Resolution is a key aspect affecting the performance of object imaging and detection tasks, including imaging the geophysical and geological features of interest beneath and on the Earth’s surface. Traditionally, improving the resolution of these images is computationally expensive and involves complex and time-consuming processes. However, artificial intelligence (AI) offers a potential solution to help accelerate this process and reduce computational costs. In this study, instead of relying on traditional, computationally intense methods, our proposed approach uses an AI-driven method to enhance the resolution of the seismic velocity imaging model and satellite imagery data by learning and correcting a pattern called the Hessian point-source image-blurring function. Therefore, it allows us to map from a blurred image/model (input) to a deblurred image/model (output). Our research focuses on developing an AI-based model that approximates complex calculations needed to refine the earth science imagery. For the subsurface imaging case with seismic imaging data, in our numerical experiments, we use realistic simulations of blurred seismic velocity models and train our model with examples at various depths and locations in the subsurface. Meanwhile, in our surface imaging case with remote sensing data, to fulfill the high desire for high spatial resolution earth observation (EO) images for many remote sensing applications, our potential solution offers to provide a finer depiction of spatial boundaries by training several bands images from the same region and at the same time interval of acquisition, with the multiple resolutions band images. Subsequently, we validate our model’s performance using hidden point sources not included in the training dataset. Through these efforts, we aim to demonstrate the efficacy of AI in significantly improving the resolution of earth imaging for enhanced subsurface imaging and surface geological feature detection.

Microscopic examination of slides prepared from tissue samples is the primary tool for detecting and classifying cancerous lesions, a process that is time-consuming and requires the expertise of experienced pathologists. Recent advances in deep learning methods hold significant potential to enhance medical diagnostics and treatment planning by improving accuracy, reproducibility, and speed, thereby reducing clinicians’ workloads and turnaround times. However, the necessity for vast amounts of labeled data to train these models remains a major obstacle to the development of effective clinical decision support systems. In this paper, we propose the integration of topological deep learning methods to enhance the accuracy and robustness of existing histopathological image analysis models. Topological data analysis (TDA) offers a unique approach by extracting essential information through the evaluation of topological patterns across different color channels. While deep learning methods capture local information from images, TDA features provide complementary global features. Our experiments on publicly available histopathological datasets demonstrate that the inclusion of topological features significantly improves the differentiation of tumor types in ovarian and breast cancers.

ANN CROWDS: INVESTIGATING DIVERSITY IN NEURAL NETWORK ARCHITECTURES This study examines how varying neural network architectures impact prediction accuracy within an ANN Crowd—a collection of 189 unique architectures designed to enhance predictions from small training sets. Each architecture is replicated 100 times with different initial weights, creating a crowd of 18,900 voices. Using statistical analysis, we explore whether architectural variations (categorized as ‘Good’ or ‘Bad’ based on key metrics) and the number of nodes in each group significantly affect accuracy. The findings provide insights into building more effective neural network ensembles for predictive modeling.

The human brain consists of complex neuronal networks that interact electrically and neurochemically on both spatial and temporal scales. Discovering the brain circuitry’s functions such as perception, memory, and behavior remains one of today’s challenges. The mechanisms in which degradation of such functions is driven by neurodegenerative disorders are not well understood and such disorders continue to pose a significant threat to the aging population. Therefore, there is a critical demand for novel neurotechnology capable of accurately mapping the spatial and temporal distribution of neurochemicals. Existing technologies, such as carbon-fiber electrodes and neurochemical fMRI, suffer from low spatial and/or temporal resolution. In this work, we have developed a multisite neural probe that is capable of monitoring the activity of neural spikes and dopamine.

Photodynamic therapy (PDT) uses light activatable nanosized drug delivery systems known as photonanomedicines. In this study we developed a solid lipid nanoparticle formulation (LNP BPD-PC) of lipidated benzoporphyrin derivative (BPD-PC) and contrasted it with an already established liposomal formulation of BPD-PC (Lipo BPD-PC). Our investigation demonstrates that Lipo BPD-PC is able to generate 17% more singlet oxygen than LNP BPD-PC, while LNP BPD-PC generates 76% more hydroxyl radical and/or peroxynitrite anions. Importantly, while 100% of BPD-PC leaches out of the Lipo BPD-PC formulation, only 28% of BPD-PC leaches out of the LNP BPD-PC formulation within 7 days of incubation in serum at 37 °C. Despite a difference in the generation of reactive oxygen species, LNP BPD-PC and Lipo BPD-PC show similar phototoxicity in CT1BA5 cells (murine pancreatic cancer cells). Additionally, no significant difference was observed in the cellular uptake of BPD-PC over 24 h when CT1BA5 cells were treated with LNP BPD-PC and Lipo BPD-PC. Further, LNP BPD-PC was more efficient at inducing ICD with IC25 and IC50 PDT doses compared to Lipo BPD-PC. Tumor delivery of BPD-PC was 2.41-fold higher with Lipo BPD-PC compared to LNP BPD-PC in mice bearing CT1BA5 tumors. In summary, these results suggest that LNP BPD-PC is a better nanoplatform to carry BPD-lipid conjugates than Lipo BPD-PC and LNP BPD-PC can be a potential nanoplatform for future drug delivery systems and photodynamic immune priming.

Harnessing wind energy at deep-water locations can be effectively achieved by deploying wind turbines on floating platforms. However, these floating offshore wind turbines (FOWTs) experience 6-degree-of-freedom (6-DOF) motion, which influences their performance, fatigue loads, and the behavior of wakes, leading to complex interactions between FOWT arrays. For this study, we tested two downscaled models of DTU 10 MW wind turbines, each with a 40 cm rotor diameter (D), installed in tandem at the UT Dallas Boundary Layer and Subsonic Wind Tunnel (BLAST), using multi-hole pressure probes, 6-DOF force sensors, and electric generators. The upstream FOWT is installed on a mechatronic emulator reproducing typical turbine motion during offshore operations, specifically imposing sway motion with non-dimensional amplitudes (A/D) ranging from 0 to 0.16 and Strouhal numbers (St) from 0 to 0.35. FOWT operational tip-speed ratios (TSRs) were set to sub-optimal (4.8), optimal (5.2), and super-optimal (7.5) values, with the downstream turbine positioned at distances of either 5D or 10D, operating at the optimal TSR. Both the turbines were tilted backward to a negative 6o pitch angle to reproduce a realistic condition. The results indicate that elevated St and A/D improve wake recovery, particularly at higher tip-speed ratios (TSRs), which boosts power generation and thrust on the downstream turbine, causing lateral wake expansion and vertical shrinkage, leading to an elliptical mean wake shape. Additionally, higher St and A/D increase Reynolds stresses and the turbulent kinetic energy within the wake.

We introduce a novel Bayesian clustering framework for analyzing simple polygon chains with n vertices, specifically n-gons, using their relative interior angles and side lengths. The primary objectives of this framework are to cluster the n-gons into K distinct groups, register the shapes for alignment, quantify the statistical properties (mean and variance) of the n-gons in each cluster, and autonomously determine the number of clusters, K. The methodology employs a Double Truncated Dirichlet (DTDir) Model, which accounts for the geometric constraints of the data normalized on a simplex space. A weighted DTDir model is further incorporated to address cases where angles and side lengths contribute differently to the clustering, updating both characteristics jointly. Two shape registration parameters—the starting vertex indicator and orientation indicator—ensure consistent alignment relative to a predetermined reference shape. To perform clustering, we use a Mixture of Finite Mixtures (MFM) model, which places a truncated Poisson prior on K and estimates the optimal number of mixture components. The model is validated through simulations with geometric shapes generated from a DTDir distribution and applied to a biological dataset of erythrocytes, distinguishing between normal circular blood cells and elongated sickle cells. We provide their clustering performances and construct each cluster’s mean shape based on the corresponding DTDir parameters, with variability quantified through the variances.

A one-way numerical coupling of the ocean and atmosphere was developed to study the effects of offshore wind turbines in the surrounding oceanic regions. Our in-house UTD-WF LES code was used to model the atmospheric boundary layer and the wind turbines, whereas FVCOM was used to model the oceanic domain. An ideal wind farm, consisting of an array of 2 x 4 turbines uniformly spaced along the direction of the flow, was modeled. Two sets of simulations, with a uniform-inlet and a turbulent-inlet at a Reynolds number of Re_D=(U_ref D)⁄υ= 9.77 x 10^7, were performed. The wind flow from these simulations were used as shear forcing for the ocean model. From the results, it is evident that the induction zone from the turbines is directly translated into the ocean surface. This high-pressure zone introduces a spanwise shear component that initiates a streamwise-normal vortex that spans the depth of the ocean domain. When turbulence is presented at the inlet, these large vortices no longer retain their structures because of the increased mixing of the flow. HPC at UTD and TACC is acknowledged for providing computational time.

