PNAS nexus最新文献

Digital tracking technologies have transformed and enhanced online data collection across industries. Their integration into healthcare systems, however, raises urgent concerns about patient privacy and security. This study provides the first large-scale empirical analysis of pixel tracking technologies on US hospital websites and their unintended consequences for data breaches. Using historical website data from the Wayback Machine (2012-2023), we find that 66% of the sample employed pixel tracking, despite stringent privacy regulations. Our results reveal that third-party pixel use significantly increases data breach risk, underscoring a previously undocumented cybersecurity vulnerability. These findings highlight a critical regulatory gap in healthcare privacy, as tracking pixels operate outside the traditional scope of Health Insurance Portability and Accountability Act protections. As hospitals increasingly rely on digital tools, our study calls for reevaluating privacy and data security safeguards and regulatory oversight to address the emerging risks of modern tracking technologies.

Automated detection of building destruction in conflict zones is crucial for human rights monitoring, humanitarian response, and academic research. However, existing approaches (i) rely on proprietary satellite imagery, both expensive and of limited availability at wartime, (ii) require labeled training data, usually not available in war-affected regions, or (iii) use optical imagery, regularly obstructed by cloud cover. This study addresses these challenges by introducing an unsupervised method to detect destruction at the building level using freely and globally available Sentinel-1 synthetic aperture radar images from the European Space Agency. By statistically assessing interferometric coherence changes over time, unlike existing approaches, our method enables the detection of destruction from a single satellite image, allowing for near real-time destruction assessments every 12 days. We provide a continuous, statistically grounded probability measure for the likelihood of destruction at both the building and pixel level, thereby quantifying the level of uncertainty of the detection. Using ground truth data and reported sequences of events, we validate our approach both quantitatively and qualitatively, across three case studies in Beirut, Mariupol, and Gaza, demonstrating its ability to accurately identify the spatial patterns and timing of destruction events. Using open-access data, our method offers a scalable, global, and cost-effective solution for monitoring building destruction in conflict zones.

[This corrects the article DOI: 10.1093/pnasnexus/pgae313.].

Long QT syndrome type 3 (LQTS3), caused by gain-of-function mutations in the SCN5A gene, encompasses a spectrum of clinical presentations, ranging from asymptomatic carriers to severe arrhythmic phenotypes, representing the "silent killer" and "brutal killer" dichotomy. The p.I239V mutation is associated with mild symptoms and minimal arrhythmic events, whereas the newly identified p.M1487K mutation is linked to life-threatening arrhythmic storms. This study aimed to explore the electrophysiological properties of these mutations and their responses to sodium channel blockers to advance precision medicine in LQTS3 management. Genetic analysis identified rare SCN5A mutations in two LQTS3 patients. Site-directed mutagenesis was used to construct mutant SCN5A plasmids, which were expressed in HEK293 cells. Electrophysiological properties were analyzed using patch-clamp techniques, and pharmacological responses to flecainide, mexiletine, and ranolazine were evaluated. Electrophysiological recordings correlated with clinical presentations. Both mutations showed increased window I_Na and faster recovery from inactivation. The p.I239V mutation lacked sustained I_Na, while p.M1487K exhibited significantly increased sustained I_Na (2.3 ± 2.15%, P < 0.0001). Mexiletine and ranolazine effectively reduced sustained I_Na by 76.15 ± 5.83, and 77.63 ± 9.41%, respectively, outperforming flecainide, aligning with clinical responses. This study highlights the role of sustained I_Na in LQTS3 severity and underscores the importance of mutation-specific treatments. By tailoring treatments to the electrophysiological characteristics of each mutation, precision medicine offers a promising approach to improving patient outcomes in LQTS3.

Spider silk proteins (spidroins) are large, block-copolymer-like proteins that must remain soluble at high concentration in the spinning dope while being primed for rapid fiber formation. Understanding how these intrinsically disordered proteins organize in solution is key to explaining the transformation from soluble dope to solid fibers with exceptional strength and toughness. Here, we show that major ampullate (Ma) spidroins from the black widow spider form dynamic ensembles that include metastable tubular substructures. Multiscale molecular dynamics (MD) simulations reveal compact, anisotropic monomers with tubular geometries ∼3-4 nm in diameter and 50 nm in contour length. Small-angle X-ray scattering (SAXS) ensemble fitting confirms that a minority population of tubular conformers is required to reproduce experimental scattering profiles. Complementary atomistic MD and solution NMR chemical shift and relaxation analyses show that these tubular conformers are enriched in β-turn and bend motifs, maintaining local flexibility while promoting overall compaction. Mutational simulations further demonstrate that alternating poly(Ala) and Gly-Gly-X sequence patterning drives amphiphilic packing that stabilizes the tubular morphology. Together, these findings reveal that spider silk proteins form dynamic, disordered ensembles with sequence-encoded tubular substructure. This model reconciles SAXS and NMR observations and provides a mechanistic framework for how amphiphilic patterning, metastability, and disorder collectively enable spider silk proteins to remain soluble yet preorganized for hierarchical self-assembly into one of nature's toughest materials.

