Explore the vast range of titles available.

Northern China harbored the world’s earliest complex societies based on millet farming, in two major centers in the Yellow (YR) and West Liao (WLR) River basins. Until now, their genetic histories have remained largely unknown. Here we present 55 ancient genomes dating to 7500-1700 BP from the YR, WLR, and Amur River (AR) regions. Contrary to the genetic stability in the AR, the YR and WLR genetic profiles substantially changed over time. The YR populations show a monotonic increase over time in their genetic affinity with present-day southern Chinese and Southeast Asians. In the WLR, intensification of farming in the Late Neolithic is correlated with increased YR affinity while the inclusion of a pastoral economy in the Bronze Age was correlated with increased AR affinity. Our results suggest a link between changes in subsistence strategy and human migration, and fuel the debate about archaeolinguistic signatures of past human migration. Northern China contains some of the world’s earliest farming societies. Here, authors use 55 ancient genomes to trace the genetic history of human migrations across northern China for the last 7500 years, and document genetic changes mirroring shifts in subsistence strategy.

Large language models (LLMs) have exploded a new heatwave of AI for their ability to engage end-users in human-level conversations with detailed and articulate answers across many knowledge domains. In response to their fast adoption in many industrial applications, this survey concerns their safety and trustworthiness. First, we review known vulnerabilities and limitations of the LLMs, categorising them into inherent issues, attacks, and unintended bugs. Then, we consider if and how the Verification and Validation (V&V) techniques, which have been widely developed for traditional software and deep learning models such as convolutional neural networks as independent processes to check the alignment of their implementations against the specifications, can be integrated and further extended throughout the lifecycle of the LLMs to provide rigorous analysis to the safety and trustworthiness of LLMs and their applications. Specifically, we consider four complementary techniques: falsification and evaluation, verification, runtime monitoring, and regulations and ethical use. In total, 370+ references are considered to support the quick understanding of the safety and trustworthiness issues from the perspective of V&V. While intensive research has been conducted to identify the safety and trustworthiness issues, rigorous yet practical methods are called for to ensure the alignment of LLMs with safety and trustworthiness requirements.

Real-time high-precision image segmentation on mobile and edge devices remains challenging due to the limited ability of traditional convolutional networks to model long-range dependencies and their high computational cost at high resolutions. We propose PUNet, a lightweight parallel U-Net variant that integrates depthwise separable convolutions (DSConv) with a structured state-space Mamba module in a dual-path encoder. The core component, the parallel structured state-space encoder, employs two branches to efficiently capture local spatial features (via DSConv) and model global semantic dependencies (via the visual Mamba layer). At the same time, a squeeze-and-excitation skip connection adaptively fuses multi-scale features. With only 0.26 M parameters and linear computational complexity, PUNet enables real-time inference on resource-constrained platforms. Experiments on the CamVid and CRACK500 datasets demonstrate superior performance, achieving validation Dice scores of 0.9208 and 0.7902, and mean Intersection-over-Union of 0.8643 and 0.6612, respectively, significantly outperforming other lightweight models.

Accurate diagnosis of bearing faults is crucial for ensuring the safe and reliable operation of rotating machinery. To enhance the recognition accuracy of rolling bearings under nonlinear and non-stationary vibration conditions, this study proposes an integrated approach combining a multi-stage signal preprocessing strategy, termed EMDWS (Empirical Mode Decomposition with Wavelet denoising and SMOTE), with a hybrid deep learning architecture that integrates a Convolutional Neural Network (CNN) and a Transformer model, hereinafter referred to as CNT (CNN-Transformer). The method first applies empirical mode decomposition (EMD) in conjunction with wavelet denoising to enhance the representation of non-stationary fault features. Subsequently, the synthetic minority oversampling technique (SMOTE) is employed to address the issue of class imbalance in the dataset. A hybrid CNN-Transformer model is constructed by integrating convolutional neural networks and Transformer modules, enabling the extraction of both local and global signal characteristics. Furthermore, Bayesian optimization is applied to fine-tune the model’s hyperparameters, thereby enhancing both the efficiency and robustness of the model. Experimental results demonstrate that the proposed method achieves a high identification accuracy of 99.83%, indicating its effectiveness in distinguishing various bearing fault types.

