenge of heat dissipation by substantially reducing thermal interface resistance and enhancing system reliability. The high compliance and flexibility of the new nanostructured thermal pads enables absorption of thermal stresses due to the different coefficients of expansion of two mating materials. The materials will also accommodate dynamic structural stresses and vibrations, which will significantly improve device or system reliability. Scalable and cost-effective manufacturing technologies will be developed to accelerate the commercialization of the nanostructured thermal pads. Through the collaboration with the industry partner, the nanostructured thermal pads will undergo comprehensive validations to meet industrial needs for commercialization. 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 Major Research Instrumentation (MRI) project provides for the purchase of an Aerodyne Vocus Proton Transfer Reaction-Mass Spectrometer (Vocus PTR-MS) that will enable the ability to simultaneously measure hundreds to thousands of different gas and particle-phase organic compounds in the atmosphere. The Vocus PTR-MS will be housed in the Center for Atmospheric Particle Studies (CAPS) laboratories at Carnegie Mellon University, complementing other state-of-the-art equipment there. Among other capabilities, the instrument is designed to measure extremely polar constituents that nucleate and grow to form atmospheric nanoparticles that can be harmful to human health. In addition to resolving differences between atmospheric carbon, oxygen, hydrogen, sulfur, and nitrogen atoms in molecules of the same nominal mass, this mass spectrometer has the sensitivity to measure highly reactive, critical free radicals such as organic peroxy radicals. The instrument design also reduces wall losses of those species as well as other sticky compounds. The Vocus is a factor of over a hundred times more sensitive than the instrument that the CMU group previously has been using for research. The new instrument will enable detection of the most lightly oxidized first-generation products at concentrations encountered in the atmosphere, radical species such as organic peroxy radicals, and gas-phase intermediate volatility organic compounds that play a central role in organic aerosol formation and aging. The high time resolution will permit sampling of source plumes in real-world conditions. Junior faculty, postdocs, and students will be trained to use the new instrument. This MRI project is funded by the NSF Atmospheric Chemistry Program. 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

The broader impact of this Partnerships for Innovation - Technology Translation (PFI-TT) project includes the development of generative artificial intelligence (AI)-powered products that unlock the information inside the written documents and books and make them available to any human being. More specifically, the project develops products that allow humans to interact and talk with books and other documents by asking questions in their natural language. The products will collect and organize these human interactions using sophisticated knowledge representation techniques, and use them, along with the informational contents of the books and documents, to power cutting edge generative AI models. Given that some of the most complex tasks and societal challenges, such as education, workforce development, learning, teaching, manufacturing, and research, rely on information, this project could be transformative. The technology could, for example, (1) reduce the cost of education, while also increasing its accessibility, (2) improve the productivity of workers in many industries by allowing them obtain answers to their questions with the click of a button, and (3) improve the speed of scientific progress by helping scientists to remain current with regard to scientific advances. This project combines techniques from several areas of computer science, including formal methods, algorithms, and artificial intelligence, to create the technical foundation for grounding Large Language Models (LLMs) in human knowledge. The project solves an important problem that limits the potential of LLMs - their tendency to "hallucinate" confidently - which make them difficult to use, due to the misleading and potentially harmful information they generate. To this end, the team will develop the technology to represent the informational content of a book (or more generally any document) and discussions pertaining to the book as a knowledge graph. Given the knowledge graph, the team will develop generative AI techniques that will mine the graph to generate the best response to a user query. The techniques work with off-the-shelf LLMs and deliver effective responses by grounding LLMs in the knowledge of human experts embedded in books or human discussions. The project will also develop a web-based platform allowing creation and sharing of knowledge, and engaging in discussions where human beings and generative AI technologies will work together for improved efficiency and scale. 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/Abstract Chronic pain is a common and debilitating medical condition, yet current treatment options are often ineffective and have side effects. Recent experimental work has identified classes of neurons in the spinal cord that mediate pain, which may be novel candidates for more targeted pharmaceutical intervention. However, there are two issues limiting clinical translation. First, Genome Wide Association Studies in humans have given us genetic markers of chronic pain, but we lack a genetic characterization of which cell types are affected by this genetic variation. Second, we need a method of targeting these specific cell types in mice, non-human-primates, and humans, the typical progression of testing for new therapeutics. I will make significant progress on both problems by analyzing the genomic cis-regulatory elements (CREs) of neural subtypes in the macaque and mouse spinal cord. Recent evidence suggests risk variants for many polygenic diseases especially accumulate in CREs. I hypothesize that neuron subtypes identified as pain-mediating in anatomical studies will be genetically enriched for chronic pain markers in CREs, compared to several controls. I further hypothesize that CREs evolutionarily conserved across macaque and mouse are even more enriched relative to those not conserved, based on the idea that CREs that are particularly important to function will be conserved. To address the goal of in vivo targeting of these cell types, I will leverage the fact that CREs are highly cell-type- specific drivers of gene expression, and can control the expression of an exogenous gene that could in principle alter neuronal excitability for therapeutic benefit. I will develop a machine learning pipeline that identifies CREs predicted to have high specific activity in target cell types. I will train models on mixed mouse and macaque CRE data, and I hypothesize that cross-species models will identify more translatable CREs. This strategy may not be optimal for pain treatment given the diversity of cell types involved in pain signaling, and my strategy may require simultaneous control of complex combinations of neuron subtypes, while minimizing off-target effects such as in motoneurons. I will therefore also design synthetic regulatory elements to preferentially target combinations of pain-mediating neurons, with minimal effects in ventral horn neurons such as motoneurons. I will take the top CRE candidates from my machine learning models and validate their cell-type-specificity activity in mouse with RNAScope experiments, and collaborators will validate CREs in macaque in a related project. In summary, I will characterize the genetic risk burden of chronic pain in evolutionarily conserved neuron populations, which will inform us of genetic pain mechanisms across experimental animal models. I will computationally identify and design CREs that target these populations, and validate their function in vivo in mouse, providing a valuable tool for pain experiments, and making progress toward a new form of pain therapy.

