Browse Articles

Prognostic RNA-splicing archetypes in breast cancer identified by extended pre-training of histopathology foundation models

Abstract Recently, histopathology foundation models (hFM) have rapidly advanced in size and complexity, achieving excellent performance in cancer diagnosis and biomarker discovery. Here, we specialise pre-trained hFMs to invasive tumour tissue and present three key contributions. First, we systematically evaluate the biological concepts encoded in hFM representations across multiple biological scales. Second, we demonstrate that informed extended pre-training transforms generalist models into tumour-specialised ones encoding richer semantic information, enabling discovery of recurrent tumour archetypes with consistent morphological and molecular identities across patients. Third, we identify dominant tumour archetypes with aberrant gene-expression programs coexisting within tumours and recurring across heterogeneous epithelial cancers, including HER2-positive and triple-negative breast cancer. Crucially, these archetypes exhibit prognostic value, with RNA splicing-associated archetypes consistently predicting poorer outcomes. Our work shows that tumour-specialised hFMs unlock rich molecular and morphological information from routine H&E slides, providing computationally efficient and biologically informed solutions for biological discovery and patient stratification.

Prevalence of undernutrition and its associated factors among pregnant women attending antenatal care service in public hospitals in Mogadishu, Somalia

Introduction Maternal undernutrition remains a major public health concern globally, particularly in Somalia, where conflict, poverty, and food insecurity exacerbate nutritional deficiencies. Undernutrition during pregnancy poses serious risks to both maternal and fetal health, including increased maternal mortality, low birth weight, and adverse developmental outcomes. Despite its critical implications, there is limited evidence on the burden and associated factors of maternal undernutrition in Somalia. This study assessed the prevalence and associated factors of undernutrition among pregnant women attending antenatal care (ANC) services in public hospitals in Mogadishu. Methods A facility-based cross-sectional study was conducted among 734 pregnant women attending antenatal care (ANC) services in public hospitals in Mogadishu. Participants were selected using a multistage sampling technique. Data were collected using interviewer-administered questionnaires and anthropometric measurements. Undernutrition was assessed using mid-upper arm circumference (MUAC < 23 cm), a measure commonly applied in clinical and humanitarian settings. Household food insecurity was assessed using the Household Food Insecurity Access Scale (HFIAS). Data were analyzed using SPSS version 26, and multivariable logistic regression was performed to identify factors associated with undernutrition at a significance level of p < 0.05. Results The prevalence of maternal undernutrition was 75.7% (95% CI: 72.6%–78.9%). Factors independently associated with undernutrition included maternal age 25–31 years (AOR = 2.61, 95% CI: 1.28–5.30) and 32–38 years (AOR = 2.18, 95% CI: 1.04–4.55), tertiary education (AOR = 3.43, 95% CI: 1.89–6.21), employee occupation (AOR = 6.38, 95% CI: 2.90–14.04), private business occupation (AOR = 9.79, 95% CI: 4.27–22.43), large household size (AOR = 1.96, 95% CI: 1.34–2.88), urban residence (AOR = 2.40, 95% CI: 1.39–4.15), household monthly income < USD 500 (AOR = 1.98, 95% CI: 1.33–2.94), lack of latrine facility (AOR = 2.00, 95% CI: 1.23–3.26), second (AOR = 8.59, 95% CI: 5.29–13.97) and third trimester (AOR = 6.91, 95% CI: 2.97–16.08), primigravidity (AOR = 2.29, 95% CI: 1.52–3.45), contraceptive use (AOR = 1.98, 95% CI: 1.34–2.90), substance use (AOR = 2.79, 95% CI: 1.93–4.04), and severe household food insecurity (AOR = 1.76, 95% CI: 1.01–3.08). However, some associations with large effect sizes should be interpreted with caution, as they may reflect residual confounding, small subgroup sizes, or model instability. Conclusion This study revealed a high prevalence of maternal undernutrition among pregnant women attending ANC services in Mogadishu, with multiple factors found to be associated with undernutrition. These findings highlight the need for targeted, multi-sectoral interventions to improve maternal nutrition through enhanced food security, health education, and access to nutrition-sensitive and nutrition-specific services. Targeted strategies such as supplementation programs, nutrition counseling, and improved water, sanitation, and hygiene (WASH) services are critical to mitigating maternal undernutrition and achieving Sustainable Development Goals related to maternal and child health.