Bladder cancer (BlCa) is the fourth most common cancer in men and has a 71% 5-year survival rate if confined to the bladder; however, this survival rate is reduced to 39% or 8% for regional or distant metastasis. Thus, it is crucial to understand the underlying mechanisms that lead to the progression of BlCa metastasis. The tumor microenvironment (TME) is replete with inflammatory factors secreted by infiltrating immune cells recruited to the TME to kill the cancer cells. Acute inflammation is anti-tumorigenic, yet, if left unresolved, chronic inflammation is a known hallmark for cancer initiation and progression. Our lab has discovered that when chronically exposed, cancer cells can develop insensitivity to the anti-tumorigenic effects of the inflammatory cytokine, interleukin-1 (IL-1). Pro-tumorigenic effects of IL-1 include tumor angiogenesis and metastasis and drug resistance. Consequently, IL-1-targeted therapies, such as the IL-1 receptor antagonist (IL-1Ra), anakinra, would not be effective against IL-1-insensitive cancers. We discovered that the BlCa cell line, 5637, secretes IL-1 and may participate in chronic autocrine IL-1 signaling. Interesting, we found that, despite potential autocrine signaling, 5637 can still respond to added concentrations of exogenous IL-1. Therefore, we generated unique chronic IL-1-exposed 5637 sublines to determine if 5637 cells could develop IL-1 insensitivity. Using this model, we began to characterize the chronic IL-1 sublines’ sensitivity to IL-1 by acutely treating them with exogenous IL-1 or IL-1Ra. We subsequently assessed protein accumulation of IL-1-induced genes, LCN2 and CXCL1/CXCL2, along with migration potential, using β-catenin to monitor cell-cell junctions. Interestingly, our data showed two consistent phenotypes in response to acute IL-1 and IL-1Ra—first, their LCN2 and CXCL1/CXCL2 levels were consistently induced in response to IL-1 and attenuated when IL-1Ra was introduced. Secondly, their cell-cell connections were lost in response to IL-1 but recovered when IL-1Ra was present. This indicates that our 5637 chronic sublines maintained their sensitivity to exogenous IL-1 and IL-1Ra; a result that deviates from what we have seen in other cancer cell lines. As such, future work would entail comparing cancer cell lines that are sensitive to IL-1 with cancer cell lines insensitive to IL-1 and identifying genetic and epigenetic biomarkers that would predict cancer patients’ sensitivity and response to IL-1-targeted therapies, like anakinra.

Single-event effects (SEEs) are radiation-induced failures in electronic components that limit the performance, reliability, and availability of systems used in space exploration. SEE testing is critical for improving system reliability but often requires access to expensive, complex, and limited facilities, such as those equipped with particle accelerators or nuclear reactors. As an alternative, two-photon absorption (TPA) using sub-bandgap wavelength pulsed lasers with a quasi-Bessel beam profile has been used to induce SEEs by simulating the trajectory of heavy-ion strikes across semiconductor devices. Previous experiments by the authors have shown strong correlation between TPA-induced SEEs and heavy-ion SEEs based on the collected charge for the same linear energy transfer (LET). However, this correlations remain remains constrained by the continued need for particle accelerators for validation. To address this limitation, Montecarlo simulations, such as stopping range or ions in materials (SRIM) estimates the charge generated by heavy ion strikes in semiconductors. However, internal charge losses which critical for correlating TPA and heavy-ion SEEs, are not described. In this work, technology computer-aided design (TCAD) is employed to model a custom-fabricated large-area PIN silicon photodiode, estimating the difference between generated and collected charge. This approach offers insight into charge loss mechanisms and enables rapid correlation without the need for particle accelerators. The generated charge is calculated with SRIM and injected into the device model’s sensitive volume, simulating the characteristic TPA quasi-Bessel beam profile and heavy-ion trajectory. A comparison is then made between the simulated and experimentally measured collected charge, from both TPA-induced SEEs and heavy-ion SEEs offering insights into the underlying charge loss mechanisms.

This study presents a comparative analysis of D-Mode GaN/AlGaN/GaN MOS-HEMTs with three different gate dielectrics: HfO2, Al2O3, and SiNx, aiming to optimize performance for power and high-frequency applications. Devices were fabricated using standard lithography techniques, followed by Atomic Layer Deposition (ALD) for HfO2 and Al2O3, and Plasma-Enhanced Chemical Vapor Deposition (PECVD) for SiNx. Transmission Line Method (TLM) measurements showed similar contact characteristics across the dielectrics due to a consistent ohmic contact process. Electrical characterizations, including output and transfer characteristics, were performed to extract key parameters such as on-resistance (Ron), maximum drain current (ID,max), threshold voltage (Vth), Ion/Ioff ratio, subthreshold swing (SS), and transconductance. While SiNx demonstrated the highest ID,max, it also exhibited the highest SS and lowest transconductance, indicating reduced gate control and less efficient switching. In contrast, HfO2 provided lower SS and higher transconductance, suggesting better channel control and switching behavior. These findings underscore the trade-offs in dielectric selection, where SiNx enhances current capacity but sacrifices switching efficiency, while HfO2 enables superior switching characteristics.

Image resolution and field-of-view are typically inversely proportional to one another in optical microscopy systems due to the sampling frequency limitations posed by the system’s pixel size, which is determined by the numerical aperture and magnification of the imaging system. This tradeoff limits the information carrying capacity, defined from the space-bandwidth product, of a given system. In cardiac studies, it is crucial to capture the highest amount of information, both in terms of field-of-view and resolution, to maximize the amount of data gained during observation. In the case where the pixel size is larger than the contributions from the point spread function, the tradeoff between field-of-view and spatial resolution can be subverted by increasing the sampling frequency, accomplished by spatially translating the sample by distances less than the system’s pixel size. We present a spatial shifting method and reconstruction algorithm that bypasses this tradeoff through spatially shifting the sample by sub-pixel increments before computationally registering via phase correlation, upscaling, and reconstructing using the shift-and-add technique, the resulting captured images. To demonstrate this method, we imaged a USAF 1951 target, demonstrating a resolution improvement of 1.42x, from 7.8 μm to 5.5 μm. Since the proposed method preserves field-of-view, this resolution improvement increased the space-bandwidth product by a factor of ~2. The findings of the presented study offer a substantial increase in spatial resolution to previously spatially-undersampled microscope systems, thereby increasing their information carrying capacity.

Kaposi’s Sarcoma Herpes Virus (KSHV) is the etiological agent of Kaposi’s Sarcoma (KS) and a leading cause of cancer in AIDS patients. With limited treatment for KS, it is essential to understand the mechanics behind KSHV infection. KSHV undergoes latent and lytic viral phases in the host cell. Only a few viral genes are expressed during latency, however, during lytic replication, KSHV encodes for ~90 genes, including a viral G protein-coupled receptor (vGPCR), a chemokine-like receptor with homology to host receptors. vGPCR signaling has been shown to alter host cell survival, angiogenesis, and metabolism. Previous metabolomic analysis showed that KSHV latently-infected endothelial cells modulate central carbon pathways (glycolysis, fatty acid synthesis, and amino-acid metabolism). Drug inhibition of these metabolic pathways resulted in reduced survival of latently infected cells and a significant reduction in virion production during the lytic phase. However, no study has measured the global alterations in the host cell metabolome during lytic KSHV to date. We hypothesize that global host cell metabolism is modulated to support maximal lytic infection. To test this hypothesis, endothelial cells overexpressing vGPCR or undergoing KSHV lytic infection will be analyzed via mass-spectrometry for metabolite levels. We expect to see significant metabolite changes in central carbon metabolic pathways and potentially reveal novel pathways modulated by KSHV lytic infection.

More than 70% of breast cancer (BCa) patients harbor tumors that are dependent on estrogen receptor alpha (ERα) for growth. Thus, ERα is a BCa therapeutic target. While initially these therapies are effective, patients can become resistant to these treatments. In the tumor microenvironment (TME), interleukin-1 (IL-1) is an inflammatory cytokine that is known to be elevated in BCa patient serum. Although IL-1 is initially secreted in the TME to combat cancer, chronic IL-1 exposure can lead to chronic inflammation, which is known to promote cancer initiation and progression. We previously reported that acute exposure (i.e. days) to IL-1 leads to downregulation of ERa, potentially selecting for a subpopulation of treatment-resistant cancer cells that survive without hormone receptors. An alternative therapeutic target is glucocorticoid receptor (GR), another hormone receptor. However, GR regulation by IL-1 in ERa positive breast cancer has not been fully explored. We previously reported that acute IL-1 represses the hormone receptors (HR), ERa and progesterone receptor (PR) in breast cancer cells and androgen receptor (AR) in prostate cancer cells, potentially selecting for a subpopulation of treatment-resistant cancer cells that can survive without HR. To mimic chronic exposure to inflammation in BCa, we generated novel sublines by chronically exposing MCF7 to two members of the IL-1 family of cytokines, IL-1a or IL-1b, to generate MCas and MCbs sublines, respectively. Using our novel cell line models, we have discovered that cells chronically exposed to IL-1 restore their ERa, PR, AR, and GR. IL-1 signals through the NFkB pathway, and upon IL-1 binding to its receptor, a series of signaling cascades ultimately free up and activate RELA (canonical NFkB signaling subunit) and RELB (noncanonical pathway). Upon activation, RELA and RELB translocate to the nucleus act as transcription factors to turn on genes for inflammation. Previously we have reported that in PCa cells, RELA silencing was sufficient to mediate AR repression and attenuates IL-1 repression of AR. Preliminary results show that RELA and RELB silencing attenuate IL-1 repression of ER and GR. Future work includes ChIP and ChIP-qPCR to determine if RELA and RELB bind to promoter regions on the ER and/or GR loci to downregulate these HR. Our studies will help us to 1) determine if the NFkB regulation is conserved for MCF7 and sublines that have been chronically exposed to IL-1 in the tumor microenvironment and 2) better understand the role of IL-1 regulation of ERa and GR in the progression of BCa tumors, and to further understand how resistance is due to chronic exposure to inflammation.