NSF Awards · FY 2024 · 2024-07

Large distributed storage systems within datacenters are primary components of cloud, Internet service, and data analytics infrastructures, and storage capacity demand is growing rapidly with the rise of data science, machine learning, and artificial intelligence. Analysts estimate that 175 zettabytes of data will be generated annually by 2025, most of which will be stored in data centers, which would require over 8 billion 20terabyte (20TB) mechanical Hard Disk Drives (HDDs) before accounting for data redundancy needed to protect data from device failures. Since such numbers are untenable, new technologies that allow higher capacity devices are being created, but they do not provide greater performance. The result is a coming performance “wall”, where the available access bandwidth-per-TB of stored data is too low to allow that data to be stored, maintained, and used. The goal of this research is to reimagine the decades-old application interfaces used for datacenter storage to enable large reductions in the bandwidth needed so that higher-capacity devices can be used. Existing input/output (IO) interfaces are imperative, such as “read this now” or “put this now”, which is easy for programmers but restrictive and inefficient for the system. This research will develop new “declarative” IO interfaces and orchestration approaches that allow programmers to express larger data access plans/needs and thereby allow the storage system to coordinate, coalesce, and schedule IO to minimize the aggregate bandwidth-per-TB needed by the various applications and data maintenance tasks required for reliable datacenter storage. The results will be more sustainable and cost-effective datacenter storage, eliminating the need to manufacture and deploy millions of HDDs, reducing Flash and DRAM cache requirements, and enabling deployment of new data maintenance activities that make data more useful, secure, and reliable. 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

The safety, security, and reliability of software applications have far-reaching impacts on everyone’s lives; therefore, formal reasoning about these guarantees has become increasingly important. While formal verification provides the highest assurance, very few companies can afford the cost of it. On the other hand, lower-effort approaches such as testing, applying linting tools, and peer code review have been widely adopted in the industry. In addition, a complex system may include differently analyzed components: some are tested, some go through a static analysis, while others undergo code review. This project aims to answer the question of how to formally reason about the safety, reliability, and security assurances of such complex systems when heterogeneous analysis methods are applied. Further, this project investigates, in the event of an incident, how counterfactual reasoning can be applied to identifying the cause of the issue and refine or harden the analyses or software to prevent future incidents. This project provides training opportunities for undergraduate and graduate students in topics including program testing, verification, vulnerability detection, and software security via research projects and dedicated course modules. The project first develops the logical and semantic foundations for compositional assurance reasoning, where modal operators such as possibility and necessity are used to express truth derived from under-approximate (incomplete) analysis and truth derived from over-approximate (complete) analysis, respectively. Reasoning principles to compose results from different types of analysis are built around these modal operators. To be concretely applicable to the analysis of programs for assurance, a Kripke semantics based on the program execution semantics is defined to give meaning to the logical formulas. Next, this project develops counterfactual reasoning principles to aid the refinement of the analysis and repairing programs when an incident occurs. The expressiveness, effectiveness, and efficiency of the reasoning systems are evaluated via case studies drawn from security incidents reported in recent years. 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