Exosomal lncRNA-miRNA-mRNA regulatory axis may drive inflammation in polycystic ovary syndrome

TNF-α induces type I IFN signalling to suppress neurogenesis and recruit T cells

Abstract Adult hippocampal neurogenesis is essential for learning, memory, and mood regulation, and its disruption is implicated in ageing, neurodegeneration, and mood disorders. However, the mechanisms linking inflammation to adult hippocampal neurogenesis impairment remain unclear. Here, we identify chronic tumour necrosis factor-alpha signalling as a key driver of neurogenic dysregulation via a previously unrecognised type I interferon autocrine/paracrine loop in human hippocampal progenitor cells. Using a female-derived human in vitro neurogenesis model, single-cell RNA sequencing, and functional T cell migration assays, we show that tumour necrosis factor-alpha induces a robust type I interferon response in hippocampal progenitor cells, promoting chemokine-mediated and CXC motif chemokine receptor 3-dependent T cell recruitment and suppressing neurogenesis. This inflammatory signalling cascade drives a fate switch in hippocampal progenitor cells from a neurogenic trajectory towards an immune-defensive phenotype, with critical implications for infectious and inflammatory disease pathogenesis. These findings uncover a key inflammatory checkpoint regulating human adult hippocampal neurogenesis and highlight potential therapeutic targets to restore neurogenesis in chronic inflammatory states.

A piezoelectric electroporator (Piezopen) for enhanced “naked” RNA vaccine delivery

Despite the success of COVID-19 mRNA vaccines, they still face challenges with high costs, complex manufacturing, off-target biodistribution, and systemic reactogenicity stemming from their inflammatory carriers: lipid nanoparticles (LNPs). While “naked” RNA delivery could in principle solve these issues, studies have suggested that it is infeasible due to rapid degradation by RNases and poor cellular entry, thereby necessitating formulations that enhance intracellular delivery and RNA stability. Now, we challenge this paradigm by showing that a simple and inexpensive (<$1), lighter-derived electroporator with microneedle electrodes (Piezopen) can augment gene expression and immunogenicity to naked mRNA leading to comparable responses to LNPs at low doses. We achieve robust responses in the absence of systemic inflammation and reactogenicity using skin-targeted delivery, administer diverse construct types (i.e., mRNA, self-amplifying RNA (saRNA), circular RNA (circRNA)), and demonstrate cross-species validation in live human skin to derisk subsequent clinical application. Our results introduce Piezopen as an inexpensive, well-tolerated, and efficacious alternative to LNPs for mRNA vaccine delivery, designed to facilitate routine vaccinations and pandemic response.

The Effects of Haze Pollution on County-level Entrepreneurship Activity in China

Spatiotemporal dynamics of scattering exceptional points

Abstract Exceptional points (EPs) offer new routes to control wave-matter interactions in non-Hermitian systems. Whereas conventional EPs arise in Hamiltonians and mark eigenstate coalescence, scattering EPs recast them in input-output responses for free-space wave control. Yet the rich topology of EPs is largely concealed in steady-state observables, exhibiting similar asymmetric patterns governed by identical Jordan-block forms. Here we exploit scattering as an evolving process in space and time to resolve spatiotemporal dynamics. We establish a cross-domain mapping from momentum-frequency to space-time, connecting static EPs to dynamic scattering singularities. By projecting a unitary scattering matrix onto lower-dimensional subspaces, EP topology is imprinted into a wave packet as a spatiotemporal vortex (STV), whose phase singularity is anchored to the EP and carries its quantized topological charge. In a liquid-surface-wave platform, we observe a charge-1 STV and a disintegrated charge-2 STV, generated by distinct EPs in a single device. Our work links fixed exceptional topology to dynamic wave topology, establishing a framework for spatiotemporal wave control.

Development of a COVID-19 Vaccination Anxiety Scale to measure COVID-19 vaccine anxiety in Japanese adults