Cisplatin is a potent alkylating agent that can induce chemotherapy-induced peripheral neuropathy (CIPN) and chronic pain. This severely limits its usefulness as a cancer treatment and negatively impacts patient outcomes. Evidence suggests that CIPN arises as a consequence of molecular alterations in dorsal root ganglion (DRG) neurons and their neighboring satellite glial cells that lead to functional changes generating hyperexcitability. CIPN is also associated with a more distal loss of intraepidermal nerve fibers in the skin. Meteorin is a secreted protein that plays a fundamental role in the development of the nervous system. Previous studies have shown that systemic treatment with recombinant mouse Meteorin (rmMeteorin) produces robust, long-lasting antinociception in rodent models of peripheral neuropathic pain and paclitaxel-induced peripheral neuropathy. To investigate a broader role for Meteorin therapy in CIPN, in the current study female ICR mice were treated with 2 cycles of cisplatin (2mg/kg, 5 x i.p.) through Days 1-15. Hind paw withdrawal thresholds to mechanical stimulation were assessed using Von Frey filaments. Once hypersensitivity was established, rmMeteorin (1.8mg/kg, 5 x s.c.) or vehicle was administered from Day 16 every other day, and withdrawal thresholds routinely assessed until Day 81. Lumbar DRGs and hind paw skin were collected from 4 mice from each treatment group on Day 34 for immunohistochemical processing. Cisplatin-induced mechanical hypersensitivity was significantly reversed after a 2nd injection of rmMeteorin and persisted throughout, and even beyond the dosing duration. Moreover, cisplatin-mediated changes in DRG expression of the gap junction protein Connexin43 and the satellite glial cell marker glutamine synthase were both restored by rmMeteorin administration. Intraepidermal nerve fibers in the skin were also protected from cisplatin-induced loss by rmMeteorin. The resolution of preclinical behavioral and cellular correlates of CIPN in cisplatin mice after rmMeteorin treatment, supports similar findings obtained in mice with paclitaxel-induced neuropathic pain.

Lead glaze on earthenware undergoes blackening in anaerobic environments such as those found in cesspits and canals. The blackening effect is attributed to the formation of black lead(II) sulfide within the glaze. However, the degradation process is not well understood; degraded glazes exhibit not only black colors but also red and orange colors. In this study, five lead-glazed ceramic pieces excavated from the canals of Amsterdam were analyzed using optical microscopy, X-ray photoelectron spectroscopy (XPS), and Raman spectroscopy to investigate the chemical and physical properties of the aged glazes. Each ceramic piece exhibited different levels of degradation, with Object 1 appearing fully black and Object 5 appearing off-white. Optical images revealed that the most visibly damaged objects had rougher surfaces and crystals growing in cracks of the glaze. Raman spectra and XPS data indicate that the objects share many of the same components, including lead oxides, tin(II) oxide, and silicon dioxide. Surprisingly, other lead compounds that are not known to be originating ingredient of glaze are also present, including lead(II) carbonate, lead(II) sulfide, and lead(II) sulfate. To determine if a correlation between the chemical compositions and the perceived colors of the objects existed, principal component analysis (PCA) was conducted, but there does not appear to be any strong correlations. This is likely due to the limited number of samples investigated.

This study presents a refined framework for Engineering Identity in undergraduate education, crucial for shaping the next generation of engineers. By leveraging design research and theory, this framework explores how students perceive and validate their identity through interests, competencies, and recognition, influenced by personal and educational experiences. Understanding and fostering engineering identity equips students with critical skills and mindsets for successful careers. This work emphasizes the need for educators to actively cultivate engineering identity, ensuring future engineers are prepared to innovate and excel across diverse industries.

Breast cancer invasion is driven by both mechanical interactions with the extracellular matrix (ECM) and metabolic reprogramming. This study examines the role of YAP/TAZ signaling in coordinating mechanical remodeling of the ECM with metabolic alterations in cancer cells invading in complex 3D micro-environments. MCF-10A spheroids with inducible YAP/TAZ mutants were embedded in 3D collagen matrices of varying densities. Upon YAP/TAZ activation, spheroids showed increased proliferation and more aggressive invasion, characterized by significant ECM remodeling and matrix stiffening, particularly in lower collagen densities. YAP activation increased glycolysis, while TAZ enhanced both glycolysis and oxidative phosphorylation, suggesting distinct metabolic roles. Fluorescence Lifetime Imaging (FLIM) revealed metabolic heterogeneity during invasion, with the core exhibiting a glycolytic signature, while the multicellular invading strands and peripheral single cells showed oxidative and glycolytic phenotypes, respectively. Pharmacological inhibition of key metabolic pathways altered invasion strategies, with glycolysis inhibition favoring collective invasion and oxidative phosphorylation inhibition promoting single-cell invasion. These results demonstrate that YAP/TAZ signaling enhances breast cancer invasion by regulating ECM mechanics and metabolic flexibility. This study highlights YAP/TAZ’s crucial role in mechano-metabolic regulation, providing insights into how breast cancer cells adapt their invasive behavior in response to both mechanical and metabolic cues in the tumor microenvironment.

Intracortical microelectrode arrays (MEAs) are implanted devices that can record or stimulate neural activity. MEAs can record neural signals in the form of extracellular action potentials, characterized by single units, and local field potentials (LFPs) and are used in both basic science and clinical work. For an MEA design to be practical, it must be able to reliably be implanted into the brain. MEAs experience compressive force during implantation. If the penetration force exceeds the critical buckling force, the MEA will buckle. The compressive force is also affected by implantation speed. Higher speeds create more force, but lower speeds cause more tissue strain and vascular damage. We have demonstrated the ability to implant a novel liquid crystal elastomer (LCE) test structure into rat motor cortex and identified and optimal insertion speed.

The subsurface earth physical property Q (seismic wave attenuation) can provide helpful information for characterizing hydrocarbon/ water/ CO2/ geothermal resource reservoirs in exploration seismology, as well as investigating the presence of partial melting and water within the deep crust and mantle in earthquake seismology. Therefore, developing accurate and efficient numerical methods for estimating subsurface Q variations from seismic wave propagation is particularly important. Full waveform inversion, which uses all useful information in seismic wavefield data to estimate high-resolution subsurface physical properties, has great potential to be used to more accurately estimate earth Q models. In this study, on the basis of the new nearly constant Q viscoelastic wave equation, a novel viscoelastic waveform inversion method is developed to estimate subsurface Q structures. In comparison to previous studies, the advantages of the proposed method include: (1) the attenuation factor Q is explicitly incorporated in the wave equation, which enables us to easily calculate the sensitivity kernels for Q without involving the Fourier transform; (2) the relaxation time (weighting function) of the viscoelastic wave equation is not perturbed within selected frequency ranges, which makes it easier to derive the adjoint wave equation and sensitivity kernels than the general standard linear solid (GSLS) method; (3) the velocity and Q sensitivity kernels can be numerically calculated in the time domain, which has higher computing efficiency compared with frequency domain approaches. Following the adjoint-state method, the Q sensitivity kernels can be constructed efficiently for the amplitude-ratio misfit function. Moreover, the L-BFGS optimization method is adopted to reduce inter-parameter trade-offs between velocity and Q kernels. Several numerical experiments, such as cross-well and distributed acoustic sensing data acquisition geometries, are used to demonstrate the feasibility and accuracy of the proposed viscoelastic waveform method which has great potential to detect partial melting and water transportation within the Earth’s mantle.

Psoriasis is a chronic inflammatory skin disease characterised by dysregulated keratinocyte proliferation leading to hyperkeratosis, redness with plaque formation, and itch. Earlier studies have revealed that psoriasis exhibits a strong up-regulation of defensin genes DEFB4A and DEFB4B. Defensins are cysteine-rich antimicrobial peptides that are produced in response to injury and invading pathogens. Despite being elevated in psoriasis, their exact involvement in psoriatic inflammation remains unknown. We recently discovered that they bind to a family of G protein-coupled receptors (GPCRs) called MRGPRs. We proposed that defensins activate Mrgpr2a/b on neutrophils and Mrgpra3 on nociceptive neurons to modulate psoriatic inflammation and itch. To investigate this hypothesis, we utilize a psoriasis mouse model by topical application of imiquimod (IMQ) on wild-type (WT) and Mrgpra2 knockout (KO) mice for 7 days. Our preliminary data confirmed the overexpression of the multiple Defb genes in the mouse model, consistent with human psoriatic skin scRNAseq analysis. Furthermore, we validated this observation by HCR RNA-FISH, which showed a significant increase in the levels of DefB14 in IMQ-treated skin compared to the control group. We also observed a significant decrease in neutrophil infiltration in the Mrgpra2 KO mice when compared to WT by flow cytometry, suggesting that Mrgpr2a/b may play a role in neutrophil activation. Further, we found that both Def cluster KO mice and Mrgpra2 KO mice exhibited less psoriatic itch. To understand this neuro-immune crosstalk, we injected human β-defensin 2 (hBD-2) into WT mice and found an increase in scratch bouts. To verify whether hBD-2 can activate neurons, we measured intracellular Ca2+ levels of cultured DRG neurons and found that small-diameter neurons responded to defensin. We also demonstrated that ablation of Mrgpra3 neurons significantly reduces defensin-mediated scratch behaviour. In summary, these findings will shed light on defensin pathways that are activated during psoriasis and enhance our understanding of the disease pathophysiology.