Biological neural circuits (BNCs) are complex neurite networks with interwoven neurons that process information and exchange communications through synapse (the small gap at the end of a neuron that allows a signal to pass from one neuron to the next) connections. Within BNCs, neuron arrangement and connectivity vary based on specific tasks. BNCs can coordinate physiological behaviors throughout the body by transmitting electric impulses and chemical signals. Together with patterned neurons, BNCs show great potential in many applications in computational neuroscience, biohybrid robotics and as a testbed for validating computational and machine learning paradigms. BNCs design requires computational tools to fully understand morphological growth and the regulation of material transport in neural circuits. The morphological growth of neurons is a very complex process involving both genetic and environmental components. How a neurite initiates from the soma (the body of the cell) and creates the axon (that carries signals from the soma to other targets) from dendrites (the receiving portion of the neuron) during growth remains challenging to predict. In addition, intracellular material transport is especially crucial to ensure necessary materials are delivered to the right locations for the development, function, and survival of neural circuits. The transport disruption can lead to abnormal accumulations of certain cellular material and extreme axonal swelling. This project will advance knowledge of the fundamental mechanism of neural growth, material transport regulation, and circuit dynamics. The resulting computational tools will support BNCs design and future development of biohybrid robotics and new therapies. The developed simulation software, research and educational materials will be disseminated broadly, including National Biomechanics Days in Pittsburgh and the network of female researchers. The goal of this project is to develop a new computational framework to predict neuron growth and transport regulation based on isogeometric analysis (IGA), phase field, partial differential equation (PDE)-constrained optimization, and machine learning techniques. This goal will be achieved through pursuit of three specific aims: (1) Feature-driven multi-stage neuron growth using isogeometric collocation-based phase field method and convolutional neural networks; (2) Data-driven material transport regulation and traffic jam simulations in neurons using physics-informed graph neural networks; and (3) Validation of BNC dynamics and prediction tools in micropatterned network cultures for healthy and degenerating cell types. This project will yield new computational tools to enable realistic 3D modeling and data-driven simulation of neuron growth and transport regulation for BNCs design and prediction. It will advance knowledge of neurobiology at the subcellular and cellular levels as well as knowledge of neural engineering on growing and controlling material transport for applications such as repair and renewal of damaged or degenerative neurons. It will develop new computer simulation software for growing and analyzing traffic control within the complex neuronal circuits. The proposed computational tools and experimental validation are critical, leading to transformative advances in BNC dynamics design in biohybrid robotics applications (e.g., biohybrid controllers). The IGA-based data-driven techniques can also be used to solve a wide variety of PDEs that describe cellular processes other than growth and transport regulation 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

Optimal transport provides a sensible mathematical framework to address the fundamental statistical question of how a statistician measures the distance between two distributions based on possibly large high-dimensional datasets. A variation of the original transportation problem featuring an entropic penalization has appeared as a more scalable alternative, fueling a wave of new results and successful applications in domains such as genomics, neuroscience, and economics, to name a few. Despite its practical success and the achieved understanding of some of its fundamental statistical properties, there is still a substantial gap between theory and practice in the entropic optimal transport framework. This project will bridge this gap through new methods grounded in an improved theoretical understanding of entropic optimal transport, potentially generating an innovative set of applications in the life sciences. Graduate students will be trained within the scope of this project. The core of this project focuses on two intimately related thrusts: first, to develop a foundation for inference in parametric models with entropic optimal transport and to identify the regimes for which this framework is best suited. This includes the problem of model-based clustering in high-dimensional, non-asymptotic regimes and a study of the robustness of entropic-optimal-transport estimators. Second, the PIs will develop statistical applications of entropic optimal transport in Alzheimer’s disease neuropathology and spatial transcriptomics. 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

In many economically and politically important situations, people's judgments and decisions require inputs from memory. For example, the decision of whether to vote for a politician will involve recalling examples of their policies and personal character. A consumer's decision about whether to purchase a particular product might involve recalling different types of information obtained at different points in time – e.g., friends’ experiences with the product, statistical summaries of online reviews, or advertisements. This research examines the consequences of different types of biases in memory storage and retrieval for the judgments people form at different points in time. Results from this research may improve our understanding of the role that memory plays in how people make decisions. The present work develops and employs a paradigm in which participants simultaneously receive two signals about a judgment object -- e.g., a potential purchase. One is a statistical signal which is highly informative about the true quality of the object but is unmemorable. The other is a less informative but more memorable signal: an account of a single person’s experience with the object. Both immediately and/or at a later point in time, participants report a belief about the object’s quality. We examine whether the differential memorability of the two signals causes judgements of the object to vary systematically over time. Immediately after receiving a positive statistical summary of a restaurant's ratings, for example, this summary might eclipse a friend's vivid story about their bad experience there; but, with the passage of time, the vivid story might be more memorable and play a more important role in decisions. In a first set of studies, we examine whether and how beliefs change over time because of differential memory decay across the two pieces of information. A second set of studies examines the mechanism behind this differential decay to identify exactly what makes information about a single experience more memorable than the statistical signal. 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 · 2024-07

Neuroscience has been able to gain major insights by relating measurements of neural activity to the brain’s sensory inputs and motor outputs. Yet most neural activity supports computations and cognitive functions (‘thoughts’) that are not directly measurable by the experimenter. The investigators for the present proposal invented a novel method to model an animal's thoughts by combining eXplainable Artificial Intelligence (XAI) cognitive models for naturalistic tasks with measurements of the animal’s sensory inputs and behavioral outputs. This model, called Inverse Rational Control (IRC), infers the internal model assumptions under which an animal's actions would be optimal. It then provides estimates of time series of subjective beliefs about the world that are consistent with this internal model. These estimates provide targets for a dimensionality reduction framework that assesses task-relevant computational dynamics within neural population activity. The investigators propose to use these analysis tools to find neural representations and transformations that implement these cognitive processes. They will apply this to a complex, naturalistic task that they developed: catching fireflies in virtual reality. The monkeys they successfully trained to perform this task demonstrably weigh uncertainty, develop predictions and long-term strategies, and apply nonlinear dynamics — all computations that are fundamental for brain function. The investigators propose first to apply their method to analyze existing behavioral data and neural recordings collected in a simple version of this task with a single target firefly. They will then collect new data on a multi-firefly version of the task, which incentivizes animals to make and implement longer-term plans. To analyze this data, the investigators will generalize their approach to allow them to learn which compressed representations are selected by the animal as the foundation for their strategies. These results will be used to form predictions about neural computations that will be tested using the electrophysiological data collected from multiple brain regions during this project. The results of this study will explain the computations required to perform a complex, strategic navigation task in the presence of uncertainty, and will demonstrate a new paradigm for understanding naturalistic brain computations. RELEVANCE (See instructions): This project will uncover the neural basis of cognitive processes in the primate brain that underlie spatial navigation, strategic planning, and behavioral control. It will demonstrate how a powerful new paradigm for understanding complex, natural brain computations can apply to a wide variety of tasks, to explain either adaptive or pathologically structured behavior. This will provide crucial guidance for understanding and improving disrupted human cognitive function.

