University Of North Texas

NSF Awards · FY 2026 · 2026-10

This three-year Research Experiences for Undergraduates (REU) site at the University of North Texas will support 10 students for 10 weeks each summer and train them to build artificial intelligence (AI) systems capable of sharing knowledge across domains through vector embeddings. Vector embeddings translate information (e.g., images, text or sound) into numerical data. These data are then converted into lists that allow the numbers to convey context and meaning. Current AI systems who are working in areas such as visual object recognition, speech recognition, and understanding natural language require extensive training, with much of the learned knowledge locked within the structure of the system, which is then difficult to reuse. However, students in this program will focus on creating and leveraging highly-trained AI systems to represent information in ways that preserve the nuanced understanding learned by these systems and make it accessible for other applications. This REU brings together an interdisciplinary team to support projects that showcase the benefits of AI systems that can exchange and reuse learned knowledge. Early in the program, each student will identify a project domain and a faculty advisor with whom they can work. Students will also participate in a long-standing AI summer research program integrating current university students and external REU students to facilitate collaboration across departments and student expertise. Specifically, the training in this REU will allow students to more efficiently represent and transfer the knowledge acquired by self-supervised deep learning models. Each year, students will create vector representations of entities that appear across multiple domains, apply these embeddings to improve prediction models, and systematically evaluate, document, and contribute them to a shared, reusable knowledge base. These efforts are coordinated through common documentation, evaluation, and sharing practices that enable comparison and reuse of embeddings across projects. For the first five weeks, the students will be exposed to different embedding strategies and machine learning applications that use them, then transition to developing, testing, and refining their individual research efforts in the last five weeks. This REU will help prepare a workforce of students not only adept at using deep learning models but also capable of extending their functionality through reusable and shareable representations. Additionally, this project will train a diverse range of students from college partners with limited research resources to work in interdisciplinary teams at a Carnegie R1 research institution. 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 2026 · 2026-08

Many critical scientific and societal challenges involve interactions among groups of entities that cannot be adequately represented as simple pairwise relationships. For example, chemical reactions, biological signaling pathways, social group dynamics, and epidemic spread involve simultaneous multi-way interactions among more than two entities. Hypergraphs, a mathematical generalization of traditional graphs, provide a more accurate representation by allowing a single edge to connect any number of entities. Despite rapidly growing adoption of hypergraph-based models across biology, health sciences, social networks, and artificial intelligence, researchers lack the scalable, comprehensive, and user-friendly software needed to analyze hypergraphs effectively. This gap forces scientists to rely on simplified graph models that risk losing critical higher-order relationships in their data, potentially leading to incomplete scientific conclusions. This motivates the project - CHAI (Cyberinfrastructure for Hypergraph-based Analysis and Innovation), an open-source parallel software framework that addresses this critical gap. CHAI serves a broad scientific user base through a tiered interface accommodating users ranging from domain scientists who need ready-to-use analytical functions, to intermediate users who wish to tune algorithms, to advanced developers creating entirely new methods. For real-world validation, the CHAI team collaborates with researchers from social network analysis, bioinformatics, food web ecology, additive manufacturing, unmanned aerial vehicles, and cyber-physical systems. The CHAI project aims to develop three foundational technical innovations: (i) a unified data structure that efficiently supports both static and dynamic hypergraph representations on high performance computing platforms including GPUs, (ii) a compact motif-based representation that reduces memory requirements and accelerates hypergraph analysis, and (iii) an extensible parallel algorithm development framework for hypergraphs, enabling researchers to build and contribute new domain-specific algorithms. Together, these innovations enable CHAI to provide scalable and accurate hypergraph analysis. The developed software will be publicly distributed through GitHub, portable software containers, and a graphical drag-and-drop workflow interface that requires no specialized programming expertise, thereby maximizing accessibility across scientific disciplines. CHAI supports the training of graduate students and postdoctoral researchers in parallel algorithm design, high-performance computing, and scientific software development, contributing to building the next generation of computational scientists. Outreach activities extend research opportunities to a broad range of undergraduate and high school students. Community-building through workshops, tutorials, and conference mini-symposia will establish CHAI as the standard cyberinfrastructure platform for hypergraph analysis by enabling advances across a wide range of scientific domains that depend on accurate modeling of complex, multi-way interactions. 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 2026 · 2026-06

The total global food demand is projected to increase significantly over the next couple of decades, yet plant diseases and invading species lead to considerable losses in global crop production. Although some species of fungi can cause plant diseases or produce harmful toxins that affect plants, humans, and the environment, other fungal species produce small molecules that help protect plants and humans by acting as fungicides or antibiotics. This project aims to investigate how the little-studied fungus Colletotrichum spinosum interacts with the “world’s worst weed” - the invasive plant Xanthium spinosum. Colletotrichum spinosum has the genetic potential to produce small molecules that might act as “chemical weapons” against certain plants, potentially acting as herbicides against weeds, but very little is known about these molecules. By identifying and studying these small molecules, this research will enhance our understanding of fungal-plant interactions and how small molecules are key to this interaction. The findings may lead to new ways of protecting crops from pests and diseases and will also contribute to the development of new biotechnological solutions for agriculture. Additionally, the project will provide valuable educational opportunities for undergraduate students, allowing them to participate in cutting-edge research and contribute to scientific discoveries. Through hands-on involvement in the research process, students will gain real-world experience, including collaborations with local artists to communicate the findings in engaging ways, that will help prepare them for future careers in science and technology. Overall, this research will advance scientific knowledge, foster innovation in biotechnology and artificial intelligence, support workforce development, while addressing important challenges in agriculture and environmental sustainability. Xanthium spinosum is an invasive noxious weed that affects essential crops such as corn, tomatoes, and sugar cane, and is difficult to contain. The fungus Colletotrichum spinosum has demonstrated ability to kill young Xanthium spinosum plants by an unknown mechanism. This project aims to identify and characterize the specialized metabolites produced by Colletotrichum spinosum and assess their role in the fungus’ interaction with the weed, such as being essential for pathogenicity, for example. The research will uncover previously unreported metabolites using genome mining, metabolomics, transcriptomics, comparative genomics, and heterologous expression in Aspergillus oryzae (a generally regarded as safe host). Once discovered, purified metabolites will be structurally elucidated using mass spectrometry and nuclear magnetic resonance techniques and tested for bioactivity against Xanthium spinosum and other plants to explore whether certain metabolites act independently of the fungal host. In parallel, a team of scientists at different career stages, will be trained to perform the research and disseminate results using traditional channels and more community-focused approaches. This research has the potential to uncover new antifungal or herbicidal compounds with applications in agriculture and plant protection. The project will also contribute to a broader understanding of fungal metabolic diversity and its ecological roles. Ultimately, it aims to provide valuable insights into how fungi adapt to their environments and could lead to new biotechnology strategies for invasive plant species and crop protection. 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 2026 · 2026-06