Although COVID-19 vaccines are widely available and effective, vaccination anxiety remains a significant public health challenge that is distinct from behavioral vaccine hesitancy. While various scales measure hesitancy, few specifically capture the multifaceted psychological anxiety regarding COVID-19 vaccines, and none have been validated for the Japanese population. To directly quantify this affective component, this study aimed to develop and validate the COVID-19 Vaccination Anxiety Scale for Japanese adults. A two-wave web-based survey was conducted with 500 Japanese adults. Participants completed the newly culturally adapted 25-item scale alongside validation measures assessing fear of COVID-19, vaccination readiness, conspiracist beliefs, and perceived vulnerability to disease. Test–retest reliability was evaluated after a four-week interval. Factor analyses revealed a three-factor structure: Anxiety Related to Vaccine Confidence, Emotional Symptoms of Anxiety, and Anxiety about Infection after Vaccination. The scale showed good internal consistency and high test–retest reliability, and multigroup analyses suggested approximate comparability of the three-factor structure across gender. Construct validity was supported through significant, theoretically aligned correlations with the external criteria. Specifically, vaccination anxiety was strongly associated with lower vaccination readiness, higher fear of infection, and greater conspiratorial beliefs. These findings highlight that vaccination anxiety is a complex emotional response that requires specific measurement. From a public health perspective, the results suggest that addressing affective concerns—rather than solely providing analytic risk data—is crucial for effective communication. The COVID-19 Vaccination Anxiety Scale provides a useful tool for monitoring population-level anxiety and evaluating targeted interventions in Japan. Future studies should examine whether the same structure is replicated in younger and more age-balanced samples and whether the scale can be adapted for use in other settings.

The interaction network of health-related quality of life in elderly people living with HIV: a cross-sectional study in Chongqing, China

Red electroluminescent Cu4I4 bipyramids with external quantum efficiency beyond 40%

Abstract The development of red cluster light-emitting devices (CLED) lags far behind blue and green congeners with respect to efficiencies, since multiple excited states of red cluster molecules are involved in non-radiation during electroluminescence. Herein, we accurately optimize excited states of a red bipyramidal [PXZDPPQ] 2 Cu 4 I 4 , whose ligand integrates conjugation-extended 2-diphenylphosphineylquinoline (DPPQ) and strong electron-donating phenoxazine (PXZ), giving rise to desired excited state locations of the outer intraligand charge transfer ( n LCT)-featured first singlet (S 1 ) and triplet (T 1 ) excited states and the inner high-lying metal-ligand charge transfer ( n MLCT) states. Its outer and highly radiative n LCT states are spatially and energetically advantageous in carrier capture and exciton confinement; meanwhile, its inner and high-lying n MLCT states are protected from collisional quenching, and support cross transitions between n LCT states to realize exactly balanced dual emissions with a ratio of 51/49. [PXZDPPQ] 2 Cu 4 I 4 achieves eightfold increased photoluminescence quantum yield of 93.6%, tenfold increased external quantum efficiency reaching 43.7% as a new record for all kinds of planar red light-emitting devices, and 20% improved exciton utilization efficiency in comparison to the congener with the reverse excited state location. These results demonstrate the unique merit of cluster materials in exciton engineering and their potential for next-generation full-color displays.

Research on fracture mechanism of rock mass with orthogonal cracks under uniaxial compression

The initiation, expansion and penetration of cracks in rock masses are important reasons for rock mass failure. Orthogonal crack is the most typical form of cracks. There is still a lack of in-depth and detailed research on the law of expansion and penetration of orthogonal cracks. In this paper, the RFPA software is used to study the failure process of the rock mass with orthogonal fractures, and the influence of the change of the angle α between the main fracture and the loading direction on the failure mode and the failure mechanical properties of the specimen is comprehensively explored. The simulation results show that the angle α between the different main fractures and the loading direction affects the internal stress state of the rock, redistributes the stress field at the tip of the fracture and the surrounding area, resulting in the difference in the energy required for the expansion of the new fracture, which is manifested in the macroscopically as the difference in the strength of the rock and acoustic emission quantity.

Curvature-weighted spectra anticipate dissipation peaks in decaying three-dimensional turbulence

Modern arc-like water content in the source of 3.1-billion-year-old volcanic rocks

Abstract Whether Archean arc-like volcanism reflects subduction remains debated. We present high-resolution geochemical data from a well-preserved 3.13-3.10 Ga arc-like volcanic succession in Australia’s Pilbara Craton, a rare Archean analog of modern arc volcanism retaining fluid-mobile element concentrations consistent with primary magmatic values. The sequence records three primitive lava series typical of modern arcs: tholeiitic, calc-alkaline, and the oldest stratigraphically extensive genuine boninites. Geochemical modelling shows this melt diversity requires at least two mantle sources with distinct depletion histories. The mantle H 2 O required for fluid-assisted melting to produce these lavas substantially exceeds primitive mantle, approaching the H 2 O-saturated solidus of modern mantle wedges. We infer hydrous melting was triggered by dripduction, the short-lived inclined foundering of hydrated lithosphere without laterally continuous plate boundaries, in an off-plateau setting. Dripduction locally recycled surface water and generated arc-like magmas without self-sustained plate tectonics, possibly promoting mantle-ocean-atmosphere volatile exchange during the Archean.