NSF Awards · FY 2024 · 2024-07

Networks, representing relationships or interactions between subjects in complex systems, are ubiquitous across diverse engineering and scientific disciplines. However, real-world relationships often go beyond simple presence or absence, which poses challenges and necessitates the development of advanced methods. This project focuses on an important class of heterogeneous networks -- “signed networks”, where relationships can be positive (for example, friendship, alliance, and mutualism) or negative (for example, enmity, disputes, and competition). Such signed relationships are prevalent and exhibit substantially different and unique interaction patterns. This project aims to provide a comprehensive investigation on signed networks through statistical model-based learning and inference, pushing the frontier of our understanding of the role of negative edges in real-world complex systems. The research outcome will stimulate interdisciplinary research and make significant contributions in a broad range of scientific domains, including political science, biochemistry, medicine, genetics, ecology, and business and marketing. The project will support and train STEM workforce members by providing research training opportunities for undergraduate and graduate students. This project will develop novel statistical methodologies and theories for analyzing signed networks, focusing on the integration of negative relationships in three core problems: (a) understanding the formation mechanism of signed networks guided by fundamental social theories; (b) detecting communities in signed networks by leveraging unique patterns; and (c) learning informative and interpretable embeddings for signed networks to assist downstream analysis. For the first problem, the investigator will provide a valid statistical inference method under novel nonparametric graphon models for signed networks and study real-world evidence of conceptual theories to understand its formation mechanism. For the second problem, new fast community detection methods will be developed under a novel stochastic block model with a hierarchical structure for signed networks, with associated theory emphasizing the positive impacts of negative relationships. Finally, the project will tackle the problem of embedding learning by developing a general latent space framework. The developed methods, algorithms, and theories in this project will be applicable to various practical problems across different domains. 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 · 2024-06

SUMMARY Short organ preservation time imposes severe constraints on transplantation, contributes to reduced organ availability and high discard rates, diminishes organ quality while exacerbating graft rejection, and limits the length and quality of life for transplant recipients. Based on studies by the U.S. organ allocation authority UNOS/OPTN, organ procurement organizations and others, the potential deceased donor organ supply has been estimated to exceed the current number of organs transplanted by up to fivefold. Consequently, the goal of developing new approaches to organ preservation has become a national priority. The current project aims at developing means for organ preservation in cryogenic temperatures, while focusing on the liver due to its high priority need. Since ice formation is the cornerstone of injury in subzero temperatures, tremendous efforts have been devoted to control and ideally circumvent it. While the addition of cryoprotective agents (CPAs) can assist in controlling ice formation, these are inherently toxic materials. One way of creating more favorable conditions for cryopreservation while reducing the adverse effects of CPAs is to elevate the pressure surrounding the organ, and thereby lowering the required CPA concentration. This approach can work if the surrounding medium remains unfrozen, whereas the relevant temperature range is known as high-subzero (HSZ). While pressure elevation can be achieved by different means, one application that drew a lot of attention in recent years relies on cooling the specimen in a constant volume container (i.e., isochoric cooling), and on the anomalous tendency of water to expand upon freezing. However, HSZ isochoric preservation does require partial crystallization somewhere in the container for the application to work, where the exact location for it to take place generally remains a source of repeated debate, speculations, and uncertainties. This research proposal aims at advancing the approach of pressure-assisted cryopreservation beyond the isochoric application, by presenting an innovative approach to eliminate the need and even the possibility for ice formation around the organ. Towards this goal, five specific aims have been formulated in an increasing level of complexity: (1) to develop an innovative cryopreservation isochoric system with superior thermal regulation means; (2) to develop a HSZ isochoric cryopreservation benchmark using the new system on an isolated rat liver perfusion model; (3) to develop an innovative pressure-modulated cryopreservation system (PMCS), which eliminates the need and even the possibility for ice formation around the cryopreserved organ; (4) to evaluate the efficacy of the PMCS on an isolated rat liver perfusion model, where results will be benchmarked against isochoric cryopreservation from SA2; and (5) to advance the scientific foundation of pressure-assisted cryopreservation, while developing multiphysics modeling tools in the service of system design, physical characterization of CPAs, and analysis of liver model experiments.

