We are pleased to present a curated collection of our recent cover articles, featuring breakthroughs ranging from enzyme–chemical interaction prediction, genome-scale modeling and novel biosensor development to metabolic engineering of industrially relevant microbial hosts and next-generation gene-editing tools. These studies highlight the field's rapid evolution from fundamental tool discovery to translational industrial applications.

Researchers developed WiFo-2, a wireless foundation model for heterogeneous future wireless systems. Pretrained on 11.6 billion CSI points, it surpassed fully supervised task-specific models in three zero-shot reconstruction tasks and used 1% of labeled data for nine downstream tasks. Its confidence estimates support reliable operation. A prototype validated three tasks across indoor and outdoor scenarios and bands spanning sub-6 GHz and millimeter-wave.

A recent study published in National Science Review has revealed that the Indian Ocean Dipole (IOD), traditionally regarded as a single climate mode, consists of two distinct types with different physical mechanisms and climate impacts. The findings advance the understanding of Indian Ocean variability and provide new insights into how IOD events influence regional climate.

Improving the subseasonal-to-seasonal (S2S) forecast skill for East Asian winter cold surges remains a grand challenge in operational meteorology. Using ECMWF hindcast data, a research team systematically identified the "high-skill windows" of week-3 cold surge forecasts. They found that a "warm-west/cold-east" sea surface temperature dipole over the mid-latitude North Pacific emerges as a consistent precursor of these windows under both El Niño and La Niña conditions, but that the dipole alone is not enough: high skill emerges only when the SST anomalies in all the key regions, including the Barents-Kara Seas, the Indian Ocean and the tropical Pacific, meet their thresholds at the same time. This makes it possible to pre-screen a subseasonal forecast's reliability before it is issued. The work was published in SCIENCE CHINA Earth Sciences.

Hafnia-based ferroelectric memories are emerging as promising candidates for next-generation nonvolatile memory because they combine excellent scalability with compatibility with CMOS technology. Further reducing the thickness of ferroelectric films can lower operating voltage and increase storage density, but maintaining reliability at high temperatures becomes increasingly challenging. Researchers at Fudan University have now shown that tailoring the size and orientation of ferroelectric grains can help address this challenge. Their ultrathin devices remain highly stable at 225 °C, pointing to a route toward dense and reliable ferroelectric memories for harsh thermal environments. The study was published in Science Bulletin.