This project outlines a two-day workshop to be held in Norman, Oklahoma, focused on developing a community-wide roadmap for leveraging ontology-driven processes in advanced manufacturing (AM) to facilitate the integration and sharing of diverse, heterogeneous materials data - from synthesis and processing to characterization and performance. The central challenge is that materials data is often stored in disparate, isolated systems, creating "data silos" that hinder the ability to connect data with its critical contextual information, such as the material-process-structure-property relationship. This fragmentation is a fundamental bottleneck that slows the pace of innovation and discovery. The conference objective is to address this bottleneck by establishing a straightforward and systematic digitalization workflow that generates high-quality, semantically structured, and linked data adhering to the FAIR principles (Findable, Accessible, Interoperable, Reusable). 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 2026 · 2026-05

Many high school students have little opportunity to understand how artificial intelligence (AI) works or to develop the computational thinking skills needed to engage with it critically. This makes it essential that teachers be not only knowledgeable about AI but also skilled enough to guide student inquiry into it. Research Experiences for Teachers (RET) programs represent one promising pathway, yet fewer than 40% of participants ever translate what they learned into actual classroom practice. This shortfall stems from several recurring challenges: research projects are often too distant from real teaching contexts, support ends with the summer program itself, and teachers are treated as beginners rather than as experts in the learning sciences. This project addresses these challenges by requiring teachers to work side-by-side with professors as genuine research partners. Together, they develop AI tools designed specifically to improve how computational thinking and AI are taught in high schools. Because teachers are central members of the development process, the resulting tools reflect actual classroom needs. Participating teachers also gain the knowledge and research skills needed to mentor their own students in AI and computational thinking projects. This project establishes a teacher-researcher co-design model in which ten high school teachers annually conduct six-week AI research experiences at the University of North Texas (UNT). Beyond the summer, the UNT team provides continued support to help teachers transfer what they have learned into their classrooms. The co-design model requires teachers to contribute their pedagogical expertise directly to the research projects they pursue alongside faculty, positioning them as active AI co-creators and ensuring that research is grounded in classroom realities from the outset. The project will also generate new theoretical knowledge about how domain expertise outside computer science and AI shapes complex technology design, and how non-technical experts can meaningfully participate in AI research. Longitudinal study of co-design processes and student interactions with the resulting systems will deepen understanding of computational thinking development and effective human-AI collaboration. Over three years, 30 teachers will be transformed into AI co-designers, each reaching several hundred students annually. Open-source tools and curricula produced through the project will be made available to schools nationwide, and the co-design model itself will demonstrate how pedagogical expertise can drive AI research that genuinely serves educational needs. 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 · 2026-05

Project Summary/Abstract Noise-induced hearing loss is one of the most prevalent and irreversible forms of sensorineural hearing loss, affecting approximately 40 million American adults. It significantly reduces quality of life and is associated with cognitive decline, depression, and social isolation. While hearing aids and cochlear implants partially restore auditory function, they fail to replicate natural hearing and may introduce further auditory damage. Prevention remains a promising pathway to preserving auditory health, especially for individuals exposed to occupational or therapeutic risks, such as firefighters, military personnel, and chemotherapy patients. Among the various approaches studied for hearing preservation therapeutic hypothermia (MTH) has emerged as the most clinically translatable option. However, evidence suggests current mild therapeutic hypothermia devices deliver suboptimal spatial temperature distributions and lack the precision needed to uniformly cool critical cochlear structures. A transformative approach is used in this project to enhance mild therapeutic hypothermia for hearing loss prevention. Computational modeling, experimental validation, and device development will be integrated to design a new non-invasive cooling strategy that accounts for anatomical variability and vascular heat transfer. Specifically, a multiphysics model of the cochlea and surrounding vasculature will be constructed to evaluate how blood flow affects cooling dynamics during mild therapeutic hypothermia. Using insights from simulation and cadaveric studies, an advanced cooling device will be fabricated to deliver targeted, uniform hypothermia to the cochlear region. Rapid onset, spatial uniformity, and safety will be prioritized, minimizing adverse effects such as overcooling. Comparative analysis will be conducted against existing devices using ex vivo and in vitro methods to quantify efficacy in achieving desired thermal profiles. The results are expected to define the thermodynamic parameters and spatial targets essential for optimal hypothermic neuroprotection of the cochlea. In addition to technical advancements, the project will serve as a training platform for undergraduates and graduate students, offering interdisciplinary exposure to computational modeling, experimental bioengineering, and auditory neuroscience. By addressing key limitations of current mild therapeutic hypothermia systems, this work will lay the foundation for next-generation cooling devices with the potential to significantly reduce the incidence of noise-induced hearing loss. These contributions will support future clinical investigations and provide a reproducible research model for non-invasive auditory preservation techniques. The in-silico model developed here may also serve as a virtual microscope to explore and optimize emerging hearing preservation strategies.