NSF Awards · FY 2024 · 2024-06

The Secure Shell Protocol (SSH) is based on a cryptographic protocol for communication between a client and a server. SSH is commonly used for remote login and for the command-line mode for services such as file transfers. OpenSSH is the open-source version of SSH. High Performance Networking-SSH (HPN-SSH) is based on OpenSSH. HPN-SSH improves performance and functionalities for services such as file transfers for high performance scientific computing. This project is a scoping project to explore community building to support the adoption and further development of HPN-SSH via an open-source ecosystem. This project will create a governance structure, craft bylaws and policies, develop onboarding processes for contributors and developers, enhance public awareness, establish fundraising strategies for sustainability, and initiate collaborations with other open-source organizations. HPN-SSH improves performance and functionalities for services such as file transfers for high performance scientific computing. HPN-SSH does this by automatically implementing tuned flow control, parallelized ciphers, inband network telemetry, the resumption of failed transfers, and other performance improvements. The planned Open-Source Ecosystem (OSE) will have a direct impact on the processes and background required for the successful development and launch of non-profit open-source software foundations. This effort will expand knowledge in the areas of open-source sustainability practices, best practices for community engagement, flexible and functional open-source foundation governance structures, and the promotion of scientific software applications. It will also impact existing HPN-SSH user and developer communities by providing new avenues for interaction with the HPN-SSH development and leadership team. This interaction will inform and guide this work. More information about HPN-SSH can be found at https://hpnssh.org/. The public code repository can be found at https://github.com/rapier1/hpn-ssh. 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 · 2024-06

Diagrams are profoundly important to science, technology, education and mathematics (STEM) education and scientific communication. Illustrative diagrams require deep domain-knowledge and extensive design work. PENROSE is a new diagram format that encodes both domain knowledge and visual design in a reusable way. The PENROSE format contains the source information of diagram design: using PENROSE, diagram authors encode domain-specific concepts and how to visually represent them in plain-text languages. The PENROSE core tools compile this format into static images, animated vector graphics, and interactive web applications. Supporting PENROSE’s expansion into a sustainable OSE would best enable broader societal impacts on education and accessibility. The development of a diverse community of users and contributors from across academic disciplines and industry areas will drive a higher number of high-quality diagrams in research papers, in classrooms, and in public-facing scientific communication efforts. In addition, PENROSE provides the best format to enable user creation of diagrams that are accessible to differently abled groups. The readable plain-text format lends itself to customization for alt text and aural feedback as well as rearranging diagram layout for language localization and color acuity. The project will employ a strategy of open communication, usable and varied forms of documentation, and a culture of highlighting contributions to encourage users. This project will be used to assess the success of transitioning PENROSE transitions to an OSE in the following ways: gathering data on the nature of contributions to open-source development and diagramming in PENROSE; quantifying and characterizing PENROSE’s adoption and prevalence in published research across domains and disciplines; quantifying and characterizing the impact of PENROSE in educational settings, both for educators and learners; as well as gathering data on the effectiveness of implementations of accessibility foci. To realize the full potential of PENROSE, a standalone diagramming tool is not enough. The aim of this project is to establish a PENROSE ecosystem of open-source diagrams and to integrate PENROSE with other platforms. The primary goals for the ecosystem are to build robust infrastructure for authoring, sharing, and reusing diagrammatic contents and develop a community of expert creators that contribute open-source diagrams in the PENROSE format. Both directions are crucial for wide adoption of the PENROSE format and for transformational impact on STEM education and communication. 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 · 2024-06

Robotic and autonomous systems (RAS) are becoming more common in our daily lives. Making sure they operate safely, however, remains challenging. RAS code is often messy, and the real world in which these systems operate is complex and unpredictable. This research project focuses on creating a family of methods to find faults in these sophisticated systems by analyzing: 1) the program artifacts they are built from to extract models that can be more easily checked for faults, and 2) the real-world data they encounter (like images taken by cameras) to determine what portions of the environment are worth simulating more accurately. If successful, the research findings will inform solutions to challenges faced in the development of RAS, affording a path to improving public safety. The research will be integrated into educational curriculum and training. The key insights enabling the project are two-fold. First, although robotics code is often messy and complex, model-relevant behavior is typically implemented via a subset of application programming interfaces and configuration files with clearly-specifiable semantics. Second, field data encodes key spatial-temporal physical constraints imposed by the real world, which provides hints on how to steer simulation to reduce the gap with reality. The investigators will leverage these insights to effectively lift useful models from real code to detect compositional faults, identify and construct simulation scenarios that capture constraints imposed by the real world, and inform and validate the evolution of robotic systems. The proposed work will produce: 1) Techniques to infer rich behavior models with physical attributes from RAS code and artifacts, and to check those models against component and system properties; 2) Techniques to infer specifications from messy real-world data, to contrast those specifications against simulation states, and to synthesize missing simulation environments; 3) Techniques to inform RAS' evolution, to understand the impact of code, physical, and world changes, and to cost-effectively correct and test changing RAS; and 4) Studies with RAS to assess the techniques, quantifying their effect on the gaps between code and models, between simulation and reality, and during evolution. 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 · 2024-06