Energy efficiency is a major driving force in the evolution of organisms, and previous research implies that humans may have evolved pleasure-based signals to guide optimal actions. But could this energy-saving heuristic also apply to aesthetic pleasure? We test this hypothesis using both an in silico model of the visual system (VGG19) and human observers, finding strong evidence in both. First, we measure the proxy for metabolic cost incurred by VGG19-either pretrained for object and scene categorization or randomly initialized-as it processes 4,914 images of objects and scenes, revealing an inverse relationship between aesthetic preferences and metabolic cost, and only in the pretrained model. Next, we compare aesthetic ratings of visual stimuli to metabolic activity in the human visual system, measured via the blood oxygen level-dependent signal during functional magnetic resonance imaging. We observe the same inverse relationship between blood oxygen level dependent signals and aesthetic preferences in both early visual regions (V1, V2, and V4) and higher-level regions (fusiform face area, occipital place area, and parahippocampal place area). These findings suggest that aesthetic preferences may at least partially arise from an affective heuristic favoring low-energy states, and they offer a unified framework linking empirical evidence on visual discomfort with theories of processing fluency, image complexity, and prototypicality, providing a straightforward model for understanding aesthetic judgments.

Social networks shape how humans form opinions, exchange information, and organize collectively. As large language models (LLMs) become embedded in social and professional environments, it is critical to understand whether their interactions resemble human network dynamics. We introduce a framework to study the network formation behaviors of multiple LLM agents and benchmark them against human decisions. Across synthetic and real-world settings, including friendship, telecommunication, and employment networks, LLMs reproduce core microlevel principles (preferential attachment, triadic closure, and homophily), and macrolevel properties (community structure, small-world effects). Their emphasis on these principles adapts to context: for example, LLMs favor homophily in friendship networks but heterophily in organizational settings, mirroring patterns of social mobility. A controlled survey shows strong alignment between LLM and human link-formation decisions. These results highlight LLMs' potential as tools for social simulation and synthetic data generation, while underscoring risks of bias and fairness in AI systems that interact with human networks.

Concentrations of nitrogen dioxide (NO₂) measured outdoors using satellites and ground-level stations across the United States are regularly used to estimate NO₂ exposures and disease burdens. In contrast, exposures attributable to sources of NO₂ indoors are neither systematically monitored nor estimated. Here, to our knowledge we produce the first nationwide, ZIP-code-level estimate of total residential NO₂ exposure that integrates both outdoor sources and the primary indoor source-gas- or propane-burning stoves. To estimate exposure by ZIP code, we combine our measurements of indoor NO₂ emissions and concentrations in more than 15 cities across seven regions with outdoor NO₂ concentrations and comprehensive housing stock data for 133 million residential dwellings and statistical samplings of occupant behavior. We estimate average total residential long-term NO₂ exposure across the United States to be ∼10 ppbv for people who own a gas stove (∼18 ppbv or more for households in the top 5% of gas burned while cooking) and ∼8 ppbv from total outdoor exposure for those with electric stoves, which cause no additional NO₂ exposure. Across the United States, the NO₂ exposure of ∼22 million people would fall below the World Health Organization (WHO)'s long-term exposure guideline (10 µg/m³ or 5.2 ppbv) if they reduced or stopped cooking with gas or propane. Gas and propane stoves are also responsible for virtually all (>99%) of the residential exceedances of the WHO's 1-h-averaged air quality guideline across the United States. Gas and propane stoves are a substantial source of residential NO₂ exposure even when compared with all outdoor sources combined.

Autoimmune mechanisms are associated with both congestive heart failure (CHF) and atrial fibrillation (AF). Optic neuritis (ON) is known to elevate the risk of systemic autoimmune disorders. However, it remains uncertain whether ON serves as a risk factor for CHF and AF. This study aimed to investigate the association between ON and the risk of CHF and AF in a nationwide, population-based cohort. The research utilized data from Korea's National Health Insurance Service, analyzing 15,587 patients newly diagnosed with ON and 77,935 age- and sex-matched controls from 2010 to 2017. Factors of demographics, medical history, lifestyle, and lab results were considered. CHF and AF incidences were identified through ICD-10 codes and analyzed using Cox regression models adjusted for confounders. During the 4-year follow-up, CHF and AF were diagnosed in 3.39 and 0.86% of participants, respectively. ON patients showed higher risks of CHF [hazard ratio (HR) = 1.341] and AF (HR = 1.215) after adjusting for potential confounders. Notably, stronger associations for CHF risk were found in patients younger than 50 years and those in the lowest income quartile. The findings provide compelling evidence of an independent association between ON and increased risks of CHF and AF, especially in younger individuals, suggesting a role of autoimmune processes in ON under these cardiovascular conditions. The study highlights the need for early cardiac evaluation in ON patients and suggests that timely interventions could improve prognosis. Further research is necessary to understand the pathophysiological links between ON and cardiovascular disorders.

Natural and human-made common goods present key challenges due to their susceptibility to degradation, overuse, and congestion. We explore the self-organization of their usage when individuals have access to several available commons but limited information on them. We propose an extension of the Win-Stay, Lose-Shift (WSLS) strategy for such systems, in which individuals use a resource repeatedly until they are unsuccessful and then shift randomly. This simple and completely decentralized strategy promotes balanced resource use based solely on individual experience, as it does not require communication between individuals nor a governing institution to coordinate usage. Selective individuals who retain information on their usage and accordingly adapt their tolerance to failure in each common good improve the experienced quality for all individuals in the population. Even in hybrid populations of adaptive and nonadaptive individuals, the WSLS strategy allows self-organization into an ideal distribution with equal experienced quality across common goods. We apply this strategy to the server selection problem faced by mobile users when accessing Internet services. Realistic simulations demonstrate its success, scalability, and robustness to dynamic system conditions. Furthermore, the WSLS strategy can be used to understand animal dispersal on grazing and foraging land, to propose solutions to operators of systems of public transport and other technological commons, and to address problems of common good usage in social systems through decentralized governance rather than control-oriented policies.