DNA polymerases play a crucial role in managing genetic information by facilitating the synthesis of DNA strands with exceptional precision, thus upholding the integrity and stability of the human genetic blueprint. Various mutations in these polymerases affect the function of DNA polymerases and cause various diseases. DNA polymerase kappa (pol κ), a human polymerase of the Y family, has been reported to be susceptible to multiple cancer related mutations. Notably, the Y432S mutation in Pol κ has been linked to multiple cancers, such as melanoma, and is also seen to decrease polymerization activity and thermal stability. Employing computational simulations, the structural and functional implications of the Y432S mutation using classical molecular dynamics (MD) and coupled quantum mechanics/molecular mechanics (QM/MM) methods were investigated. The findings suggest that the Y432S mutation induces structural changes in the domains responsible for nucleotide addition and ternary complex stabilization, while still preserving catalytic capability. Analysis of the energy profiles associated with the reaction mechanism of both wild type (WT) and Y432S Pol κ reveals that both enzymes remain catalytically proficient, albeit with the cancer-associated mutation being energetically less favorable, with an elevated catalytic barrier. Additionally, the interactions with the third magnesium ion and the environmental influences on the non-covalent interactions on the crucial residues, contribute to the kinetic and thermodynamic disparities between the WT and mutant enzymes during the catalytic process. The energetic and electronic assessments also highlight the significance of active site residues in promoting the catalytic reaction with protonated incoming nucleotide (dNTP3–) over deprotonated incoming nucleotide (dNTP4–).

Continuous monitoring of physiological parameters from deep-seated organs, including bladder monitoring, is essential for preventive healthcare and early disease diagnosis. However, these monitoring technologies face constraints related to power consumption, form factor, cost, and technical issues such as depth of signal penetration and difficulties in resolving signals from specific tissues, which have hindered their widespread adoption, especially for deep tissues like the bladder. Recent advancements in wearable ultrasound sensors have demonstrated the potential for low-power, unobtrusive, and high-quality sensors, but they still face significant technical challenges in achieving power efficiency, reducing computational complexity, maintaining wireless connectivity, and capturing high-quality imaging at high channel counts with low-power consumption and small form factor. In this work, we present an integrated autonomous signal chain solution platform that consumes 0.898 W of power, enabling a prolonged battery life. The platform features a doubled transducer aperture, which provides a broader field-of-view and improved image quality without requiring additional piezoelectric elements, transmitters, or AFE resources using a deep learning processing unit (DPU) accelerator design inside the FPGA. This approach results in lower power consumption, reduced cost, and a smaller footprint. This device leverages a lightweight UNET (L-UNET) convolutional neural network (CNN) model optimized for efficient computation in addition to 1 core of DPUCZDX8G with an 8×14×14 PE array in the B3136 configuration. This device has the balance between the high degree of parallelism and power efficiency to generate the predicted ultrasound radio frequency (RF) data for the main transducer elements to double the field-of-view of the proposed platform without sacrificing the power consumption, cost, and fingerprint. This idea of recovering full-array ultrasound RF data from the sparse set of channels using a CNN-based framework running on the FPGA-based deep learning accelerator core in real-time enables high-performance, low-power, and low-cost wearable ultrasound imaging, paving the way for continuous bladder volume monitoring and broader applications in preventive healthcare.

Urinary tract infections caused by uropathogenic Escherichia coli (UPEC) are among the most common bacterial infections and disproportionally impact women. My thesis project focuses on revealing the bladder immune mechanisms against UPEC infections, with the hope of identifying novel drug targets to treat UTI. I discovered two G protein-couple receptors, Mrgprb2 and Mrgpra6, that suppress bladder immunity. Mrgprb2 is expressed on mast cells, while Mrgpra6 is specific for basophils. Both cell types are granulocytes involved in allergy and type 2 immunity, with mast cell being tissue resident and basophils circulating in the blood and recruited post-infection. Mutant mice lacking either Mrgprb2 or Mrgpra6 are better at clearing UPEC after infection. Using a new Mrgpra6Tdtomato reporter mouse, we show that basophils are rapidly recruited to the bladder epithelium after infection. This is the first time this rare immune cell was examined in UTI. qPCR analysis revealed that in the absence of these granulocyte GPCRs, the bladder produced higher Th17 cytokines IL-17, which may contribute to enhance clearance of bacterial infections. Together, these results have highlighted the roles of granulocytes during UTI. Antagonists for Mrgprs may enhance bladder immunity and improve UTI outcomes.

Primase-polymerase (Prim-Pol) is a family of enzymes that are critical for genome stability. These enzymes combine primase and DNA polymerase activities to participate in DNA replication, repair, and damage tolerance pathways. In Mycobacteria, several Prim-Pol enzymes are essential for repairing DNA lesions, with Prim-PolC specifically facilitating the repair of short DNA gaps of 1–3 nucleotides during excision repair. A distinctive feature of Prim-PolC is its C-terminal extension, known as loop 3, which is proposed to play an important role in recognizing DNA gaps. Experimental data suggest that the R179 residue acts as a gating side chain in the enzyme’s active site, providing greater stability in the closed conformation compared to the open conformation in the pre-catalytic structure. To investigate this, we performed 1.5 μs molecular dynamics (MD) simulations in each of the R179 conformations for two pre-catalytic complexes, as well as two post-catalytic complexes, all bound to gapped DNA substrates. Our analysis of root mean square fluctuations (RMSF), matrix-correlation, and normal mode analysis revealed significant differences in catalytically important residues between the pre- and post-catalytic structures. Additionally, non-bonded interaction energy calculations showed that the closed R179 conformation stabilizes the active site. Quantum mechanics/molecular mechanics (QM/MM) calculations further indicated that the ternary complex is more stable when R179 is in closed conformation, after the nucleotide addition reaction. The computational findings from these analyses will be further discussed and compared to experimental outcomes.

Suspended particles in fluid flows are typically modeled as spheres due to their isotropic properties, which simplify the resolution of particle shape, dynamics, and kinematics. While existing literature has extensively studied spheroidal particles that can either be prolate(rod-shaped) or oblate(disk-shaped), natural sediment grains often exhibit angularity. These edges as well as the non-isotropic shape can lead to different dynamic and kinematic behaviors compared to non-angular particles. In this study, we model angular particles as octahedrons using the Immersed Boundary Method. The first part of the study involves direct numerical simulations of oscillatory flow over a bed of angular particles to compare stress distributions with those over spherical beds. The second part examines the dynamic behavior differences among spherical, spheroidal, and octahedral particles in steady flow. Particle layout significantly influences rotation statistics, with upstream particle wakes affecting the torque on downstream particles.

Mammalian prostate is a androgen-dependent glandular organ which contributes to the production of seminal fluid. Given that the prostate comprises heterogenous cell types including epithelial and stromal cells, understanding the normal contribution of these cells during development is critical for deciphering their role in fertility and cancer. While prior research has focused on mature luminal, basal epithelial cells and stem cells of the prostatic epithelium, the role of prostate stromal cells remains elusive. In mammary gland ducts, which are structurally similar to prostatic ducts, mesenchymal cells identified by platelet-derived growth factor receptor alpha (PDGFRα) are reported to contribute to epithelial lineages during tissue regeneration. Here, we interrogated the functional contribution of PDGFRα+ cells to prostate development.

Adolescent disclosure is associated with closer parental relationships, less risky behaviors, and positive adolescent adjustment. However, some topics are more sensitive than others, and adolescents are less likely to share information about certain topics. This study aimed to determine which of four different topics (problem behaviors, school experiences, social experiences, and romantic interests) were most relevant in a sample of one hundred 14- to 18-year-old adolescents across seven consecutive days and whether they shared more with their mothers about those topics. The study found that adolescents considered the more sensitive topics of problem behaviors and romantic interests to be more relevant than the more mundane topics of school and socialization. However, they were more likely to share information about mundane topics with mothers. The study also identified demographic and relational factors that influenced the relevance of certain topics to adolescents and the amount of information they disclosed.

Passive thermoregulation of buildings presents a sustainable means to alleviate the ever-growing demand for fossil-fuel derived energy for thermal comfort due to climate change. Emerging radiative cooling (RC) technologies effectively achieve passive daytime cooling, but alone cannot fully satisfy the thermal comfort needs that arise from all types of weather – namely, the need for warmth in cold weather. Herein, we report a dual-functional material, composed of microencapsulated phase change materials (MPCMs) embedded in delignified wood pulp cellulose fibers (CFs) derived from wood wastes, for energy-efficient thermal management of buildings through RC and thermal energy storage (TES). In warm weather, RC assists in the re-crystallization of MPCMs. In cold weather, TES serves as a valuable complement to RC by offering space heating to offset excessive cooling. The dual-functional material exhibits 95% solar reflection that ascribes the scattering by the CFs and MPCMs, whereas the intrinsic emissivity of cellulose produces a strong radiative cooling effect. Meanwhile, TES through MPCMs achieves a latent heat of 156 J/g with excellent shape stability. Benefiting from the synergy of RC and TES, this dual-functional composite demonstrates substantial energy-saving potential through outdoor testing and COMSOL computational modeling. Furthermore, whole-building simulations by EnergyPlus show RCTES could reduce the annual energy use by heating, ventilation, and air conditioning by 7.2% in a hot and dry desert climate (Phoenix, Arizona)—an increase of 10 kWh/month and 2.4 kWh/month from the solely TES and RC cases, respectively. Lastly, fabricating this material from abundant wood waste promotes carbon sequestration for the development of sustainable building materials.

