NiFe-based layered double hydroxides (LDHs), as engineered nanomaterials (ENMs), are widely utilized in industrial applications, environmental remediation, and soil improvement. However, the biological impacts of such ENMs on plant-soil systems remain insufficiently explored. A 50-day soil cultivation experiment was conducted to assess the effects of two-dimensional and three-dimensional NiFe-based LDHs (2D NiFe-LDHs and 3D NiFeS-LDHs) on ryegrass growth and soil ecosystem. Generally, morphology and structure differences influence the biological effects of NiFe-based LDHs. 2D NiFe-LDHs promote ryegrass growth at specific concentrations, whereas 3D NiFeS-LDHs exhibit significant inhibitory effects on ryegrass growth. NiFe-based LDHs directly impacted soil geochemistry. 3D NiFeS-LDHs decreased soil pH while increasing electrical conductivity and soil organic carbon. 2D NiFe-LDHs inhibited soil sucrase (S-SC) activity, but enhanced soil urease (S-UE) activity. Conversely, NiFeS-LDHs inhibited catalase (S-CAT) activity and S-UE activity, while promoting neutral phosphatase (S-NP) activity However, as concentrations increased, the effects of 2D NiFe-LDHs and 3D NiFeS-LDHs on soil properties and enzyme activities became more similar. Both NiFe-based LDHs altered the diversity of soil bacterial and fungal communities, with 3D NiFeS-LDHs exerting a stronger inhibitory effect on fungal diversity at higher concentrations. The structural differences in NiFe-based LDHs exert a distinct and substantial impact on the composition of soil microbial communities and soil enzyme activities. The findings provide critical insights into bio-effects of NiFe-based LDHs, which provide a theoretical basis for the study of potential ecological risks in soil.

Global agriculture faces growing threats from pests, pathogens, and invasive species, intensified by climate change and biodiversity losses. Conventional approaches are limited in both precision and scale, and artificial intelligence (AI) is now reshaping integrated pest management (IPM). Modern agricultural monitoring leverages high-resolution observations, hyperspectral sensors, and the Internet of Things (IoT) to facilitate early detection for diseases. Hybrid AI systems can integrate multi-source data to enhance the accuracy of real-time monitoring of pest and disease dynamics, including the detection and tracking of sparse invasive populations and their biomass even under shifting climate scenarios. They further enable predictive forecasting and the optimization of management strategies. When AI-driven diagnostics are integrated with autonomous robotics, they form a robust framework for epidemic mitigation. Here, we provide a systematic macro-perspective review, bridging foundational AI mechanisms with actionable IPM intelligence. We analyze the evolution of agricultural AI from cross-modal architectures to multi-scale diagnostics and full-lifecycle interventions. Finally, we propose a framework for the global agroecological network, offering a sustainable path toward maximizing productivity while ensuring ecological resilience.

Detecting rare cell populations that drive development, differentiation, and disease-associated transformation remains a central challenge in biology and medicine. Although these populations often represent promising targets for intervention, they are difficult to resolve from single-cell transcriptomic data because most methods rely on homophily-based cell–cell similarity, which can merge rare cells into dominant populations and mask their subtle transcriptional signatures. The challenge is further amplified in multi-sample analyses, where batch correction can dilute rare-cell-specific signals. Here, we present scFormer, a heterogeneous graph transformer (HGT) framework for sensitive and robust rare-cell discovery. scFormer constructs a Z-score-guided cell-gene heterogeneous graph in which highly specific marker genes serve as informational bridges, embedding rare-cell features directly into the graph topology rather than inferring them from global neighbors. This design provides a clear biological rationale for rare-cell recovery, as low-abundance cells can remain connected through shared high-specificity genes even when local cell–cell neighborhoods are sparse. An integrated optimization strategy jointly performs representation learning, clustering, and optional batch correction, enabling rare-cell discovery while preserving biological structure. Across 125 simulated and 18 real datasets, scFormer consistently achieved competitive or superior performance relative to existing approaches. Applied to diverse multi-sample single-cell and spatial transcriptomics datasets, scFormer recovered known but weakly represented populations and revealed previously obscured cell states, including proliferative club cells in the airway epithelium, revival stem cells during intestinal regeneration, and rare embryonic cell states from spatial transcriptomics. Overall, scFormer provides a unified framework for identifying biologically meaningful rare populations while mitigating batch effects in multi-sample datasets.

A-to-I RNA editing, one of the most frequent RNA modifications, is essential for most aspects of RNA metabolism. However, the accuracy of quantifying A-to-I editing levels has been largely neglected, particularly when contrasting next-generation sequencing (NGS) with long-read sequencing (LRS) RNA-seq. To address it, we performed paired NGS and LRS cDNA RNA-seq of both HEK293T and U2OS cells, revealing that the A-to-I editing levels were estimated to be lower by NGS compared with LRS, a conclusion that was confirmed through full-length amplicon sequencing. Furthermore, the lower estimation of A-to-I editing levels is widespread, while we also identified this consequence using released public both same-study and cross-study NGS and LRS RNA-seq data in various human cancer cell lines. In addition, our analyses indicate that the primary factor contributing to this lower estimation is the sequencing read length difference, and a significantly positive correlation between the quantification and read length is identified by analyzing the public NGS RNA-seq data of various cell lines. We additionally found that shorter sequencing reads are more prone to aligning incorrectly to the reference genome, such as uniquely-to-multiply mapped and uniquely mapped-to-unmapped, inducing the lower estimation of A-to-I editing levels. In summary, our analyses indicate that LRS RNA-seq is preferable for accurately quantifying A-to-I editing levels.

Cereals are emerging as attractive platforms for the sustainable production of high-value lipids through metabolic engineering. Although plant lipids play essential biological roles and have considerable economic value, their conventional production from natural sources is often limited by sustainability, scalability and cost. Recent advances in synthetic biology enable the reprogramming of seed lipid metabolism for the tailored synthesis of valuable lipid compounds. In this review, we first summarize the core pathways of fatty acid biosynthesis and triacylglycerol assembly in seeds, together with the genetic transformation and genome editing toolkits available for major cereals. We then highlight recent progress in the heterologous production of specialized lipids, including eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), wax esters, and insect sex pheromones, in engineered plant systems. Finally, we discuss the potential of cereals as scalable and sustainable platforms for the production of high-value lipids. Together, these advances position engineered cereals as promising plant-based factories for applications in agriculture, nutrition, and the emerging bio-based economy.