NIH Research Projects · FY 2026 · 2026-04

Project Summary The Digital Health Research and AI Training Program at the University of North Texas (UNT) provides a hands-on summer research experience for high school science teachers in the Dallas-Fort Worth (DFW) region. This initiative immerses educators in digital health, AI-driven healthcare solutions, and wearable health technologies, equipping them with the skills to integrate real-world digital health applications into STEM curricula. Under the guidance of faculty mentors from engineering, health sciences, and education, participating teachers conduct research in active digital health and AI labs, gaining direct exposure to cutting-edge work in embedded systems, biosensors, medical imaging, and AI-powered diagnostics. The program places ten high school teachers in research laboratories, where they engage in structured study design, bio-signal data collection, and machine learning analysis of physiological signals such as ECG, PPG, and motion data. Educators also participate in targeted workshops designed to provide foundational knowledge in AI, digital health research, and embedded systems, ensuring they have the necessary prerequisites to apply these concepts in their teaching. Through hands-on training, teachers develop wearable health devices, implement real-time bio-signal processing, and explore AI-driven digital health applications, culminating in collaborative research studies with university faculty. Using a Community of Practice model, teachers work alongside faculty and peers to develop STEM lesson plans, curriculum maps, and instructional materials aligned with Next Generation Science Standards (NGSS) and Texas Essential Knowledge and Skills (TEKS). This program directly addresses the lack of hands-on AI and digital health research training opportunities for high school educators in the DFW region, bridging the gap between academic research and STEM education. By integrating biosensors, real-time signal processing, and embedded AI concepts into classroom instruction, educators empower students to engage with real-world biomedical challenges and explore careers in digital health, healthcare AI, and digital health technologies. The program includes a comprehensive evaluation plan assessing curriculum improvement, classroom implementation, and student engagement in biomedical AI topics. Teachers participate in pre- and post-program assessments, classroom observation studies, and long-term tracking of student learning outcomes. The initiative also incorporates structured dissemination opportunities, enabling educators to present their research experiences and curriculum materials at regional STEM education conferences, school district professional development workshops, and AI in healthcare summits. This initiative enhances teacher expertise in digital health and AI, expands the STEM workforce pipeline, and fosters student interest in healthcare innovation by providing a sustainable professional development model. Through collaborations with local research institutions, industry partners, and K-12 networks, the program ensures a lasting impact on high school STEM education in the DFW region.

NIH Research Projects · FY 2026 · 2026-03

Project Summary The demand for novel synthetic chiral molecules in biomedical research and drug development is increasingly urgent. Chiral compounds and natural products play critical roles in biological and pharmaceutical applications. While direct conversion of C‒H bonds into functional groups presents new synthetic opportunities, achieving enantioselectivity for complex molecules remains a challenge. Recent advances in C–H activation have led to highly efficient catalytic systems, yet their widespread application in synthesizing complex organic building blocks, natural products, and pharmaceuticals remains underdeveloped. This research proposal aims to address these challenges through the development of ruthenium(II)-catalyzed enantioselective functionalization of arene and alkene derivatives, and their synthetic applications for asymmetric syntheses of natural products and structurally complex molecules. Readily accessible chiral catalysts and inexpensive ruthenium resources will be used for the development of practical catalytic systems and synthesis. Specifically, it targets the enantioselective functionalization of ortho- and meta-aryl C‒H bonds, leveraging a non-covalent interaction-based chiral induction model to enhance stereochemical control. Key objectives include the total synthesis of dendrofalconerol A and asymmetric synthesis of dinoxyline, utilizing enantioselective C‒H activation as a pivotal step. Additionally, the project aims to develop intermolecular enantioselective ortho- and meta-C–H functionalization of arenes, potentially yielding novel axially chiral indole-aryl compounds with applications in ligand design and pharmaceuticals. Furthermore, the research seeks to develop new enantioselective alkenyl C–H activation systems, exploring new hydrovinylation routes for generating unconventional products. This project offers undergraduate students at the University of North Texas hands-on experience in the cutting-edge field of C–H functionalization. By focusing on enantioselectivity in the catalytic processes and target molecule syntheses, students will deepen their understanding of chemical principles while gaining valuable skills for careers in biomedical, pharmaceutical, and health- related sciences. Engaging in Ru(II)-catalysis research, they will work at the intersection of catalysis, organic chemistry, and medicinal sciences, contributing to impactful discoveries and mastering the latest synthetic technologies. This experience will provide them with both technical expertise and the mindset needed to lead in the evolving landscape of pharmaceutical research.

NSF Awards · FY 2026 · 2026-01

Humans depend on aquatic habitats for numerous ecosystem services including drinking water, food production, and recreation. The provision and maintenance of these services is underpinned by ecological processes that change under different ecological contexts and human uses. Ecosystem energy flow, the movement of energy from basal species such as plants and algae up food webs to top predators, is a crucial ecological process that can influence the number of different species, their biomass, and population persistence across space and time. However, understanding the patterns and drivers of ecosystem energy flow can be a difficult task because different ecosystems have different species with unique ecological and evolutionary histories and the collection of ecological information is labor intensive. This project overcomes these challenges by focusing on a universal trait, body size, that is tied to many biological processes such as individual metabolism, lifespan, movement, and feeding and therefore has an important role in structuring ecosystems. For example, the range of body sizes and the rate of decline of large individuals relative to small individuals represents an ecosystem-level proxy for ecosystem energy flow. Using information on the range of body sizes, their relative abundance, and a suite of additional variables, this project will examine the mechanisms that drive energy flow through ecosystems. This project will accomplish this goal by leveraging crucial, continental-scale infrastructure supported by NSF, the National Ecological Observatory Network (NEON), and the associated stream and river ecosystems across the United States to determine the environmental and ecological drivers of energy flow and their importance in the provision and maintenance of crucial ecological functions in aquatic ecosystems. This project will train a postdoctoral scholar, graduate student, and multiple undergraduate students and will provide publicly available and permanently archived data and code of body size analyses across the United States. The relationship of declining abundance (N) with increasing body mass (M) is among the most consistent patterns in ecology. More formally, the relative rarity of large individuals across all trophic levels reflects not abundance or biomass per se, but the frequency distribution of individual body sizes in an ecosystem known as the individual size distribution (ISD) or community size spectrum. The ISD is described by a power law of the form f(x) ~ (cM)^, where c is a constant, M is individual mass, and λ measures the rate of decline in frequency from the smallest to largest individual in an ecosystem. Across ecosystems and environmental gradients, values of λ fall in a remarkably narrow range, from -1 to -2, a consistency that has motivated using the ISD as a universal ecological indicator. A common understanding is that the ISD reflects the efficiency of energy flow from small individuals to large individuals in size-structured food webs. Here, less negative values of λ represent more efficient energy transfer. Further work has attempted to identify the mechanisms behind the ISD and a bottom-up model of energetics proposes the relationship between N and M is tied to individual energetics and metabolic scaling with mass, (M)^α. Refinements on the energetic perspective have incorporated trophic structure and the loss of energy among trophic levels and propose a trophic level correction on the basic physiological model by including terms for the trophic transfer efficiency (TTE) and the predator-to-prey mass ratio (PPMR). Together, this model suggests the exponent of the power law should be a function of the metabolic mass scaling α, TTE, and PPMR.   The research will determine the macroecological patterns of the underlying model variables and its ability to explain the distribution of individual body sizes in ecosystems. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-11