The small-scale turbulence resulting from wind-wave interactions profoundly affects the interfacial air-sea flux exchanges and, consequently, the long-term climate trends and short-term weather events. However, the correlation between this turbulence and surface wave characteristics has yet to remain challenging due to the complexity of near-surface dynamics. It is, in fact, extremely challenging to resolve the near-surface turbulence using either high-resolution experimental/field measurements or numerical techniques in most wind-wave conditions. Over the past few years, deep learning methods have been increasingly used to estimate turbulence from indirect and limited measurements. In this study, we developed a CNN model to reconstruct the turbulent flow above surface waves based on surface observations, such as surface elevation and velocity. The model is trained to minimize the induced reconstruction errors using the existing dataset of high-resolution measurements above wind-generated surface waves obtained by particle image velocimetry (PIV) and laser-induced fluorescence (LIF) techniques in the wind-wave tunnel facility. This novel approach seeks to enhance the understanding of near-surface turbulence and improve predictive capabilities crucial for wind-wave interactions.

DNA electrochemistry has proven beneficial for understanding fundamental charge transport features of DNA, DNA-protein interactions, enzymatic kinetics, and DNA-damaging anticancer drug activity. Still, fundamental insight on the overall electrical and electrochemical behavior of DNA electrochemistry is elusive due to the interplay of ionic and electronic effects. Electrochemical Impedance Spectroscopy (EIS) has been proven to be a useful tool in analyzing the ionic and electronic features of electrochemical systems. Here, we use EIS with equivalent circuit modeling to develop a fundamental equivalent circuit to represent our DNA electrochemistry system. When analyzing a working electrode modified with electrochemically active DNA monolayers and a mercaptohexanol backfilling agent, key capacitive and diffusive elements were uncovered. A modeled parallel-plate capacitor with 6.7-nm spacing matched to the height of the DNA monolayer revealed a solution dielectric constant of 76, consistent with an aqueous buffer solution. A capacitor attributed to double layer formation yielded a characteristic spacing of 0.9 nm, closely matching the ionically-blocking mercaptohexanol agent. A modeled Warburg diffusion element produced a diffusion constant of 6.2 x 1010 cm2/s, consistent with sodium ion diffusion across the DNA monolayer. Comparing to a DNA-free control revealed that the negatively charged DNA monolayer drew approximately triple the ionic charge to the electrode. surface. Additional mechanistic insight was revealed through protein studies with DNA helicases. These experiments provide a different outlook on the ionic and electronic features of DNA, giving a better understanding of how DNA reacts to charge.

Electrochemistry has evolved and matured alongside the need for consistent and precise measurements of modern sensors, electronics, and biochemical devices. Nonetheless, many electrochemical measurements require either human judgment, increasing the time for analysis and introducing user error, or analytic algorithms, which may be limited in scope and adaptability to nonideal voltammograms. In recent years, machine learning (ML) has emerged as a powerful and versatile tool for analysis. We utilize multiple ML approaches to create adaptive and predictive models to determine the parameters of electron transfer kinetics from limited and heterogeneous datasets of experimental square wave voltammograms. Our models include Gaussian process regressions (GPRs), randomized forests of decision trees, and ML ensemble techniques, all trained on experimental data modeled as quasi-reversible first-order reactions. The GPR method ardExp performed with the greatest accuracy in most situations but exhibited the longest computational time. Randomized forests produced respectable kinetic parameter estimates at a fraction of the computing time, while ensembles balanced accuracy and time. Adding one to three kinetic parameters successively improved the training of these ML models with each parameter added. Overall, this shows that ML can successfully predict the suite of kinetic parameters for surface-bound electrochemistry, improving when additional parameters are utilized as inputs for training.

Graph embeddings serve as the cornerstone for graph representation learning, facilitating the exploration of graphs by machine learning methods. However, prevalent deep learning techniques rely on black-box, high-dimensional graph embeddings. There is a pressing need for an interpretable, low-dimensional embedding approach to empower efficient graph visualization and provide practical tools to study graph datasets effectively. In this paper, we present a novel low-dimensional graph embedding method called \textit{Topological Evolution Rate (TopER)}, which simplifies a key concept of topological data analysis known as \textit{filtration}. TopER calculates the evolution rate of graph substructures induced by a filtration function on nodes or edges, resulting in interpretable 2D visualizations of graph datasets. Our experiments demonstrate that this new embedding method achieves highly competitive performance compared to the latest deep learning models in graph classification tasks on benchmark datasets. We further provide theoretical stability guarantees for TopER.

The escalating extreme climate conditions in the foreseeing future necessitate innovative solutions that provide energy-efficient building thermal regulation with a low carbon footprint, as the building sector accounts for 30-50% of the energy consumption in the United States and 30% of global carbon emissions. Thermal energy storage (TES) utilizing the latent heat stored in phase change materials (PCMs) during phase transition is one of the promising technologies to improve building energy efficiency by shifting and reducing peak loads. A critical challenge in PCM-based TES applications is the limited tunability of the operating temperature, especially for near-ambient applications, as most PCMs have a fixed transition temperature as designed. For instance, within buildings, the required operating temperature can significantly fluctuate between summer and winter, and even exhibit notable diurnal variations. This results in suboptimal PCM utilization, often leading to incomplete melting or no phase transition at all. Although a cascaded system utilizing multiple PCMs with different phase transition temperatures has been proposed to meet the varying use temperature requirement, the energy density of TES for each use case is reduced. Additionally, most PCMs store the energy through a solid-to-liquid phase change transition. Handling the liquid phase of PCMs during phase transitions (melting) has hindered practical TES implementation, especially in building envelopes. To address these challenges, this work reports a solid-state, shape-stabilized PCM with tunable phase transition temperatures for TES applications requiring various use temperatures. The tunability of phase transition temperatures of PCMs is achieved through moisture absorption/desorption into/from hygroscopic and semicrystalline PCMs. More moisture absorption into the PCMs reduces the crystallinity of the PCMs, leading to a lower phase transition temperature. Our experimental results demonstrate TES with an impressive phase transition temperature tunability of up to 7.2 °C (from ~18 °C to 35 °C). Meanwhile, outstanding shape stability over 120 hours without leakage during melting is also achieved by converting the traditional solid-to-liquid phase transition to a sold-to-gel transition through a sol-gel synthesis. Furthermore, the tunable TES exhibits exceptional cyclability, maintaining TES storage capacity of more than 500 cycles, without degradation in energy density and storage temperature, thus presenting a promising avenue for practical all-season building thermal comfort control applications. Our developed tunable TES will lead to adaptable energy storage for variable use cases and wide-ranging needs from end users, as well as benefit the grid flexibility requiring dynamic demand, which are keys to unlocking high-efficiency TES systems for decarbonization

Zero-dimensional (0D) hybrid organic-inorganic metal halides with luminescent properties have drawn attention for optical applications like solid-state lighting, X-ray detection, etc. Developing highly emissive and environment-friendly materials that show high performance, but low cost was one of the goals for this project. In this research work, we worked on a series of A2SbCl5. We varied materials for A site with different long-range materials and made a correlation about how their performances are changing with the increasing carbon chain. Here in this research work, we reported three novel hybrid compounds [BDTA]2SbCl5 (BDTA+= benzyldimethyltetradecylammonium), [BDHA]2SbCl5 (BDHA+=benzyl dimethyl hexadecyl ammonium) and [BDSA]2SbCl5 (BDSA+=benzyldimethylstearylammonium). All three compounds exhibit a strong bright orange emission where [BDSA]2SbCl5 emits at 620nm for a low energy excitation at 360nm. In addition to that, it also exhibits an emission at around 470nm for high energy excitation at 370nm. This compound [BDSA]2SbCl5 has a large stoke shift of 260nm and a PLQY above 90%. Apart from optical properties, these compounds show structural variation in the single crystal XRD at low and room temperature which might also have a significant impact on their optical properties as well. Here, we have mentioned the synthesis process for these crystals and the unique single-crystal synthesis procedure which works for long-range structures where growing single crystals in conventional ways like the solvent method, antisolvent method, and slow evaporation failed to grow single crystals. This research work explains the struggle of growing single crystals with long-range carbon chain materials and how the size of A site correlated with photoluminescent quantum yield as well as probable applications due to high PLQY.

This work has demonstrated that LSE significantly enhances the cycling performance and capacity retention of LNMO spinel cathodes under high-voltage conditions (4.85V). Our results indicate that LSE maintains 94% capacity retention after 200 cycles in Li|LNMO half cells and up to 99% in LTO|LNMO full cells, significantly outperforming conventional electrolytes. These improvements are critical in addressing the challenges of high-voltage battery operations, which often suffer from rapid capacity degradation and electrolyte instability. The addition of FEC to the electrolyte further improved their electrochemical performance, demonstrating that strategic molecular enhancements can substantially augment battery performance. MD simulations provided further insight into the mechanisms underpinning the enhanced performance for the first time. These simulations revealed that LSE optimizes the solvation structure around Li+ ions, effectively reducing the presence of free solvent molecules and thus minimizing deleterious side reactions at the electrode-electrolyte interface. These findings advance our understanding of high-voltage electrolyte behavior and underscore the potential of advanced electrolyte formulations to improve the stability and longevity of lithium-ion batteries. By fostering a deeper comprehension of solvation dynamics and interface chemistry, this research supports the development of more robust and efficient energy storage systems, paving the way for the wider adoption of high-energy-density cathodes like LNMO in commercial applications. This study not only highlights the benefits of localized saturated electrolytes but also sets the stage for future innovations in battery technology.

