Browse Articles
Discover research articles across all indexed journals
Conservative port-to-port funneling of light in nonlinear photonic lattices
Machine learning prediction of overall survival in patients with cT1b renal cell carcinoma after surgical resection using the SEER database
Observation of returning Thouless pumping
Simulation-based inference for subject-specific tuning of middle ear finite-element models towards personalized objective diagnosis
Abstract Computational models, particularly finite-element (FE) models, are essential for interpreting experimental data and predicting system behavior, especially when direct measurements are limited. Tuning these models is particularly challenging when a large number of parameters are involved. Traditional methods, such as sensitivity analyses, are time-consuming and often provide only a single set of parameter values, focusing on reproducing averaged trends rather than capturing experimental variability. New approaches are needed to make computational models more adaptable to patient-specific clinical applications. We applied simulation-based inference (SBI) using neural posterior estimation (NPE) to tune an FE model of the human middle ear against subject-specific data. The training dataset consisted of 10,000 FE simulations of stapes velocity, ear-canal input impedance, and absorbance, paired with seven FE parameter values sampled within plausible ranges. By using simulated data, we generated a diverse training dataset, enabling efficient learning by the neural network (NN). The NN learned the association between parameters and simulation outcomes, providing a probability distribution of parameter values, which could be used to produce subject-specific computational inferences. By accounting for noise and test–retest variability, the method provided a probability distribution of parameters, rather than a single set, fitting three experimental datasets simultaneously. Importantly, examining the inferred parameter distributions alongside prior knowledge of normal ranges enables individualized differential inference used for diagnosis. SBI offers an objective alternative to sensitivity analyses, uncovering parameter interactions, supporting personalized diagnosis and treatment, and compensating for limited clinical training data. This method is applicable to any computational model, enhancing its potential for improved patient outcomes.
Rethinking infrastructure design from component failure to systemic resilience
Abstract Bridge design typically uses load-based design criteria focused on risk thresholds from engineering practice and standards, overlooking cascading effects on connected infrastructure and regional economies. We argue for a systems-based design that balances risk reduction with resilience — the capacity to recover from disruptions. Using the Francis Scott Key Bridge collapse as a case study, we estimate economic impacts assuming impact on local transportation networks only as well as integrating cascading failures on surrounding infrastructure (e.g., closure of the Port of Baltimore), employing the regional economic model TranSight. Results show combined bridge-and-port disruptions produce substantially larger losses in GDP, employment, disposable income, and labor force, with some indicators not recovering until 2040. The Baltimore region exhibits lower resilience to compounding shocks, highlighting the need for a resilience-based framework that considers interconnected infrastructure. We conclude infrastructure design must move beyond component-focused risk criteria toward an explicit, quantifiable resilience framework.
Long-term susceptible fractions in networked epidemic models and their relation to the basic reproduction number
Abstract Compartmental epidemic models with dynamics that evolve over a graph network have gained considerable importance in recent years. Fundamental to these models is an important threshold known as the basic reproduction number (BRN) that aims to capture, on average, the tendency of a communicable disease to spread. In this paper, we develop two complementing frameworks that provide insights into the long term evolution of a wide range of compartmental epidemic models, including group and networked processes, exploring the positive feedback that is inherent in such models. Specifically, for the case of a group (resp. networked) process, we show that the proportion of the population that is susceptible to a disease (resp. the susceptible proportion in at least one subgroup) tends to a limit that is bounded from above by the reciprocal of the BRN of the respective model, thereby establishing that the BRN encodes critical information on the level of penetration of the disease into a subpopulation. The two substantially distinct scenarios, where the disease remains always present in the population or not, are discussed and the significance of the bound explained. To verify the validity of our conclusions, we apply the developed frameworks to examining various networked epidemic models, including a model that was recently introduced for a bi-virus process.
Development and validation of a clinical wearable deep learning based continuous inhospital deterioration prediction model
Runge-kutta method inspired aerial image dehazing network in YUV space
Unveiling the cut-and-repair cycle of designer nucleases in human stem and T cells via CLEAR-time dPCR
Abstract DNA repair mechanisms in human primary cells, including error-free repair, and, recurrent nuclease cleavage events, remain largely uncharacterised. We elucidate gene-editing related repair processes using Cleavage and Lesion Evaluation via Absolute Real-time dPCR (CLEAR-time dPCR), an ensemble of multiplexed dPCR assays that quantifies genome integrity at targeted sites. Utilising CLEAR-time dPCR we track active DSBs, small indels, large deletions, and other aberrations in absolute terms in clinically relevant edited cells, including HSPCs, iPSCs, and T-cells. By quantifying up to 90% of loci with unresolved DSBs, CLEAR-time dPCR reveals biases inherent to conventional mutation screening assays. Furthermore, we accurately quantify DNA repair precision, revealing prevalent scarless repair after blunt and staggered end DSBs and recurrent nucleases cleavage. This work provides one of the most precise analyses of DNA repair and mutation dynamics, paving the way for mechanistic studies to advance gene therapy, designer editors, and small molecule discovery.