Epidemiological characteristics of amebiasis in Japan from 2001 to 2022

Amebiasis cases in Japan are reported to the government according to the Infectious Diseases Control Law. Previous studies have shown significant reductions in total case numbers after 2018 and during the COVID-19 pandemic. This study aimed to clarify the recent trends of amebiasis cases in Japan, including during the pandemic period, with details on places of infection, using government surveillance data from January 1, 2001, to December 31, 2022. Change of time trends were modeled through piecewise mixed-effect regression model with knots set at 2018 and 2020. Year-to-year differences in case numbers were statistically assessed using Poisson regression model. And descriptive analyses of amebiasis cases reported in Japan by sex, age class, and prefecture (from 2001 to 2022) were conducted. Piecewise time trends of male-domestic cases showed increasing trend by 59.5 cases per year (p < 0.0001) before 2008. The trend slowed but still increased during 2008–2018, showing annual increase of 15.2 cases per year (p = 0.0014). A sharp decline occurred during 2018–2020, with cases decreasing by 219.0 cases per year (p < 0.0001). After 2020, the trend did not show statistically significant change (−16.9 cases per year, p = 0.6532). Poisson regression confirmed significant reductions in total and domestic cases between 2017–2018 and 2019–2020, while imported cases declined significantly only between 2019 and 2020. Male cases predominated, with most male cases in their 40s and 50s. Most cases of amebiasis have been reported in metropolitan areas. These results suggest that the decreased case numbers during the COVID-19 pandemic were due to not only the travel ban, but less socioeconomic activity. Furthermore, the epidemiology of amebiasis is similar to that of HIV infection in Japan, but the case numbers of amebiasis have not yet increased through 2022, showing a different trend from HIV infection and syphilis, the reason of which is unclear and needs further investigation.

Research on adaptive collaborative dispatch optimization algorithms for drones in distribution networks

Rapid GC content evolution in rice through GC-biased gene conversion and selection for translation efficiency

Effects of 5D built environment and non-built-environment factors on injury crash risk: An interpretable machine learning analysis

To identify the key determinants of traffic injury risk and clarify the relative roles of built environment factors and crash-context factors, this study develops a traffic injury risk identification framework that integrates 5D built environment variables with non-5D factors, using traffic crash data collected in Changsha, Hunan Province, China, from 2017 to 2019. After data cleaning, screening, and spatial matching, a total of 9,743 valid samples were obtained, and injury occurrence in a crash was defined as a binary dependent variable. On this basis, three feature sets were constructed, including a 5D feature set, a non-5D feature set, and a combined 5D + non-5D feature set. Random Forest (RF), eXtreme Gradient Boosting (XGBoost), and Categorical Boosting (CatBoost) models were then developed and compared, and the best-performing model was further interpreted using the Shapley Additive Explanations (SHAP) method. The results showed that the combined 5D + non-5D feature set consistently outperformed the models using either the 5D or non-5D feature set alone, indicating that traffic injury risk arises from the joint influence of the built environment and immediate crash-context conditions. Among the three models, CatBoost achieved the best performance under the combined feature set and produced the highest receiver operating characteristic-area under the curve (ROC-AUC) value, demonstrating superior overall discriminative ability. The SHAP results further revealed that, within the combined CatBoost model, 5D variables accounted for a larger share of the model-based contribution than non-5D variables. However, given the relatively small improvement in accuracy after adding 5D variables, this finding should be interpreted as evidence of complementary explanatory information rather than as a dominant source of predictive performance. Distance to the nearest metro station, lighting condition, road network density, point of interest (POI) mix, and distance to the nearest bus stop were identified as the most influential factors. Further dependence analysis showed that a greater distance to the nearest metro station generally increased traffic injury risk, whereas higher road network density and greater POI mix were generally associated with lower injury risk. From the perspective of the interaction between 5D built environment characteristics and non-5D factors, this study reveals the multidimensional pathways through which traffic injury risk is shaped. The findings also confirm the effectiveness of the CatBoost–SHAP framework for traffic injury risk identification and interpretation, and provide empirical support for urban traffic safety risk assessment, road environment optimization, and refined governance strategies.