Halide perovskites are premier photovoltaics due to their remarkable properties, such as defect tolerance and charge transport that arise from dynamic disorder of the crystal lattice. These structural dynamics have been abundantly explored in 3D perovskites and attributed to expression of ns2 lone pairs, but few studies have examined such dynamics in 2D perovskites. 3D perovskites have the formula ABX3 and consist of a B2+ cation coordinated with 6 X- halide anions in 3D corner-connected octahedra, with cuboctahedral voids filled by A+ cations. The A2BX4 Ruddlesden-Popper perovskite (RP) is a quasi-2D <100> structure variant where B2+ ion occupies corner connected octahedra in the ab plane, while A+ ions space out the layers. We aim to determine if inorganic 2D RP perovskites Cs2MI2Cl2 (M=Sr, Cd, Sn, & Pb) exhibit the same anharmonic lone-pair driven lattice dynamics as their 3D counterparts. Our series enables us to connect these dynamics to the B2+ lone pair (or lack thereof), while ongoing work on their photoluminescence and Raman spectroscopy enables us to relate these dynamics to optoelectronic properties.

On June 1, 1942, under the rule of Marshall Ion Antonescu, Romania began deporting Roma to the newly-acquired region east of the Dniester River, named Transnistria. The first wave of deportations – that targeting nomadic Roma – ran from June 1 to August 15. 1942. The second wave – of which sedentary Roma fell victim – began on 12 September and concluded somewhere around 21 September. Newly acquired data from the Romanian Central National Historic Archives sheds light on this deportation process. Integrating data from this archive with additional primary source data available, and further aided by the Power BI program, I have created visualizations, which I hope tell the story of the Romanian Roma from the abrupt nature of their removal to the shocking breadth of their deportation.

Background: The most frequently reported medicinal use of cannabis in adults is for the management of sleep and pain. Because cannabis provides a dual reinforcing effect for chronic pain sufferers by alleviating pain and aiding in sleep, pain sufferers may be at heightened sensitivity to the effects of cannabis. Objectives: We examined the acute and chronic effects of cannabis use on chronic pain and sleep and tested whether slow-wave sleep (SWS) mediates the relationship between cannabis use and pain. Methods: We recruited 62 adults for a home-based study where participants were asked to record their sleep via ambulatory electroencephalogram (EEG), pain levels, and cannabis use over seven consecutive days/nights. A Mixed model repeated measures ANCOVA was used to test the interactions between presence of regular cannabis use (chronic effect), cannabis use days during the in-home data collection (acute effect), and chronic pain on SWS. Hayes Process MACRO was used to test the mediating effect of SWS on cannabis use and pain levels. Results: The results indicated that chronic effects of cannabis use was associated with decreased SWS. There was an interaction between acute cannabis use and pain such that pain sufferers had increased SWS on days when cannabis was used, but decreased SWS on days without cannabis use. Additionally, SWS mediated the relationship between years of cannabis use and pain levels, where more years of using cannabis decreased SWS and increased pain. Conclusions: Our results showed that there are acute benefits of cannabis on sleep for pain sufferers, but these benefits diminish with continued cannabis use.

Aquaporin 4 (AQP4) is a bidirectional water channel primarily expressed at the endfeet of astrocytes in the Central Nervous System (CNS). This protein plays a critical role in maintaining fluid homeostasis, by facilitating the exchange between cerebrospinal fluid and interstitial fluid which enables waste clearance from the brain parenchyma. Our lab has discovered a novel variant of AQP4, termed AQP4X, which has a C-terminally extended region arising from translational readthrough allowing for exclusive expression in the perivascular regions throughout the CNS. Uniquely, in AD perivascular AQP4 is redistributed into the parenchyma and downregulation of AQP4X which leads us to investigate if a similar differential expression of AQP4X also occurs in Huntington’s Disease (HD). We quantified the expression of AQP4 and AQP4X using immunostaining and western blot and found that both are significantly downregulated in the cortex of 9-month ZQ175 – a HD mouse model. We next asked if its downregulation can be redressed by crossing the HD mouse to our Aqp4All_X mouse. The Aqp4All_X mouse has its Aqp4 stop codon mutated to a sense codon to allow 100% readthrough (as compared to 20% readthrough in the wild-type mouse). Indeed, we observed the rescue of AQP4X in the ZQ175; Aqp4All_X double heterozygote mouse. Our ongoing experiments explore the role of AQP4X in the clearance of mutant Huntingtin (mHTT) from the brain. Overall, our findings suggest that both AQP4 and AQP4X are downregulated in HD, and the expression of at least AQP4X may be restorable with our genetic models of AQP4X.

Inflammatory bowel disease (IBD), which includes both ulcerative colitis and Crohn’s disease, is a relapsing disease of the gastrointestinal tract. IBD is characterized by an excessive immune response leading to tissue damage and loss of function. Damage of the large intestines, including both the colon and rectum, is associated with visceral pain and common clinical manifestations of IBD. From a pathophysiological standpoint, inflammation is associated with a wound-healing response aimed at restoring tissue structure. In turn, wound healing leads to alterations in the content and organization of the extracellular matrix (ECM). This key process, commonly referred to as ECM remodeling, was recently identified as an integral part of IBD progression. However, ECM remodeling includes deposition, degradation, and reorganization of key structural constituents such as fibrillar collagen, thus leading to disparate changes in the mechanical function of colorectal tissues. While both tissue softening and stiffening have been implicated with IBD, the regional biomechanical changes that compromise colonic tissue function in IBD are not well understood. This is due, in part, to a lack of rigorous experimental data on a complex and heterogeneous tissue such as the colon. We utilized a custom biaxial testing device to map the regional biomechanical properties of an inflamed colon. To model the inflamed colon, we administered dextran sulphate sodium (DSS), which induces acute colitis in mice, into drinking water. Then, we divided the entire mouse colon into four sections – two proximal and two distal – and for each section we investigated passive biomechanics, pertaining to near physiological conditions, in both the axial and circumferential direction. Using our biaxial setup, we report complex patterns of colonic tissue mechanics in response to acute colitis, with a proximal loss in distensibility and a distal loss in extensibility. With our most recent study, we expanded on this research methodology by incorporating high-fat diet into the DSS-induced murine model to determine its effect on biomechanical changes of acute colitis. Overall, this work highlights the importance of investigating regional biaxial changes in colonic tissue mechanics to gain an improved understanding of the role played by ECM remodeling in IBD progression.

Étienne-Jules Marey (1830–1904) is a critical figure in nineteenth-century studies of animal and human motion. His research moved fluidly from observation and the graphical notation of data to photographic recording of still images and their animation in both two- and three-dimensions via optical devices as he pursued deeper understanding of movement and forms of translation. His use of the three-dimensional zoetrope and the research trajectory that led to this and other sculptural outcomes embraced the unique strengths of different media: moving from the body to the image and back to the body (in the form of cast birds) extended understanding at each step while also signaling a certain embrace of the aesthetic, perhaps poetic, nature of his subject matter. Marey is distinguished by this translation of form and thinking across media and systems — thinking by analogy — such as in the application of his understanding of mechanical systems to anatomical ones. Both for his lasting diverse influence as well as his expansive body of research, Marey remains exceptional for his analogical thinking, the translation of form across time and space, and the poetics embedded in his work — compelled by what I read as a sensibility of great curiosity and wonder.

Cetuximab (Cet)-IRDye800CW, among other antibody-IRDye800CW conjugates, is a potentially effective tool for delineating tumor margins during fluorescence image-guided surgery (IGS). However, residual disease often leads to recurrence. Photodynamic therapy (PDT) following IGS is proposed as an approach to eliminate residual disease but suffers from a lack of molecular specificity for cancer cells. Antibody-targeted PDT offers a potential solution for this specificity problem. In this study, we show, for the first time, that Cet-IRDye800CW is capable of antibody-targeted PDT in vitro when the payload of dye molecules is increased from 2 (clinical version) to 11 per antibody. Cet-IRDye800CW (1:11) produces singlet oxygen, hydroxyl radicals, and peroxynitrite upon activation with 810 nm light. In vitro assays on FaDu head and neck cancer cells confirm that Cet-IRDye800CW (1:11) maintains cancer cell binding specificity and is capable of inducing up to ∼90% phototoxicity in FaDu cancer cells. The phototoxicity of Cet-IRDye800CW conjugates using 810 nm light follows a dye payload-dependent trend. Cet-IRDye800CW (1:11) is also found to be more phototoxic to FaDu cancer cells and less toxic in the dark than the approved chromophore indocyanine green, which can also act as a PDT agent. We propose that antibody-targeted PDT using high-payload Cet-IRDye800CW (1:11) could hold potential for eliminating residual disease postoperatively when using sustained illumination devices, such as fiber optic patches and implantable surgical bed balloon applicators. This approach could also potentially be applicable to a wide variety of resectable cancers that are amenable to IGS-PDT, using their respective approved full-length antibodies as a template for high-payload IRDye800CW conjugation.

If a work of art is the sum of its parts, then Phil Tippett’s stop-motion epic Mad God has a very large sum indeed. This animated horror feature is a Frankenstein’s monster of stitched-together influences ranging from Dante’s Divine Comedy to Hieronymous Bosch’s chaotic religious paintings, from comics artists such as Moebius and Richard Corben to the psychoanalytic theories of Freud. These influences are unusually potent due to Tippett’s strong authorial position as an independent director who crafted Mad God over the course of three decades. This poster depicts one half of a larger argument that I make in a longer paper, which is linked with a QR code on my poster. In this paper, I argue that Tippett leans on his strong authorial position to effectively bring together grotesque adult horror and stop-motion animation in Mad God, and that in doing so, he magnifies and reaffirms a visceral materiality that has always made stop-motion compelling. Tippett lays his carefully-crafted network of influences as a foundation for this visceral materiality; in doing so, he not only incorporates the histories of cinema and animation as a source of Mad God’s inspiration, he also writes this unique masterpiece into those histories through his uncommonly passionate dedication to craft.