Management of waste plastics and other polymeric materials has become a critical environmental issue in recent years. Effective mitigation has generated research and development efforts to “up-cycle” waste plastic to more valuable chemicals or embrace “circular” plastics process technology. When combined with more efficient manufacturing processes, the re-use of plastics, in any form, stands to contribute significantly to net-zero carbon emissions. To those goals, the project investigates a novel catalyst design to break down polyolefin-based plastics (e.g., polyethylene or polypropylene) into smaller molecules that can be used as building blocks for either the remanufacture of polymeric materials or production of commodity chemicals. The novel catalyst design improves overall manufacturing efficiency by lowering energy requirements and directing the chemistry towards desired products. Beyond the technical aspects, the project includes educational and outreach initiatives promoting STEM opportunities for K-12 students, and training opportunities for both undergraduate and PhD students. The project investigates an aspect of supported heterogeneous supported metal catalyst design that is typically not considered or controlled, i.e., the relationship between particle size and particle density. Specifically, this will be accomplished by varying the particle size and site density of rutile-RuO2 supported on rutile-SnxTi1-xO2 where x ranges from 0 to 1 and to study the effects of these parameters on polyolefin hydrogenolysis. The project leverages the lead investigator’s skills in catalysis with polymer characterization and synthesis capabilities at the University of Akron, thus enabling new methods to analyze polymer upcycling kinetics on well-controlled polymer samples. The RuO2 particle size will be controlled by varying the calcination temperature and/or the misfit strain during epitaxial growth of RuO2 on the rutile-oxide support. Site density will be controlled by the metal precursor loading. Polyolefins (POs) are ideal substrates, and PO hydrogenolysis is a well-chosen probe reaction for elucidating the kinetics as related to the effects of particle size vs density, given the large substrate size and potential for oligomers to bridge multiple particles. The study thus carries practical implications for polymer upcycling, benefitting from new “drop-in” catalyst technology potentially applicable to a broad range of catalytic processes. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-10

Today, Europe is leading the way in developing laws to manage the increasing range of harm that artificial intelligence (AI) can and does create. Europe's AI Act focuses on regulating how its businesses develop and use AI. Many experts worry that this heavy-handed intervention into private enterprise will stifle AI innovation, including, ironically, innovations that could ultimately make European citizens safer. The United States urgently needs its own legal approach for managing AI harms. This project's novelty lies in developing an objective standard for deciding who is at fault after an AI harm occurs. This approach avoids Europe's tussle between law and computing because an after-the-harm standard does not directly interfere with business operations. Rather, it provides businesses the space to innovate and the incentive to figure out how best to design accountable software systems that minimize avoidable AI harms. The objective of this project is to bring accountability for AI without impeding the businesses' and researchers' ability to continually innovate and lead in AI. To do so, this project aims to design an objective-fault standard for AI that does not prohibit or censure any AI behavior outright but instead compares AI’s behavior with an external negligence benchmark. Then, by calibrating the benchmark’s standard to the social value and the current safety profile of the AI conduct at issue, the AI law could be applied flexibly, progressively, and across broad domains. The researchers plan to attain this objective by (i) laying the legal foundation for a negligence standard for AI, (ii) developing AI negligence benchmarks for three representative applications, and (iii) evaluating this new standard against the European AI Act. Upon completion, this research would support establishing a distinctly American alternative to the European style AI 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 2025 · 2025-09

This I-Corps project is based on the development of a new class of biometric technologies that securely identifies individuals without physical contact. Current biometric systems, such as fingerprint or facial recognition, often face difficulties in everyday settings due to changes in lighting, user appearance, and other environmental factors. These challenges result in errors and delays, creating security risks and inefficiencies in critical areas such as healthcare, finance, government, and infrastructure. This solution applies advanced machine learning techniques to improve the way biometric systems learn and recognize unique features, increasing accuracy, reliability, and scalability for broad use. The technology performs effectively in real-world environments, enabling fast and secure identity verification without requiring physical interaction. The solution addresses the growing problem of identity theft and unauthorized access, which impacts millions of individuals annually. By reducing these risks, the technology enhances public safety, protects sensitive data, and increases operational efficiency. The technology also lowers the need for manual identity checks, saving time and resources. The project advances national interests by fostering scientific progress in secure digital identification, supporting economic stability, and strengthening infrastructure essential to public welfare. This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of the technology. This solution is based on the development of a new class of biometric algorithms that leverage advanced, generative, artificial intelligence-based data augmentation pipelines; deep biometric feature learning techniques using supervised and self-supervised learning; and sophisticated decision post-processing frameworks utilizing decision uncertainty estimation techniques. These innovations improve the accuracy, robustness, and scalability of visual contactless biometric systems, especially in uncontrolled operational environments. The approach overcomes challenges related to environmental variability, user presentation differences, and large-scale deployment by integrating adaptive learning models capable of generalizing across varying environmental conditions. This solution advances the state of the art by introducing advanced data augmentation, feature extraction, and decision post-processing pipelines, thereby enhancing reliability in dynamic, real-world applications. Users benefit from highly reliable identity verification systems with faster processing times and reduced false acceptance and rejection rates. The technology supports secure, user-friendly, and contactless authentication, ideal for sectors such as healthcare, finance, and government. It contributes to the broader field of biometric solutions that enhance universal accessibility and operational resilience. Broader adoption of this innovation may reduce identity fraud risks, limits unauthorized access, and improves the efficiency and security of authentication workflows across a wide range of applications. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-09