In the past fifteen years, there has been explosive growth in the number of applications seeking to leverage Machine Learning (ML) models. Before an ML model can be deployed, the model must be “trained” by repeatedly processing examples. Each example helps the model to make progressively more accurate predictions, "learning" how to solve the problem at hand. Unfortunately, this training step requires a large amount of expensive, highly-specialized hardware and can take several hours to complete. Given limited hardware resources, it is not obvious how to allocate (share) these resources across a stream of ML training jobs. The goal of this project is to develop new resource allocation policies that allow us to produce highly-accurate ML models, quickly, and with limited resources. The main challenge in this work is that ML training jobs present a number of unique characteristics compared to other computing workloads. ML training jobs are highly parallelizable, meaning a single training job might run across multiple servers. Each job also has a large number of configuration options that affect how fast the model learns a particular problem. This project uses mathematical modeling to develop new allocation policies specifically designed for ML training jobs. This research will be accompanied by the development of new courses and mentorship programs designed to recruit students into research on the modeling of computer systems. Machine Learning (ML) models are increasingly being deployed across a wide variety of applications. The growth in ML models has been accompanied by the development of specialized hardware accelerators that help reduce the training time for ML models. However, there has not been a similar degree of specialization in the scheduling algorithms used to train ML models on clusters of specialized hardware. The question of how to best schedule ML training jobs is non-trivial. ML training jobs present many unique characteristics compared to other computing workloads. First, ML training jobs vary in their degree of parallelizability (ability to scale out across servers), with some jobs being highly elastic and others being inelastic in their scalability. It is not clear how to allocate (share) hardware resources among jobs with such different characteristics. To make things more complicated, each training job is parameterized by a set of tuning parameters called hyperparameters. The parallelizability of an ML training job varies over time as the job hyperparameters change. Hence, systems which either rely on static resource reservations or existing heuristic scheduling policies are poorly suited for ML training jobs. Finally, the inherent work associated with an ML job is not a fixed quantity. Instead, ML jobs are generally trained until the model meets a desired level of accuracy. Hence, the existing scheduling theory that favors “short” jobs does not apply in this case. This proposal develops a new theoretic framework for modeling the dynamics of ML jobs running in a shared cluster. This framework will be used to develop scheduling and allocation policies that are specialized to ML training jobs, and to prove performance guarantees on these policies. 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 · 2024-06

Project Summary There is a critical need to develop therapies targeting Streptococcus pneumoniae (Spn): a major human pathogen contributing to global morbidity and mortality. Bacterial communication is a critical contributor to virulence, and we propose that bacterial extracellular vesicles (EVs) contribute to communication across bacteria as well as between bacteria and the mammalian host, and in doing so, contribute to pathogenesis and serve as therapeutic targets. This proposal is based on preliminary results that demonstrate not only that viable Spn produce EVs (pEVs), but also that these pEVs are associated with DNA and RNA. Our hypothesis is that the nucleic acids on pEVs are delivered to other bacteria, and that uptake of pEV RNA by bacterial cells promotes changes in gene expression and subsequent changes in phenotypes. Conclusion of this proposal will reveal the molecules tethering nucleic acids to pEVs, uncover the sequence and localization of pEV-bound RNA, and test whether and how internalization of EVs, and RNA on pEVs, modifies the transcriptional profile of the recipient bacterial cells. This work will set the stage for future studies focused on the role of pEV-nucleic acids in driving infection-associated phenotypes and host immunity.

NIH Research Projects · FY 2025 · 2024-05

Potable water is a major source of opportunistic pathogens that cause a variety of life-threatening infections in immunocompromised people, with a yearly US economic burden of $2.4 billion dollars. Biofilms ubiquitously coat surfaces within building premise plumbing that delivers potable water (e.g. water pipes, faucets, shower heads) and serve as a reservoir for opportunistic pathogens, including Pseudomonas aeruginosa. My postdoctoral studies and work from other laboratories implicate potable water as a source of P. aeruginosa that causes deadly respiratory infections in people with cystic fibrosis (CF). While P. aeruginosa persists in premise plumbing biofilms for months or years prior to infection, it is unknown how evolution in this niche impacts its pathogenic potential. The success of P. aeruginosa as a CF pathogen is owed in part to how it adaptively evolves in response to selective pressures in the airways, including nutrient limitation, antimicrobials, phagocytosis by immune cells, and microbiota interactions. Selective pressures in plumbing resemble those in the airways, including scarce nutrients, residual disinfectant (e.g. monochloramine), phagocytosis by amoebae, and polymicrobial competition. Preliminary data suggests evolution in plumbing impacts traits associated with infection of CF airways. The long- term goal of the candidate, Dr. Catherine Armbruster, is to establish an independent research program focused on how evolution of opportunistic pathogens in response to selective pressures faced in plumbing biofilms impacts downstream pathogenesis at different sites of infection. To achieve this goal, the immediate career objective of Dr. Armbruster is to obtain an independent faculty position using research proposed in this application as the foundation of her job applications. The overall research objective is to understand how evolution in response to selective pressures in plumbing biofilms impacts colonization and persistence in CF airways. The hypothesis is that low nutrients, polymicrobial competition, and exposure to monochloramine in plumbing biofilms select for P. aeruginosa traits that promote biofilm formation, protection from antimicrobials and immune responses, and increase competitiveness against microbes in CF airways. To test this hypothesis, two specific aims are proposed. The first aim identifies key P. aeruginosa pathways for survival in plumbing and how they evolve in response to specific pressures applied in a potable water biofilm model system. The second aim defines the subset of these adaptations that enhance fitness during host-pathogen interactions, using CF as a model system. These aims are expected to demonstrate how opportunistic pathogens are shaped by specific ecological and evolutionary factors prior to infection that impact downstream host-pathogen interactions. Successful completion of these aims will provide preliminary data for a competitive R01 application within two years of this award. Finally, proposed career development activities will complement Dr. Armbruster’s prior training. Dr. Bomberger, the postdoctoral advisor to Dr. Armbruster, is included as a scientific advisor because she is a leader in CF research with a strong track record of training successful independent academic scientists.