A comparative study of baked and traditional fermented milk: physicochemical characteristics and biological effects in wistar albino rats

Abstract Baked fermented milk (BFM) is produced by heating milk at high temperatures before fermentation, which promotes Maillard reactions. These reactions enhance sensory and functional properties and generate bioactive compounds; however, they may also lead to the formation of potentially harmful products. This study aimed to evaluate the physicochemical properties and biological effects of BFM, with and without probiotic supplementation, compared with traditional fermented milk (TFM). Cow’s milk was either heated at 90 °C for 5 min or baked at 115 °C for 20 min before inoculation with starter cultures. Half of each preparation was supplemented with probiotics, yielding four formulations: TFM, probiotic TFM (PTFM), BFM, and probiotic BFM (PBFM). Thirty-five male Wistar albino rats were randomly assigned to five groups and fed experimental diets for 45 days, including a control group receiving liquid milk and four groups receiving fermented milk formulations. Physicochemical analysis showed that BFM had higher 5-HMF content and stronger antioxidant activity than TFM, along with slightly higher diacetyl levels but lower apparent viscosity. Biologically, BFM reduced serum glucose levels and oxidative stress markers without adversely affecting body weight gain (BWG), serum lipid profiles, or protein levels. However, rats fed BFM exhibited the highest serum ASAT activity and urea levels. Probiotic supplementation with Bifidobacterium bifidum NRRL B-41,410 and Lacticaseibacillus rhamnosus NRRL B-442 provided synergistic benefits in the PBFM group, including further reductions in glucose levels, improved lipid profiles, decreased oxidative stress, lower BWG, and mitigation of elevated serum urea and ASAT activity. Overall, although daily consumption of BFM may provide health benefits, it may also induce mild metabolic alterations, which are effectively alleviated through probiotic supplementation.

Assessing corporate sustainability with large language models: evidence from Europe

Abstract Companies play a crucial role in achieving global sustainability goals, yet evidence on their progress across environmental, social, and governance (ESG) dimensions remains limited. We develop a machine learning framework to systematically extract ESG indicators from corporate reports. Applying this approach to annual and sustainability reports of 600 large European firms (2014–2023), we construct a dataset of 2.9 million ESG observations across environmental, social, and governance topics. We assess ESG transparency based on disclosures aligned with the European Sustainability Reporting Standards (ESRS) and evaluate ESG performance using extracted numerical indicators. Results reveal a pronounced transparency gap: firms in the top ESG rating decile disclose 22% more indicators than those in the bottom decile, although this gap narrows over time. Performance trends are uneven: while most social indicators remain largely stagnant, except for gains in gender equality, environmental indicators show some improvement. Reported scope 3 emissions increase sharply, largely reflecting improved disclosure. Our open-source framework enables systematic tracking of corporate ESG efforts.

SST subseasonal-to-seasonal forecasting: A heterogeneous three-path fusion model with wavelet decomposition

El Niño-Southern Oscillation (ENSO), the Earth’s dominant mode of interannual climate variability, is closely linked to sea surface temperature (SST) variability in the central and eastern equatorial Pacific. Accurate subseasonal-to-seasonal SST forecasting in this region supports ocean monitoring and environmental assessment. We propose WDFormer (Wave Decomposition Former), a daily SST forecasting model built on a “decomposition, multi-path processing, and adaptive fusion” paradigm. The model uses 365 daily SST observations from local 4 × 4 grids around representative ENSO monitoring stations to predict daily SST values over 7–90-day horizons. WDFormer first applies wavelet decomposition to separate the input sequence into a smoother low-frequency component and a higher-frequency residual. Lightweight MLP-based branches process the original and low-frequency components, while a Transformer-based branch models the residual dynamics. Outputs are integrated through a learnable gating mechanism. Experiments on daily OISST data from five representative ENSO subregions show that WDFormer outperforms several deep learning baselines at most 30–90-day horizons. Comparisons with persistence and daily climatology confirm that persistence dominates at very short lead times, whereas WDFormer provides clearer advantages as the horizon extends. An ONI-based ENSO-phase and El Niño-stage stratified evaluation shows that WDFormer achieves the lowest RMSE under El Niño, La Niña, and Neutral conditions, as well as during El Niño onset/development, mature/peak, and decay/termination stages for the 90-day task. Bootstrap confidence intervals and paired RMSE comparisons further support the robustness of the 90-day results. These results support decomposition-based heterogeneous modeling for daily SST forecasting at 30–90-day horizons and motivate broader extensions to probabilistic and multi-variable forecasting.