Irrigation pipelines are critical agricultural hydraulic facilities that often develop minor bending defects due to ground settlement or improper installation. This study employs Lagrange equations for non-material volumes and the Absolute Nodal Coordinate Formulation (ANCF) to model the multi-span viscoelastic micro-bending irrigation pipelines, investigating the influence of micro-bending defects on critical flow velocity. The material parameters of the pipeline wall are determined via uniaxial tensile tests, and the effectiveness of the proposed model is validated through comparison with degraded models and field tests. Further numerical analysis demonstrates that modifying the micro-bend defect of the pipeline from a parabolic to a sinusoidal shape yields a 13.9% enhancement in critical flow velocity. This improvement is particularly significant for irrigation projects with limited pipe material options, tight flow design margins, and low economic returns.

China, with the third largest share of global tuberculosis cases, faces a substantial challenge in its healthcare system as a result of the high burden of multidrug-resistant and rifampicin-resistant tuberculosis (MDR/RR-TB). This study employs a genomic epidemiological approach to assess recent tuberculosis transmissions between individuals, identifying potential risk factors and discerning the role of transmitted resistant isolates in the emergence of drug-resistant tuberculosis in China. We conducted a population-based retrospective study on 5052 (MTB) isolates from 70 surveillance sites using whole genome sequencing (WGS). Minimum spanning tree analysis identified resistance mutations, while epidemiological data analysis pinpointed transmission risk factors. Of the 5052 isolates, 23% (1160) formed 452 genomic clusters, with 85.6% (387) of the transmissions occurring within the same counties. Individuals with younger age, larger family size, new cases, smear positive, and MDR/RR were at higher odds for recent transmission, while higher education (university and above) and occupation as a non-physical workers emerged as protective factors. At least 61.4% (251/409) of MDR/RR-TB were likely a result of recent transmission of MDR/RR isolates, with previous treatment (crude OR =   2.77), smear-positive (cOR =   2.07) and larger family population (cOR = 1.13) established as risk factors. Our findings highlight that local transmission remains the predominant form of TB transmission in China. Correspondingly, drug-resistant tuberculosis is primarily driven by the transmission of resistant tuberculosis isolates. Targeted interventions for high-risk populations to interrupt transmission within the country will likely provide an opportunity to reduce the prevalence of both tuberculosis and drug-resistant tuberculosis.

Background: Joint inflammation is significantly connected with progressive joint deterioration, potentially increasing the incidence of persistent major clinical challenges and global disability. Nutrient-based preventive strategies have been explored to investigate the interventive efficacy of the proposed prescribed formula for joint inflammation. However, the synergistic ameliorative effects of the nutritional formula should be evaluated to investigate its impact on joint inflammation. Methods: A prescribed formula including turmeric (T), N-acetylglucosamine (G), enzymatically hydrolyzed bone powder (E), and undenatured type II collagen (U) was comprehensively evaluated for its synergistic effects on joint inflammation and the underlying mechanisms. A rat model established using the Hulth method was used to evaluate the interventive effects in vivo. Moreover, in vitro analysis using the murine chondrogenic cell line ATDC5 was performed to validate the intervention and its mechanism of action. Results: The prescribed formula was shown to synergistically reduce levels of inflammation-related cytokines, reduce oxidative stress, and enhance bone metabolism to promote joint regeneration. Micro-Computed Tomography (Micro-CT) analysis revealed restoration of joint architecture and ameliorated physiological status upon formula intervention. In vitro analysis further validated the synergistic alleviation of inflammation and oxidation, as well as reductions in MMP13 and CTX-1 levels, which implies that modulating bone metabolism alleviates the deterioration and inflammation of joint architecture. Conclusions: The synergistic formula in this study achieves synchronous modulation of several core pathological pathways, yielding synergistic modulation of joint inflammation. Nutrient-based interventions or preventive strategies show promising effects against joint inflammation and progressive mechanistic deterioration.