NSF Awards · FY 2024 · 2024-04

Understanding and designing how light interacts with materials play an important role in semiconductor technology, with applications in our daily lives ranging from solar energy harvesting and light-emitting devices to high-speed internet and high-performance computing. Engineering and enhancing this light-matter interaction are critical for advancing new technologies and even the new field of quantum information science and engineering (QISE). One strategy is to couple exciton, an optically excited electron and hole pair, to a nanoscale cavity and form exciton-polariton (EP), which is a half-material and half-light hybrid that inherits the advantages of both worlds. The photon nature allows us to manipulate the EP via optical engineering, and the exciton nature enables strong interaction. The light-matter interaction can be further enhanced in atomically thin semiconductors and engineered by stacking two of these atomic sheets together and precisely controlling their twist angle, forming a semiconducting moiré superlattice. In this proposal, we will design a systematic way to couple the nano-cavities with the semiconducting moiré superlattice to enhance the light-matter interaction to an unprecedented level, in which the device function can be drastically altered even by a single photon. This level of strong light-matter interaction can be utilized to implement photonic quantum simulations, a way to simulate new materials whose properties arise from complicated interactions between electrons. Strategically aligned with the National Quantum Initiative and Semiconductor Technology Initiative, this proposal will develop new course materials to train students for careers in high-demand, cutting-edge semiconductor, optical science, and QISE fields. Hands-on summer workshops on optics and nanofabrication for K-12 and under-represented minority students will be organized. The results from this proposal will be disseminated to both the scientific community and the general public to raise national awareness of the importance of QISE and semiconductor technology innovations. This proposal aims to develop a quantum nonlinear optical device platform to understand and engineer light-matter interactions in two-dimensional (2D) materials for analog quantum simulations. The project will accomplish a hybrid device that couples the robust excitons in a semiconducting moiré superlattice to nanophotonic resonators, forming a quasiparticle known as moiré exciton-polariton (EP), a half-material and half-light hybrid. The exciton in the semiconducting moiré superlattices formed by atomically thin transition metal dichalcogenides (TMDCs) can be tailored with even stronger interaction thanks to the electronic and excitonic flatbands. Therefore, the moiré superlattice hosts fascinating correlated insulating electronic states, and the exciton resonances are modified due to the moiré potential confinement. Strong coupling of the moiré excitons with the ultra-small mode-volume nanophotonic resonators and resonator arrays will lead to a unique moiré-EP platform for studying correlated excitonic physics and realizing nonlinear phonon-phonon interactions down to the single-photon level, paving the way to transformative quantum nano-optoelectronics such as analog quantum simulations. The proposed research will also transform the current state of power-efficient optical information processing and quantum optoelectronics. This proposal will develop education components well integrated with the proposed research to train students for the future workforce in semiconductor, optical science, and QISE fields. This proposal will develop learning opportunities on optics and nanofabrication for K-12 and under-represented minority students. The results from this proposal will be disseminated to a wide scientific audience and shared with the general public. 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 · 2024-04

A technology that enables biomolecules, such as nucleic acids and proteins, to be imaged at nanoscale precision throughout preserved 3-D specimens would enable a greater understanding of life processes and disease detection. This would be beneficial in fields such as neuroscience and oncology where many complex questions remain unanswered. The ability to map the locations of specific types of biomolecules within subcellular compartments would give insight into cellular organization and any altered state in disease. With the advent of expanding microscopy (ExM), it is now possible to obtain nanoscale images using only a diffraction-limited light microscope. This simple yet novel approach uses water-swellable polymers to physically expand biological specimens to be imaged at approximately 70 nm resolution. While the protocol for expansion and the retention of intracellular antigen have progressed rapidly since ExM was first developed, currently available methods are limited by linear expansion of 4- 5 times of their original size. A higher expansion factor is needed to reveal the subtle changes in the size, shape, and texture ratio of subcellular organelles in health status or disease. More importantly, almost all the current ExM methods require a specific anchoring step to ensure targeted biomolecules are covalently linked to the newly synthesized hydrogel. A universal biomolecule anchor that works for thick tissues remains elusive. Additionally, few existing ExM approaches can expand diverse tissue types other than the brain without losing most of the epitopes. Thus there is tremendous pent-up demand for a method of nanoscale imaging for extended 3-D specimens and/or with highly versatile molecular contrast. Given its potential impact, we now propose a new framework called Magnify to bring the power and versatility of ExM to next generation. We will focus on the following aspects: (1) Develop a robust one-shot 11× expansion microscopy with universal biomolecule retention; (2) Further develop Magnify for expanding mm-scale thick tissues and whole small animals; (3) Extend Magnify for super-resolution vibrational imaging, including label-free, metabolic and multicolor super-resolution imaging. We will demonstrate the potential of Magnify as a powerful tool for mapping subcellular proteomic changes in diverse tissues, cells, and organelles by visualizing molecular spatial patterns at unprecedented high spatial resolution throughout preserved specimens.

