Researchers have developed an optical "knob" using Poincaré sphere engineering to continuously tune topological states in ferroelectric materials. Unlike prior methods relying on abrupt electrical or mechanical changes, this approach manipulates structured light’s orbital angular momentum to smoothly create and control skyrmions, antiskyrmions, and hybrids on ultrafast timescales, promising new routes for dynamic, precise topological control in nanoscale memory and photonic devices.

Researchers from Nanjing University and Nanjing Medical University have identified sperm microRNAs as key messengers that transmit the benefits of paternal exercise to the next generation. Published in Cell Metabolism, the study shows that offspring of exercise-trained fathers inherit improved exercise endurance and metabolic health. Mechanistically, sperm microRNAs from exercised fathers suppress NCoR1 in early embryos, reprogramming gene networks to enhance mitochondrial biogenesis and glucose metabolism. These findings establish a PGC-1α–sperm microRNA–NCoR1 axis for intergenerational inheritance and suggest that paternal exercise before conception may offer a simple, cost-effective way to improve offspring health and help break the cycle of obesity and chronic disease.

Scientists report the first direct real-time measurement and control of quantum uncertainty dynamics using ultrafast squeezed light pulses. This attosecond-resolved breakthrough not only advances fundamental quantum physics but also demonstrates petahertz-scale secure communication protocols, paving the way for next-generation quantum technologies in secure communication, quantum computing, and ultrafast spectroscopy.

Regression analysis is essential in biomedical research for exploring relationships between phenotypic or clinical outcomes and diverse predictors. However, constructing multiple univariate and multivariate models is often hindered by the lack of robust tools for batch regression in R, forcing researchers to rely on custom scripts. To address this gap, we developed bregr, an open-source R package built in the tidyverse style, leveraging the object-oriented programming strategy for enhanced extensibility. bregr streamlines batch processing of diverse regression models, including generalized linear, Cox proportional hazards, and mixed-effects, using native R pipes. It provides tidy outputs, integrated visualization, parallel computing capabilities, and a cohesive workflow, enabling efficient execution of hundreds of models with structured results for downstream analysis. Available on CRAN, bregr enhances efficiency, reproducibility, and scalability in biomedical research and beyond.

Scientists have decoded the molecular pathway behind the striking yellow petals of Camellia nitidissima and applied these insights to boost the nutritional profile of tomatoes. 

Researchers from Shandong University and South China University of Technology have developed E2E, an effective and comprehensive R package to simplify the construction of ensemble models for diagnosis and prognosis. The package integrates bagging, stacking, and voting strategies, offering flexible model selection and improved handling of imbalanced data. Validated on TCGA and CHARLS datasets, E2E demonstrates strong performance in both diagnosis and prognosis tasks.

Banana production worldwide is threatened by narrow genetic diversity, pests, diseases, and climate pressures.

Grapevines worldwide face heavy losses from fungal diseases, with white rot posing one of the greatest threats. 

The world’s growing livestock industry produces millions of tons of animal manure every year, creating both environmental and public health challenges. A new study shows that this abundant waste material could be transformed into a valuable solution for one of today’s most pressing problems—wastewater pollution.

Prof. Shigui Chen's team at the Institute of Advanced Studies at Wuhan University recently achieved, through a step-by-step construction strategy, the stabilization of the [N⋯Cl⋯N]⁺ halogen bond at the framework level, thus breaking through the bottleneck of the inability of chlorine (I) halogen bonds to form bonds under conventional conditions and successfully constructed a class of stable chlorine (I) bridged two-dimensional halogen bond organic frameworks (XOF(Cl)-TPy-BF₄/OTf). This framework not only has excellent crystallinity, chemical stability and thermal stability, but also provides an ideal platform for subsequent functionalization (such as the introduction of catalytic sites). As a cationic framework, XOF(Cl) is anion-exchanged with PdCl₄²⁻ and then reduced to generate stable Pd(0) clusters within the framework. The Pd(0) cluster-loaded XOF(Cl)-TPy-Pd⁰ exhibits excellent catalytic performance in various palladium-catalyzed coupling reactions (such as Suzuki, Heck, and Sonogashira couplings), achieving high yields even under mild conditions. In the synthesis of industrial biphenyl liquid crystal molecules, no palladium residue was detected in XOF(Cl)-TPy-Pd^0. In addition, the catalyst can be efficiently recovered by simple filtration and can be recycled many times, showing excellent practicality. These results were published as an open access article in CCS Chemistry, the flagship journal of the Chinese Chemical Society.