PROJECT SUMMARY In the United States, it is estimated that >100,000 people will die from non-small cell lung cancer (NSCLC) in 2024. Emerging targeted therapies, such as checkpoint immunotherapies have resulted in significant success; still, 50% of patients treated with anti-PD1 or anti-CTLA4 have poor clinical efficacy. The effectiveness of immunotherapies has been strongly correlated with the abundance of immune T cells in the tumor stroma. Indeed, increased T cell infiltration is a strong prognostic indicator across 17 solid cancer types, including NSCLC. Nonetheless, immunotherapies have largely remained ineffective in tumors that have reduced T cell infiltration, also called immune-cold tumors. Therefore, understanding the mechanism(s) that lead to reduced tumor infiltrating T cells is of significant interest in the treatment of immunosuppressive NSCLC. Despite an extensive understanding of T cell suppression mechanisms in immune-cold NSCLC, current therapies have shown modest clinical benefit, indicating that additional mechanisms contribute to this process. One such emerging mechanism is tumor-derived extracellular vesicles (EVs), however, identification of specific EV subpopulations responsible for T cell suppression has remained a challenge. Overcoming this challenge, we have identified a specific subpopulation of NSCLC-derived EVs, called TGN46+ EVs, that strongly inhibit the proliferation of T cells. In an immunosuppressive NSCLC model (H358), total H358 EVs inhibit T cell proliferation; however, specifically depleting TGN46+ EVs from total H358 EVs significantly reduces their ability to suppress T cells. Our findings are clinically relevant, and we have further verified them in two independent NSCLC models as well as in prostate cancer, which are well-known to form immune-cold tumors. Given these findings, we hypothesize that TGN46+ EVs play a key role in modulating T cell tumor infiltration in immune-cold tumors. In the proposed work, we plan to evaluate how this single EV subpopulation, i.e. TGN46+ EVs, influences T cell infiltration and overall tumor microenvironment. To evaluate this, we will use (i) a humanized mouse model that supports the development of the human immune system and allows the study of human-specific tumor-immune interactions; (ii) a human NSCLC model (HCC827) that exhibits high T cell infiltration in humanized mice, and (iii) EVs from an immunosuppressive NSCLC model (H358). We will evaluate how T cell infiltration and abundance of other immune cell types in HCC827 tumors change upon introduction of (i) total H358 EVs (which includes TGN46+ EVs), or (ii) H358 EVs depleted of TGN46+ EV subpopulation, in the tumor stroma. Following the successful completion of this study, we anticipate that we will have evaluated EV-specific modulation of the tumor microenvironment and its physiological implications. These accomplishments would have an enormous positive impact on our understanding of intercellular communication in immune-cold NSCLC. In the long term, these insights have the potential to inform the development of effective strategies to overcome immunotherapy resistance, thereby benefiting a considerable number of lung cancer patients.

NSF Awards · FY 2025 · 2025-09

Language is a fundamental part of all human societies. Studies of language evolution, from ancient forms to modern languages, contribute to our understanding of populations and the development of modern societies. Recent developments in AI technology make it possible for us to solve questions we have not been able to solve such as reconstructing earlier versions of today’s languages using AI and Artificial Neural Networks. This project creates such AI models to study language evolution using large amounts of data from many past and present languages and advances the development of AI tools in STEM. Reconstructing ancestor languages and earlier states of known languages is pivotal in understanding pathways of linguistic change. The standard method of reconstructing such protolanguages is the comparative method, which works by analyzing systematic correspondences across attested languages of the same family; however, it has several limitations in its present form. The goal of this project is to develop an Artificial Neural Network approach to comparative reconstruction. AI models can take in large and diverse linguistic data and detect complex associations in descendant languages to make accurate inferences about ancestor forms. The project aims at building Neural Networks that can run on large datasets with multiple language families and learn to reconstruct proto-forms by harnessing cross-linguistic patterns. Different model types, architectures, and dataset approaches will be explored and afterwards applied to current reconstruction problems in various language families. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

The Large Cardinal Axioms (LCAs) are extensions of the standard axioms of set theory, Zermelo-Fraenkel set theory with the axiom of choice (ZFC). LCAs are designed to settle all natural theories that are independent of ZFC. This is Gödel's program in set theory. How can one test for “correctness” of an LCA? The inner model program, a major program in modern set theory, justifies correctness by constructing canonical models for LCAs much like the natural numbers are the canonical model for the Peano Axioms of arithmetic (PA). The canonicity of the models justifies the correctness of the LCAs much like the canonicity of the natural numbers justifies the correctness of PA. This research project contributes to the inner model program by advancing the state of the current knowledge regarding canonical models for LCAs and their relationship with other foundational frameworks of set theory. The project provides research opportunities for graduate students. This project builds on and expands the PI’s previous work on computing consistency lower bounds for the Proper Forcing Axiom (PFA), on studying canonical models of AD^+ up to the minimal model of the Largest Suslin Axiom (LSA), and on the Sealing phenomenon concerning universally Baire sets. These are important theories in our set-theoretic landscape. The PI plans to explore further various aspects of Sealing (particularly a weak form of Tower Sealing and its implications) and develop further techniques of the core model induction with the eye towards determining the consistency strength of various fragments of Martin’s Maximum. The PI plans to study the theory of short-tree strategy mice for the least-branch hierarchy; this not only completes the general structure theory for least-branch hod mice but also will have potential applications, particularly in core model induction contexts. The PI also plans to continue the ongoing joint work with W. Chan and S. Jackson on the general analysis of combinatorial structures of determinacy; this allows the PI and his co-authors to understand deeper how sets are related to one another in this context. This, in particular, leads to the discovery and resolution of the ABCD Conjecture, solutions to several classical descriptive set theoretic conjectures, and the study of general cardinalities and cofinalities of sets in choiceless contexts. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