The monitoring of neurochemical activities in freely behaving animals enables novel studies into the dopaminergic neural circuits for behavioral studies regarding reward, aversion, and learning.

Oral cavity and pharyngeal cancer affect approximately 560,000 individuals globally each year. Despite advances in therapeutic interventions, survival rates remain dismal, with prognoses typically improving through early detection. However, a critical gap persists in rigorous quantitative methods for analyzing histopathological features pertinent to clinical diagnosis, such as cellular morphology and the number of epithelial layers. In response, we developed a Bayesian nonparametric approach to detect epithelial layer counts based on features extracted from H\&E-stained whole-slide images. This method integrates AI-driven histological reconstruction, utilizing both spatial and morphological data, within a Bayesian finite mixture modeling framework to estimate layer numbers. Simulation results validate the feasibility of our approach for segmenting oral epithelium and accurately counting its layers. Furthermore, a case study of oral dysplasia shows the clinical relevance between detected oral epithelium layer numbers and severities of dysplasia.

The primary problem addressed in this research is the inadequate performance of current ASR systems in noisy and adverse acoustic environments. While deep learning has significantly improved ASR accuracy in controlled structured settings, these systems still struggle in real-world conditions where subjects speak freely (e.g. conversational interactions between 2 or more team members; not prompted/read speech), and most importantly noise, reverberation, and other distortions that are prevalent. Traditional acoustic features, such as MFCC and FBANK, are not sufficient to address these challenges, as they are highly sensitive to such distortions. Additionally, existing ASR models often treat different acoustic conditions uniformly, without explicitly considering the varying nature of these conditions. This lack of scenario awareness further limits the models’ ability to adapt and perform well in diverse environments. This research aims to provide a comprehensive solution to the problem of robust speech recognition by integrating advanced acoustic modeling techniques and feature fusion strategies. The proposed methods will be evaluated using well-established speech corpora, including the Wall Street Journal (WSJ), LibriSpeech 100-hour corpus, CHiME-4, and Fearless Steps Challenge Phase 2/Phase 4 corpus, to demonstrate their effectiveness in real-world scenarios. By addressing the limitations of current ASR systems and introducing novel approaches for feature extraction and modeling, this research has the potential to significantly enhance the robustness and reliability of ASR technology, making it more suitable for practical applications across various domains.

Cyberattacks have become one of the most critical threats for firms in today’s business environment. Our study investigates how CSR and CSIR affect a firm’s likelihood of becoming a cyberattack target. We identify four key findings: (1) both CSR and CSIR increase cyberattack risk, (2) CSR activities may signal resource abundance, attracting external attackers, (3) CSIR has a stronger effect on internal cybercriminals sensitive to a firm’s perceived moral value, and (4) stock returns weaken the relationship between CSR and external cyberattack. These insights reveal unexpected connections between a firm’s social practices and its cybersecurity vulnerabilities, expanding our understanding of strategic management in the digital age and highlighting the complex interplay between corporate behavior and cyber risks.

In this work, we study the potential of bilayer transition metal dichalcogenides (TMDs) for next-generation electronic devices. TMDs, consisting of stacked layers of transition metals (Mo, W) and chalcogens (S, Se, Te), show a strong dependence of their electronic properties on the number of layers. Using first-principles methods, we investigate the impact of different insulating substrates (SiO2, HfO2, hBN) on carrier transport and scattering processes. Our results demonstrate that bilayer configurations, especially in materials like WSe2 and WS2, can enhance carrier mobility, making them promising candidates for high-performance transistors.

P-type ATPases are a superfamily of ubiquitous, primary active transmembrane ion pumps, essential for maintaining cellular homeostasis and known for their complex multi-domain architectures. Among them, P1B-type ATPases, which transport first-row transition metals, are comparatively understudied. In this work, we employed an interdisciplinary computational and experimental approach to design functional De Novo P1B-type ATPases. Our collaborators used unsupervised machine learning (ML) to generate a sequence landscape of 250,000 potential candidates from a multiple sequence alignment of over 13,500 native P1B-type proteins. Molecular dynamics (MD) simulations were then applied to rationally filter and assess a subset of these sequences. MD simulations helped evaluate whether key conformational changes in the A-Domain and transmembrane (TM) domain, critical for protein functionality, were preserved in the selected De Novo candidates. Experimental validation confirmed that despite extensive sequence modifications (approximately 200 mutations), the MD-screened sequences retained the catalytic activity and ion transport properties characteristic of native P1B-type ATPases. Our findings highlight the growing utility of computational tools in protein engineering, revealing latent information within sequence data. This study underscores the potential of combining ML and MD in advancing De Novo protein design, opening new avenues for in silico-driven engineering of complex biological systems. Our next aim is to use this approach for the rational understanding and design of these proteins’ selectivity.

Children’s behaviors are known to impact parents’ responsiveness (Belsky, 1984). During family mealtimes, parents may find it difficult to remain responsive towards their preschool children when they are highly active, particularly in the absence of a supportive coparent. We analyzed children’s observed activity levels, parents’ supportive coparenting, and mothers’ and fathers’ emotional responsiveness during family mealtimes from 99 families across a 7-day period using a moderated multilevel model. On days when coders observed more supportive coparenting, mothers and fathers were more emotionally responsive to children. On days when children had higher activity levels at the meal, coders rated more supportive coparenting between parents. Coparenting did not moderate relations between child activity level and parent responsiveness. Findings suggest that coparenting during this shared time as a family can improve the quality of parent-child interaction, and that children’s active behaviors may necessitate greater cooperation among the parenting team.

Cochlear implants (CIs) are surgically implanted medical devices that rely on real-time digital signal processing (DSP) strategies for acoustic-to-sound conversion. Because most fixed strategies have been implemented and tested only in clinical and laboratory settings, the ability for CI systems to adapt to varied feedback in spontaneous environments is limited. To help allocate real-time CI feedback in naturalistic spaces, this study proposes the first CI framework for situational signal processing: “Emaging” and considers CI vocoded testing approaches to help record and document collected data when CI users are often difficult to recruit for experimental testing. This unprecedented application implements ecological momentary assessment (EMA), an “on-the-go” data collection method for instantaneous feedback from CI subjects. The “Emaging” algorithm solution runs on wearables and portable devices alongside CCi-MOBILE, a customized portable CI signal processing platform. This study evaluates two parameters of EMA for the CI participant: sound source localization (SSL) and sound source identification (SSI) for non-spoken sounds. With “Emaging”, CI users document and “tag” situational data from their naturalistic environments in real-time. Due to the many constraints with CI subject recruitment and testing, vocoded simulations with normal hearing (NH) participants can contribute valuable information and considerations aptly integrated with CI algorithm development. “Emaging” and its collected responses from CI, NH, and vocoded (V) subjects provides a unique opportunity for next generational CI processing design that integrates effective sound coding strategies for non-linguistic sound intelligibility and source localization.

Autistic children show characteristic differences in social development and challenges with engagement during social interactions compared to their non-autistic peers. Importantly, autistic children do not often use their language to communicate socially (i.e., for the purpose of sharing information or attention with a communicative partner). However, most language measures do not distinguish the social intention of children’s utterances, and previous studies have not examined the relationships between different types of engagement with a parent and language produced within an interaction. The current study investigated the overall verbosity and social motivation behind the language produced by young autistic children within various types of engagement (ranging from less social to more social) during naturalistic parent-child interactions. Results indicate that children spent the most time in and talked most during less social types of engagement. Notably, however, the children in this sample produced very few utterances with high social motivation scores. These findings emphasize the difference between autistic children’s verbosity and their use of socially motivated utterances, highlighting the need to differentiate expressive vocabulary from social communication as two distinct outcomes for this population.

Accumulating evidence has demonstrated that alcohol disturbs metabolic functions in adipose tissues, causing progressive liver damage. Adipocyte Gs-coupled signaling plays a crucial role in regulating lipolysis and thermogenesis. However, the role of Gs-coupled signaling in alcohol-associated liver disease is largely unknown. To address this question, we generated an inducible transgenic mouse model (Adipo.GsD), which allows specific and spatiotemporal stimulation of Gs signaling in adipose tissues by administration of an exogenous agent (clozapine N-oxide, CNO). Male mice were fed a liquid diet containing either 5% alcohol or pair-fed a control diet along with CNO administration for 10 days, following by a single binge of alcohol (5g/kg body weight) or isocaloric maltose dextrin on the 11th day. We found that selective activation of adipocyte Gs-signaling protected mice from acute-on-chronic alcohol-induced fatty liver and hypertriglyceridemia. In addition, Gs activated mice exhibited decreased body weight and fat mass as well as increased expression of thermogenic genes in inguinal white adipose tissue, indicating an enhancement in energy expenditure. Another interesting finding was that mice with Gs activation in adipocytes showed significantly higher plasma adiponectin levels. Given the important role of adiponectin in lipid metabolism and thermoregulation, these results indicate that adipocyte Gs-signaling activation protects against alcoholic liver damage via adiponectin. Our findings could be applied in the future for the identification of potential therapeutic targets to prevent and combat alcohol-induced liver injury.