NIH Research Projects · FY 2025 · 2024-03

PROJECT SUMMARY/ABSTRACT Type 1 diabetes is an exceptionally challenging disease to manage, requiring a host of complex behaviors to manually regulate one’s blood glucose. Importantly, these disease management behaviors must be responsive to the ever-changing needs of one’s day-to-day environment; thus, one’s diabetes management must be fundamentally adaptive. While the psychological burden of diabetes is well documented, investigations into the implications of psychosocial factors like mood, stress, and social relationships on blood glucose have tended to focus on broad cross-sectional or between-person findings. However, research has failed to adequately capture how these psychosocial factors impact blood glucose in real-time (e.g., day-to-day or hour-to-hour), and thus have limited implications for interventions to reduce disease burden. The proposed research will address this gap by combining ecological momentary assessments (EMAs) with continuous glucose monitoring (CGM) data among 50 adults with type 1 diabetes. Participants will complete six brief surveys per day for fourteen days, reporting on mood, diabetes-specific and overall stress, the quality of their social interactions, self-care behaviors, food intake, energy expenditure, and insulin dosing. Given recent surges in CGM use, blood glucose data will be downloaded from participants’ existing devices; the mean, coefficient of variation, and time in, above, and below range will be calculated across the two hours following each EMA. Aim 1 will identify links of mood, diabetes distress, and overall stress to blood glucose outcomes across the subsequent two hours. It is hypothesized that negative mood and stress will lead to worse glucose outcomes, while positive mood will lead to better glucose outcomes. For links that are supported, self-care will be investigated as a potential mechanism underlying these relations. Aim 2 will examine the effects of conflictual and supportive social interactions on blood glucose. It is hypothesized that conflictual interactions will lead to poor glucose outcomes, while supportive interactions will lead to improved glucose outcomes. Again, self-care will be investigated as a mediator for any observed links. Exploratory Aim 3 will investigate whether within-person links of these psychosocial factors to blood glucose differ based on gender and overall relationship quality. It is hypothesized that females will overall show stronger relations than males, and that higher quality relationships will buffer against the negative effects of conflictual interactions. This research will contribute to our understanding of the real-time barriers to diabetes care, paving the way for just-in-time interventions to improve disease outcomes. In support of this project, the Principal Investigator will collaborate with a team of leaders in the field. This mentorship team will facilitate additional training opportunities in the patient experience of diabetes and CGM use, the endocrine functions underlying diabetes and blood glucose regulation, and the quantitative methods required for dense CGM data. Ultimately, the training plan and proposed research will enable the PI to establish her future career as a researcher at the intersection of diabetes and digital health.

NIH Research Projects · FY 2026 · 2024-03

Cell and tissue mechanical properties play critical roles in physically shaping animals, organs, and tissues during development, growth, maintenance, regeneration, and disease. Early embryonic development and later growth utilize cell-generated physical forces to sculpt the body and organs. The micro-architecture and composition of these tissues is spatially, geometrically, and temporally complex. Adjacent tissues can differ widely in elastic modulus and can change greatly in a few hours as cells differentiate and gene expression changes. However, real-world knowledge of these properties is limited to a few model systems, where physical access and size of samples allow direct measurement. This gap is not due to lack of effort. Numerous technologies have been developed to sense these properties, but the applicability of these technologies has been limited by the need for samples with regular shape or with a quasi-planar geometry that is amenable to scanning with table-mounted mechanisms. To advance our understanding of embryonic development, there is a great unmet need for mechanical testing instrumentation that can handle more complex 3D geometry. We propose to address this problem by developing a handheld tissue force microscope. Our team has experience in atomic force microscopy and related techniques, and has developed an actuated handheld micromanipulator that enhances accuracy by performing active compensation of physiological hand tremor and that incorporates force sensing, highly accurate optical tracking of both the handle and the manipulated instrument tip, and camera-based visual tracking through the optical microscope. Based on our expertise in this area, we propose to develop a convenient and easy-to-use active handheld instrument that can perform dynamic mechanical analysis of embryos, organoids, and small tissue samples. Tremor compensation provided by the instrument will enable the user to precisely target desired locations for testing. Frequency sweeps will be performed automatically by the instrument. The sensing capabilities of the instrument will automatically ensure that the sinusoidal oscillation is applied perpendicularly to the local tissue surface. In data analysis, optical tracking of the instrument handle will allow the system to automatically correct for any residual motion disturbance that remains after active tremor compensation. Visual tracking will automatically register local point measurements to photographic images of the sample, enabling scans of any area of interest, up to and including the entire surface if desired. The specific aims are to develop a prototype capable of manipulating with accuracy of approximately 1 μm. (This will represent the “coarse” portion of a coarse-fine manipulation system); to develop a “fine” manipulator and sensing system for the tip of the instrument; and to integrate and evaluate the full prototype.