The project involvew research in two main directions in descriptive set theory, which is a branch of mathematics in which modern set-theoretic methods are used to develop the theory of subsets of the real line and related structures which are in turn the fundamental objects of mathematical analysis used throughout mathematics and its applications. One of the directions concerns developing the structural theory of sets in models of the axiom of determinacy. This is important as this axiom holds in various mathematical universes which interact with the ``true'' universe and also because it is giving us the theory of definable objects which is a central goal of descriptive set theory. A second direction concerns the theory of definable equivalence relations. This is a relatively recent area of descriptive set theory which has been investigated extensively the past several decades. This area interacts heavily with a number of different areas of mathematics and provides a unifying framework for them. There are a number of fundamental problems that remain open in this area concerning the structure of countable Borel equivalence relations in particular, but significant progress has been made in the last few years. The project will further develop these methods with a goal of attacking some of these problems. This project will involve graduate students. Specifically, the combinatorics of non-wellordered sets in models of determinacy is a main line of the first direction of the proposal. An example is Chan's recent proof of the ``ABCD'' conjecture describing the relation between cardinalities of the form A^B for ordinals A and B (the conjecture roughly speaking asserts that the only relations between these cardinalities are the obvious ones). The principal investigator along with Chan and Trang have isolated a general principle about infinity Borel sets which also proves the ABCD conjecture and seems likely to have other applications. There are, however, many fundamental questions concerning these non-wellordered cardinalities that remain open which the proposal plans to investigate. The theory of countable Borel equivalence relations has shown much progress in recent years. For example, one of the central questions is the hyperfiniteness question which asks which groups have the property that all of their Borel actions generate only hyperfinite (an increasing union of equivalence relations with finite classes) equivalence relations. Recent work has shown that this class extends to include the polycyclic groups, so, in particular contains finitely generated groups of exponential growth. How much further this extends is an open problem. Some new techniques have been introduced recently by the principal investigator and co-authors which will help answer several structuring questions for actions of Z^n and other fairly simple groups. This hopefully leads to marking the boundary of what types of Borel or continuous structurings can be done in an invariant manner. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-09

With support from the Chemical Structure and Dynamics (CSD) program in the Division of Chemistry, Professor Francis D’Souza of the University of North Texas, Denton, TX, is building wide-band capturing chromophore dimers that can form excitonic states upon photo-illumination and undergo symmetry-breaking charge separation (SB-CS). Comprehending photoexcited charge transfer in molecular assemblies is of paramount interest as it directly relates to the process of light energy to chemical energy conversion, with direct applications in photocatalysis and building optoelectronic devices. By performing systematic experimental and theoretical studies, Professor D’Souza and his students will address the unanswered questions regarding the interplay between exciton dynamics and SB-CS, and the role of the molecular structure and geometry and solvent surroundings in governing these events. Achieving a strong visible-near-infrared spectral response in these highly absorptive dimers will be transformative to building high-efficiency devices while providing an excellent training platform for graduate and undergraduate students to help strengthen teaching, mentoring, and leadership skills. Photoinduced SB-CS is a process where a symmetrical pair of identical chromophores forms a charge-separated excited state with the hole and electron on different chromophores, and molecular systems exhibiting such a property are highly sought after owing to the minimal energy loss during the charge-separated state formation. However, a critical knowledge gap exists due to the lack of synthetic molecular dimers capable of undergoing SB-CS with low-energy light excitation from the far-red and near-IR regions. This study addresses this critical issue by building far-red and near-IR capturing chromophore dimers of different geometries and orientations and performing systematic studies at broad temporal and spatial time scales, to unravel the underlying mechanistic details of SB-CS and to derive meaningful structure-property relationships. The outcomes from this broad research project are expected to have a widespread impact across many fields of science. Additionally, the project provides advanced training opportunities for graduate and undergraduate research students, while also engaging local communities through Professor D’Souza's outreach activities. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NIH Research Projects · FY 2025 · 2025-08

ABSTRACT Chromosome conformation capture techniques, particularly Hi-C, have benefitted the study of the spatial proximity, interaction, genome conformation of cells, and genome architecture leading to the development of several three-dimensional (3D) chromosome structure modeling methods. Many observations become more apparent in 3D because some relationships—for example, evolutionary constraints or cell-to-cell variability of mammalian chromosome structures—cannot be surmised by genomic sequences alone. Although members of the bioinformatics community, including the PI, have developed many algorithms for reconstructing 3D genome structures based on population Hi-C data, we lack computationally effective methods to precisely model at a high-resolution (<=5 kilobase (kb)). One difficulty is the exponentially increasing number of fragments at this resolution. My work in the last five years provides the premise for the current proposal and uniquely positions my interdisciplinary research program to carry out the proposed studies. The PI proposes to conduct leading research to overcome this challenge and address important questions that remain about how (and why) 3D genome structures across cells are organized and about the relationship between 3D structure and genetic and epigenetic mechanisms for gene expression. During the next five years, the PI’s objective is to develop computational and machine learning-based models to further highlight the hierarchical organization of, and the refined structures within, the genome. The PI proposes to explore the development of innovative models for 3D chromosome and genome reconstruction using a novel noninstance-based generalizable model based on a graph convolutional neural network to generalize across resolutions, chromosomes, restriction enzymes, and cell populations. Given the PI’s background, track record, and productivity in the genomic research field, the computational objectives defined here are not only feasible but also computationally and biologically rewarding to the bioinformatics community at large. Computationally, our methodology will resemble a robust one-size-fits- all model that can be sufficiently trained at a lower computational cost on less complex data and be used across multiple higher resolutions for 3D structural modeling. Biologically, our proposed reconstruction algorithms will aid diseases diagnosis, prevention or treatment by shedding light on the relationship between long-range interaction and gene expression in human cells and how disruptions in physical interactions between genes and the enhancers or silencers could aberrantly alter gene expression. Thus, this research demonstrates the potential impact of knowing the architecture of the genome to the understanding of biological processes and human disease. Once the proposed objectives are completed, the PI will ultimately have been well established as an independent investigator, and will have proposed leading robust, high-performing, and efficient computational algorithms that will provide new vertical advancement in the chromatin genomics research field.