Alcohol-associated liver disease (ALD) is a major global health concern and a leading preventable cause of death, driven by both chronic and acute excessive alcohol consumption. While the detrimental effects of alcohol on liver health are well-documented, its impact on adipose tissue and the underlying mechanisms remain largely unexplored. Alcohol consumption has been proven to cause change in the fatty acid synthase (FASN) expression levels in the adipose tissues. FASN is a key enzyme involved in lipogenesis – converting malonyl CoA into palmitate, which is essential for lipid metabolism. To investigate how FASN ablation influences adipose tissue function and liver lipid metabolism under acute and acute-on-chronic alcohol exposure, we generated adipocyte specific FASN knockout mice (FASNFKO). Following acute alcohol intake, we found that FASNFKO mice exhibited an increase in free fatty acid (FFA) mobilization and liver triglyceride (TG) accumulation, indicating elevated lipolysis in adipose tissues and the development of alcoholic fatty liver disease. In contrast, after acute-on-chronic alcohol drinking, FASNFKO mice showed less plasma FFA and increased expression of thermogenic genes in inguinal white adipose tissue (iWAT). Moreover, alcohol-fed FASNFKO mice had reduced hepatic content of malondialdehyde (MDA), a marker of oxidative stress. Further investigations will focus on elucidating the metabolic consequences of FASN deficiency in the liver in response to different alcohol drinking patterns and the underlying mechanisms how adipocyte FASN ablation affects thermogenic function after alcohol exposures.

Wind farms play an essential role in increasing green energy usage. However, the chaotic nature of the atmosphere and the large size of wind farms, coupled with the cost of traditional measurement techniques such as large meteorological towers instrumented with wind anemometers, make studying wind farms challenging. New developments in the technology of light detection and ranging (LiDAR) make measuring large regions of the atmosphere much easier. Combining the data from LiDARs with massive amounts of data collected from operating wind turbines is still difficult, but new advances in machine learning (ML) are perfectly suited to processing big data in efficient ways. We present the different ways the WindFluX lab at UT Dallas combines LiDAR measurements and ML to study wind farms along with new insights into the interaction between wind turbines and the atmosphere.

We extend the work of D. Armstrong, V. Reiner, and B. Rhoades on noncrossing parking functions in classical real reflection groups to irreducible well-generated complex reflection groups, W. We define two parking spaces for such W, one defined purely combinatorially and the other defined algebraically. Each space carries a WxC-action, and we describe an equivariant isomorphism between the two. In doing so, we enumerate the noncrossing parking functions. We also extend our results to the Fuss case.

UTDrive-MOBILE-App: The next generation platform for intelligent transportation and driver modeling

Chronic neck pain is a debilitating condition affecting millions worldwide. Recent studies suggest that immunoglobulin gamma 4 (IgG4) may play a key role in modulating dorsal root ganglia (DRG) excitability by influencing potassium channel function. Furthermore, studies using rodent DRG models and human serum have demonstrated that the B cell–IgG–Fc gamma receptor axis is critical for the development of neuropathic pain. We report the first case study involving the clinical and spatial transcriptomic characterization of DRG tissue from a patient with elevated immunoglobulin heavy constant gamma 4 gene (IGHG4) expression and chronic neck pain. The patient, a 63-year-old male, initially experienced neck and arm pain localized to the left cervical region, which later spread to the right. Over two years, his symptoms progressed, leading to weakness, clumsiness, and numbness in both arms, with more recent symptoms affecting the legs. Quantitative Sensory Testing (QST) revealed a 30% higher pain level on the left side of the neck compared to the right (average pain scores of 8 vs. 6.5), along with diminished light touch sensitivity in both extremities. Radiographic imaging showed narrowing of the spinal canal (spinal canal stenosis) from the foramen magnum to C7/T1, ossification of the posterior longitudinal ligament (OPLL), and compression of the spinal nerves (neural foraminal stenosis) at multiple cervical levels, leading to a C1-T2 posterior fusion and C1-7 laminectomy. We retrieved bilateral C1/C2 fresh frozen DRGs during surgery that underwent transcriptomic analysis using Visium 10X Genomics spatial sequencing. We identified elevated IGHG4 gene expression, particularly in the peripheral and certain central regions of the DRG. Computational analysis using the Seurat and SONAR packages in R, in conjunction with previously published human whole-cell sequencing DRG datasets for cell type identification, revealed B cell infiltration and colocalization in areas of high IGHG4 expression. Our future studies aim to validate IGHG4 expression at the protein level through immunohistochemistry and further explore the role of IGHG4 in the pathogenesis of peripheral neuropathies. Acknowledgments: This project supported by NIH grants U19NS130608 to TJP and MC, and NIH 1R01AR078192-01A1 to MC, CPH and TJP

Despite advancements in cardiovascular engineering, heart diseases remain a leading cause of mortality. The limited understanding of the underlying mechanisms of cardiac dysfunction at the cellular level restricts the development of effective screening and therapeutic methods. To address this, we have developed a framework incorporating light field detection and individual cell tracking to capture real-time volumetric data in zebrafish hearts, which share structural and electrical similarities with the human heart and generate 120 to 180 beats per minute. Our results indicate that the in-house system achieves an acquisition speed of 200 volumes per second, with resolutions of up to 5.02 ± 0.54 µm laterally and 9.02 ± 1.11 µm axially across the entire depth, using the estimated-maximized-smoothed (EMS) deconvolution method. The subsequent deep learning-based cell trackers allow us to investigate contractile dynamics, including cellular displacement and velocity, followed by volumetric tracking of specific cells of interest from end-systole to end-diastole in an interactive environment. Collectively, our strategy facilitates real-time volumetric imaging and assessment of contractile dynamics across the entire ventricle at the cellular resolution over multiple cycles, providing significant potential for exploring intercellular interactions in both health and disease.

Introduction: Substance use disorders (SUDs) have emerged as a pressing public health crisis in the United States, with adolescent substance use often leading to SUD in adulthood. Alcohol and cannabis are two of the most commonly used substances among adolescents. Thus it is important to develop tools that can help stem the progress of adolescent/young adult substance users to SUD. To this end, we developed the first-ever absolute risk prediction model that predicts the personalized risk of developing alcohol use disorder (AUD) and cannabis use disorder (CUD) for an adolescent or young adult alcohol/cannabis user over any user-specified time periods. Methods: We train a Bayesian machine learning model using data from the National Longitudinal Study of Adolescent to Adult Health. The model provides the risks of developing AUD and CUD in a given time period based on personal risk factors of an adolescent/youth user of alcohol or cannabis after accounting for the competing risk of mortality. Model performance is evaluated using 5-fold cross-validation (CV), with the area under the curve (AUC) and the ratio of expected to observed cases (E/O) as performance metrics. External validation of the final model was conducted using an independent dataset from Add Health. Results: The proposed model has 9 predictors and 3 sub-models, each predicting risks for different user groups: alcohol-only users, cannabis-only users, and users of both substances. For predicting the risks of developing AUD and CUD within 10 years from the age of first substance use, the respective AUC for AUD and CUD are (a) 0.69 and 0.66 using 5-fold CV, and (b) 0.70 and 0.61 on the validation dataset. The respective E/O for AUD and CUD are (a) 0.87 and 0.84 via 5-fold CV, and (b) 1.19 and 0.90 on the validation dataset. This indicates good discrimination and calibration performances of the model. Conclusion: The personalized absolute risk of AUD and CUD provided by the proposed model may assist clinicians in identifying adolescent and young adult substance users who are at high risk of developing SUD, enabling clinically appropriate interventions.

Our mechanistic knowledge of how the microbiota influence human physiology is in its infancy. Key to understanding the function of the microbiome is dissection of metabolic interactions between the host and the resident microbiota. This is especially true of the microbiota colonizing the female urogenital tract which includes the vagina, urethra, and urinary bladder. The female urogenital microbiome (FUM) is a low diversity microbial ecosystem predominated by specific species of the genus Lactobacillus (Lb) that are adapted to mammalian hosts. Lactobacillus crispatus and Lactobacillus gasseri, for example, produce D-lactate, hydrogen peroxide, and bacteriocins that chemically modify the local environment to inhibit pathogen colonization. Dysbiosis, or disruption of Lactobacillus populations, within the FUM is associated with conditions like recurrent urinary tract infection (rUTI), urinary incontinence, and overactive bladder. Despite their importance to female urogenital health, the metabolic requirements of FUM Lactobacillus species remain largely undefined. We have discovered these FUM Lactobacillus species exhibit fatty acid (FA) auxotrophy, meaning that they cannot synthesize long-chain fatty acids necessary for lipid biosynthesis and rely on external fatty acid and lipid sources. Through gene synteny analysis we identified the absence of the Fab operon—responsible for type II fatty acid biosynthesis in Gram-positive bacteria—in specifically in key FUM members like L. crispatus and L. gasseri. Growth assays using a chemically defined media (CDM) demonstrated that L. crispatus and L. gasseri require fatty acid supplementation, in the form of both cis and trans unsaturated fatty acids, while non-urogenital species of the related genera, Lacticaseibacillus and Lactiplantibacillus, require no supplementation. Metabolic tracing experiments with universally labeled glucose confirmed that de novo fatty acid biosynthesis occurs in non-urogenital species (e.g. Lacticaseibacillus rhamnosus) but not in L. crispatus and L. rhamnosus. Finally, we measured the direct incorporation of host lipids purified from human bladder epithelial cells and found that L. crispatus and L. gasseri could not only use them as a fatty acid source, but directly incorporate diverse host-derived fatty acids into their membranes. These findings greatly expand our understanding of the metabolic relationship between FUM Lactobacillus species and the human host and will be essential in the design of effective prebiotic and probiotic therapies to improve female urogenital health.