Tender coconut water has been very popular as a natural beverage rich in various electrolytes, amino acids, and vitamins, and hence a large amount of tender coconut kernel is left without efficient utilization. To explore the possibility of making infused tender coconut kernel, we investigated the effects of two osmosis methods, including solid-state osmotic dehydration and liquid-state osmotic dehydration, as well as two osmosis agents such as sorbitol and sucrose, on the mass transfer of coconut kernel under solid-state osmotic dehydration conditions. The results showed that under the conditions of solid-state osmosis using sucrose and liquid-state osmosis using sucrose solution, the water diffusion coefficients were 9.0396 h−1/2 and 2.9940 h−1/2, respectively, with corresponding water mass transfer coefficients of 0.3373 and 0.2452, and the equilibrium water loss rates of 49.04% and 17.31%, respectively, indicating that the mass transfer efficiency of solid-state osmotic dehydration of tender coconut kernel was significantly higher than that of liquid-state osmotic dehydration. Under solid osmosis conditions, the water loss rates using sucrose and sorbitol were 38.64% and 41.95%, respectively, with dry basis yield increments of 61.38% and 71.09%, respectively, demonstrating superior dehydration efficiency of sorbitol over sucrose under solid-state osmosis. This study can provide a reference for the theoretical study of the mass transfer of tender coconut kernel through osmotic dehydration, and also provide technical support for the development and utilization of tender coconut kernel.

Background Tropomyosin (TM) is a major and highly stable allergen in crustaceans such as shrimp. The high stability of TM during food processing or gastrointestinal digestion impels the substantial allergenic risk to sensitive individuals, which strengthen the urgent requirement of rapid and sensitive detection to guarantee the avoidance of relevant food allergens. Results In this study, we report the development of a nanobody (Nb)-based “colorimetric-photothermal” dual-readout immunosensing system for ultrasensitive TM detection based on the utilization of a nanozyme of manganese-gold nanoflowers (MnAuNPs). TM-specific Nbs with high affinity and good specificity were isolated from an immune library using phage display technology. To enhance the detection signal, MnAuNPs with a core-shell structure were prepared by a hydrothermal method. The Mn-rich shell and gold core confer intrinsic oxidase-like activity and strong near-infrared absorption capacity to MnAuNPs, enabling simultaneous catalysis of 3,3’,5,5’-tetramethylbenzidine oxidation and photothermal conversion of the oxidized product. By decorating TM-specific Nbs with MnAuNPs, a dual-mode immunoassay combining enzymatic and photothermal signal amplification was established. The colorimetric detection limit was 6.17 ng/mL, while the sensitivity of the photothermal mode was further enhanced to 0.25 ng/mL, approximately 48 times higher than that of the conventional ELISA. The immunosensing system exhibited high specificity, reproducibility, and accuracy in real food matrices. The recovery rates of the dual-signal immunosensing system ranged from 79.1% to 108.8%, with a coefficient of variation of 0.6% to 12.3%. Conclusions We have generated a highly sensitive detection platform based on TM specific Nbs and nanozyme-mediated enhanced signaling. The versatile Nb-nanozyme detection platform integrates electron-transfer catalysis and photothermal transduction for dual-mode readout. The established method could offer a promising foundation and inspire further investigation on portable, low-cost point-of-care testing of food allergens and other biomolecular targets. Graphical abstract

Serum ferritin (SF) is an iron-rich protein tightly connected with iron homeostasis, and the variations are frequently observed in diseased states, including iron-deficiency anemia, inflammation, liver disease, and tumors, which renders SF level an indicator of potential malignancies in clinical practice. Nanobodies (Nbs) have been widely explored and developed into theranostic reagents. Surprisingly, no reports stated the identification of anti-SF Nbs, nor the potential of such Nbs as a diagnostic tool. In this study, we generated SF-specific Nbs and provided novel clinical diagnostic approaches to develop an immunoassay. An immune library was constructed after immunizing an alpaca with SF, and five Nbs specifically targeting human SF were retrieved. The obtained Nbs exhibited robust properties including high stability, affinity, and specificity. Then, an ELISA-based test using a heterologous Nb-pair was developed. The calibration curve demonstrated a linear range of SF between 9.0 to 1100 ng/mL, and a limit of detection (LOD) of 1.01 ng/mL. The detecting recovery and coefficient variation (CV) were determined by spiking different concentrations of SF into the serum sample, to verify the successful application of our selected Nbs for SF monitoring. In general, this study generated SF-specific Nbs and demonstrated their potential as diagnostic immunoassay tools.