NSF Awards · FY 2025 · 2025-08

An oxoanion is a negatively charged ion composed of one or more oxygen atoms along with another element. In groundwater, oxoanions are pollutants and are a threat to human health. This project will create better methods for removing such pollutants from water. Specifically, it will provide insight about pollution removal methods that work effectively even when multiple oxoanions are present in water simultaneously. The investigators will study the behavior of several oxoanions in water and show why some materials can remove them better than others. This understanding will enable the design of new materials and improved methods for water treatment. Additional benefits will come from training new scientists, and from hands-on activities at local science museums to educate the general public about water pollution. In situ infrared spectroscopy (IR) will be coupled with ab initio modeling to elucidate the behavior of oxoanions adsorbed onto porous surfaces in contact with water. This project will study the adsorption and competition of different oxoanion pollutants within porous materials, focusing on metal-organic frameworks (MOFs) as the porous materials. MOFs are selected due to their structural diversity and functional flexibility. Adsorption will be studied by infrared spectroscopy, a sensitive tool for discriminating chemical bonds and detecting a wide range of adsorbent/host interactions ranging from weak van der Waals forces to hydrogen and covalent bonds. Several basic processes involved in the adsorption of oxoanions will be examined in detail, including their diffusion, binding, competition, and molecular exchange. The key objectives are to decouple and analyze two processes involved in oxoanion adsorption: diffusion through the porous channels and binding to local sites. The project will identify the competition of different oxoanions during their diffusion through pores, and also their competition for occupying local binding sites. The findings obtained through studying MOFs will translate to other types of nanoporous materials and will provide the scientific foundations necessary for developing novel adsorption-based technologies for water purification. Education and outreach activities in this project will include summer camps, workshops and field trips to local museums to improve the public’s scientific literacy about water the cycle and pollutants. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-07

This I-Corps project focuses on the development of a software platform that enables the rapid, low-effort creation of software for building, deploying, and managing modern applications in cloud computing environments. This technology addresses the growing need to simplify application development across cloud and Internet of Things systems by providing a solution that unifies data, logic, and performance management into a single package. The current approach to building such applications requires high expertise, prolonged development cycles, and costly deployment optimization. These impediments create barriers for small businesses, educational users, and non-technical innovators. By reducing the complexity and effort required to build cloud-based applications, this technology expands access to powerful digital tools and supports innovation. The solution holds potential to increase workforce readiness, improve efficiency in industries reliant on cloud and connected systems, and expand the digital transformation. This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of a software platform that enables rapid, low-effort creation of cloud-native applications. This solution is based on the development of a serverless application abstraction that encapsulates data handling, workflow coordination, and deployment constraints in a unified, object-oriented structure. Current paradigms treat computing and data separately leading to inefficiencies and high tuning overhead. This technology recognizes data locality and enables declarative specification of quality-of-service preferences then satisfies those preferences, automatically minimizing the need for user intervention. This encapsulation extends across edge and cloud systems, enabling fault-tolerant and responsive operation even in unreliable network environments. The resulting architecture reduces software development burden and supports performance and consistency at scale, providing a technical advancement over existing cloud service abstractions. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-07

Large-scale deep learning models are widely used and are critical across many fields. However, the need to adapt and refine these models to changing conditions is becoming increasingly important. Achieving real-time adaptation poses a significant challenge due to its resource-intensive nature. This research focuses on developing methods to dynamically refine and adapt existing models, enabling real-time updates and reducing the reliance on time-consuming retraining. Such advancements are critical for scenarios where static pre-trained models often underperform, including cases with incomplete data, dynamic environments, or evolving research objectives. These methods support a wide range of applications requiring real-time adaptability in large-scale models, such as climate modeling, real-time traffic management, digital twins of fusion energy in plasma physics, and virtual infrastructure twins of supercomputer networks. By addressing this challenge, the project not only promotes innovation but also boosts scientific research and practical applications. This project contributes to the U.S. national goal of broadening participation in science and engineering, developing the research workforce for advanced cyberinfrastructure, promoting the innovation economy, and maintaining a leading position in international technology competitions. The developed software as a foundational tool is shared with researchers and engineers in academia, national laboratories, and industry for broader societal impacts. Also, this project promotes K-12, undergraduate, female, and underrepresented minority populations in the science, technology, engineering, and mathematics (STEM) fields and strengthens collaboration between academia and national laboratories/industries. The primary objective of this project is to pioneer the effort of developing a time-sensitive large model training platform for dynamic data analytics in practice by taking advantage of the world-class accelerator computing infrastructure to fine-tune trillion-parameter large models cost-efficiently. Towards this end, this project has three research thrusts and one educational thrust. First, this research automatically generates a parallelization plan at an affordable cost to minimize training iteration latency. Second, this project progressively grows models from pre-trained small models during fine-tuning to reduce the number of training iterations. Third, this project validates the practicality of the developed platform using realistic applications in weather forecasting, fusion energy experiment control, resilient streaming event prediction in virtual infrastructure twins, and scooter-sharing demand prediction. Fourth, this project designs integrated research and education activities, including broadening the adoption of the time-sensitive training platform, undergraduate student research advising, curriculum development, and outreach for professional development and K-12 education. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-06

University of North Texas (UNT) will establish a three-year Research Experiences for Undergraduates (REU) site focused on making Generative Artificial Intelligence (AI) responsible. This initiative aims to address critical issues related to impartiality, ethics, privacy, and usability in Generative AI systems. Each summer, undergraduates will participate in a ten-week residential research program designed to ensure that advanced Generative AI technologies are developed and deployed responsibly. This REU site aligns with national priorities for ethical AI innovation and workforce development, preparing students to address real-world challenges such as model impartiality, data privacy, and social impact. UNT will leverage its resources, faculty expertise, and commitment to its mission to support this program. Recruitment efforts will focus on identifying and nurturing talent to build a skilled AI workforce that advances the nation’s technological progress and societal well-being. Over three years, the REU site will provide undergraduates with interdisciplinary research opportunities involving adversarial testing, prompt engineering, hardware-based privacy-preserving techniques, and error detection and mitigation in multimodal AI systems. Participants will be guided by seven faculty mentors and supported by graduate students, utilizing high-end Graphics Processing Unit (GPU) servers and resources at the Texas Advanced Computing Center (TACC) to conduct experiments and develop innovative solutions for responsible Generative AI. The program includes daily lessons, research activities, and professional workshops that expose students to cutting-edge methods for designing and evaluating responsible Generative AI models. By the conclusion of each ten-week session, participants will have produced publication-ready papers and presented their findings in poster sessions. Through a combination of specialized coursework, hands-on research, and mentorship, this REU site aims to train a new generation of AI researchers capable of addressing complex issues related to impartiality, ethics, privacy, and usability in Generative AI. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-06