Wave energy and other environmental drivers as predictors of seeded-coral performance on the great barrier reef
Abstract Wave energy shapes coral reef communities, yet its influence on early coral survival and growth remains poorly understood, limiting its use in reef restoration planning. This study investigated the survival and growth of three Acropora species deployed on seeding devices across a wave energy gradient at three reefs on the Great Barrier Reef. After 1.5-2 years, survival varied significantly within reefs, among sites, and among species, with highest average yield at Moore Reef ( A. millepora , 32% after 554 days) followed by Davies Reef ( A. hyacinthus , 24% after 527 days) and Heron Reef ( A. hyacinthus : 13% and A. cf. kenti : 23% after 834 days). However, no single environmental variable, including nominal wave energy, bottom stress, flow velocity, sedimentation or benthic community composition consistently predicted survival, and effects weakened over time. Coral size and survival varied more at the device level than across sites, indicating the importance of fine-scale spatial and transient factors. These findings underscore the limitations of broad-scale environmental models to guide restoration and highlight the need for flexible, site-specific strategies. While seeding devices show promise as a scalable restoration tool, their success depends on matching species to suitable microhabitats and monitoring local conditions over time to support long-term outcomes.
Mitochondrial ABHD11 inhibition drives sterol metabolism to modulate T-cell effector function
Abstract α/β-hydrolase domain-containing protein 11 (ABHD11) is a mitochondrial hydrolase that maintains the catalytic function of α-ketoglutarate dehydrogenase (α-KGDH), and its expression in CD4 + T-cells has been linked to remission status in rheumatoid arthritis (RA). However, the importance of ABHD11 in regulating T-cell metabolism and function is yet to be explored. Here, we show that pharmacological inhibition of ABHD11 dampens cytokine production by human and mouse T-cells. Mechanistically, the anti-inflammatory effects of ABHD11 inhibition are attributed to increased 24,25-epoxycholesterol (24,25-EC) biosynthesis and subsequent liver X receptor (LXR) activation, which arise from a compromised TCA cycle. The impaired cytokine profile established by ABHD11 inhibition is extended to two patient cohorts of autoimmunity. Importantly, using murine models of accelerated type 1 diabetes (T1D), we show that targeting ABHD11 suppresses cytokine production in antigen-specific T-cells and delays the onset of diabetes in vivo in female mice. Collectively, our work provides pre-clinical evidence that ABHD11 is an encouraging drug target in T-cell-mediated inflammation.
Computational and pharmacophore-based study of Camellia sinensis phytochemicals targeting BRAF in melanoma
Structure of the human astrovirus capsid spike in complex with the neonatal Fc receptor
Abstract Human astroviruses (HAstVs) are a leading cause of viral gastroenteritis in children worldwide. Recently the neonatal Fc receptor (FcRn) was identified as a receptor for HAstV, however the molecular basis for the FcRn-HAstV interaction remained unclear. Here, we report the crystal structure of FcRn bound to the HAstV capsid spike domain at 3.4 angstroms resolution. We show that all classical HAstV spikes bind to FcRn and we identify three conserved HAstV spike residues that mediate binding to FcRn. Using competition binding assays, we show that the HAstV spike competes with IgG for binding to FcRn. Additionally, we demonstrate that the FcRn inhibitor, nipocalimab, and anti-HAstV neutralizing monoclonal antibodies block HAstV spike binding to FcRn, revealing their neutralization mechanisms and supporting their therapeutic potential. Overall, our findings illuminate a crucial interaction in the HAstV life cycle, which may help to inform the development of a HAstV vaccine and antibody therapies.
Nested named entity recognition in traditional Chinese medicine electronic medical records via dual-granularity feature augmentation and span classification
Genome-wide analysis of heart failure yields insights into disease heterogeneity and enables prognostic prediction in the Japanese population
Making smartglasses accessible: perspectives and prototypes from co-design with people with aphasia
Abstract Smartglasses are set to become a mainstream consumer technology in the near future. However, emerging technologies often overlook the accessibility needs of users with disabilities and older adults, increasing the risk of alienation and marginalization of these vulnerable communities. To address this tension, we conducted three co-design workshops with people living with aphasia (N=14) to examine their perspectives on smartglasses, envision potential applications, and identify anticipated barriers. Co-designers proposed and prototyped a variety of smartglass applications for both general and aphasia-specific support. However, co-designers also raised concerns about interaction difficulties and the socially conspicuous form-factor of current smartglass designs. Additionally, we evaluated smartglass functionalities using a mixed reality HoloLens head-mounted display (HMD). While participants expressed enthusiasm, qualitative and quantitative findings revealed mixed reactions. The standard hands-free interaction of the HoloLens was perceived as publicly awkward and deemed inaccessible for participants with vision impairments or post-stroke paralysis. Our findings highlight the need for more inclusive design practices to ensure emerging smartglasses empower all users.