This study addresses a longstanding issue in our understanding of hazard adjustment decisions: what pushes individuals thinking about hazard adjustment to adopt these adjustments? To address this question, the research team employs a longitudinal survey design to understand the mechanisms that lead to adjustment behaviors over time. By addressing this issue, emergency managers and other key stakeholders can create programs that reduce risk by targeting barriers to adopting hazard adjustment behaviors. The team leverages these insights by co-developing a toolkit to support local and state efforts to improve flood adjustment program design, participant experiences, and community outcomes. The project also provides an opportunity for experiential learning and graduate student training. The findings are transferable to other locations affected by natural and induced technological hazards. This project builds on Paton’s Social-Cognitive Preparation Model to examine how people develop expectations, intentions, and behaviors related to flood hazard adjustments over a 3-year period. The research team surveys households living in coastal areas within 100-year flood zones. The survey introduces an experimental intervention to measure how flood hazard adjustment behaviors change after participants receive a brochure on flood hazard protections. The study also uses advanced statistical methods such as Structure Equation Modeling and Latent Growth Curve Mediation Modeling to analyze how different factors influence people’s flood hazard adjustment decisions, as well as how behavioral intentions and actual behaviors change over six time points across 3 years. This study extends beyond typical one-time or short-term surveys by tracking how intentions to adopt flood adjustments translate into actual behaviors over time. The research model is transferable to understand drivers that affect protection behaviors across different types of hazards, cultures, and places. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-06

With the support of the Chemical Catalysis program in the Division of Chemistry, Professor Shengqian Ma of the University of North Texas is studying a new type of metal-free porous solid catalyst to accomplish challenging chemical transformations historically limited to rare and expensive metals. This new catalyst class may overcome deficiencies of both existing catalysts, providing a robust framework to facilitate precise tunability. By fine-tuning electron-poor (Lewis acid) and electron-rich (Lewis base) moieties, as well as the microscopic environment surrounding them, the so-called frustrated Lewis pairs embedded in an easily tailorable framework will allow non-metals to achieve metal-like reactivity. Moreover, as a metal-free material, this new class of catalyst will be well-positioned for use in pharmaceutical and food industries, inherently circumventing rigorous purification otherwise necessary to remove potentially toxic metals. This project will support research efforts of a broad team, with researchers spanning high-school, undergraduate, and graduate levels. Students will be trained to utilize an exhaustive set of cutting-edge instrumentation for synthesis, characterization, analysis, and modelling, setting the foundation for impactful careers in critical STEM disciplines. With the support of the Chemical Catalysis program in the Division of Chemistry, Professor Shengqian Ma of the University of North Texas is studying a new class of metal-free heterogeneous catalysts for asymmetric hydrogenation reactivity based on frustrated Lewis pairs (FLPs) confined within the nanospace of covalent organic frameworks (COFs). The project will build upon Professor Ma’s recent advances in the field, systematically quantifying primary, secondary, and tertiary sphere chiral induction effects towards achieving high enantioselectivity for traditionally promiscuous substrates such as the asymmetric hydrogenation of tetrasubstituted enamides – a key challenge in the synthesis of emerging pharmaceuticals. Strategies will be developed to rationally tailor COF-supported FLP microenvironments followed by systematic evaluation of the resulting enantioselective and chemoselective hydrogenation of an established set of substrates. Additionally, in situ spectroscopy supported by computational modelling will elucidate outstanding questions in FLP-based hydrogenation catalysis, leveraging methods pioneered in Professor Ma’s group. Insights garnered from this project will be directly applicable to other framework-based systems for asymmetric catalysis and beyond. This project will involve researchers spanning high-school, undergraduate, and graduate levels, setting the foundation for their impactful careers in critical STEM disciplines. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Awards · FY 2025 · 2025-05

This Research Experiences for Undergraduates (REU) site award to the University of North Texas (UNT), located in Denton, TX, supports the training of 10 students for 10 weeks during the summers of 2025-2027. In this program, funded by the Division of Chemistry, early-stage college students will engage in highly collaborative and interdisciplinary summer research projects involving two or more faculty mentors with different expertise. Students will be recruited through a network of faculty partners at primarily undergraduate Hispanic-serving institutions (HSIs) in the state of Texas. A special focus of the program is the training of partnering HSI faculty in collaborative mentorship skills. These HSI partners will continue to mentor students at their home institutions in the academic year following the summer REU program, while remaining engaged with UNT collaborators and benefitting from access to research facilities at UNT that they lack at their home institutions. Involving students in a full year of collaborative research is expected to make them more excited about STEM careers and better prepared to tackle research at the cutting edge of important new technologies that benefit the nation. Students in this REU program will pursue highly interdisciplinary summer research projects that draw on the expertise of a team of mentors, with topics ranging from joint computational/experimental studies of nitrogen and carbon dioxide activation by metal catalysts, to applications of machine learning in solvation and crystal polymorphism, to corrosion-resistant nanocomposite coatings, among others. Partnering HSI faculty will be encouraged to build on the collaborative research skills that they and their students gain in the summer program as they continue to mentor students in research over the following academic year. The collaborative research training will be coupled with a professional development program focused on scientific communication, scientific ethics, and STEM career paths. The program is expected to contribute to a more ethical, collaborative, and creative STEM workforce, with enduring benefits for the economic competitiveness of the state of Texas and the nation. 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.