DiaCardia, a novel artificial intelligence model that can accurately identify individuals with prediabetes using either 12-lead or single-lead electrocardiogram (ECG) data, has been recently developed. This breakthrough holds promise for future home-based prediabetes screening using consumer wearable devices, without requiring invasive blood tests. This study emphasizes the utility of the ECG as a powerful biomarker and highlights that the innovative AI model can contribute to the prevention of diabetes.

A £3.5 million UK-Japan research project will transform sign language AI by ensuring training is on real conversations between Deaf people, not interpreted signing. 

A nanostructure made of silver and an atomically thin semiconductor layer can be turned into an ultrafast switching mirror device that may function as an optical transistor – with a switching speed around 10,000 times faster than an electronic transistor. An international team of researchers led by the University of Oldenburg, Germany, describes this effect in a paper published in the current issue of Nature Nanotechnology. Ultrafast light switches offer interesting prospects for optical data processing, chip manufacturing, optical sensors or quantum computers.

Scientists have developed a new way to fabricate three-dimensional nanoscale devices from single-crystal materials using a focused ion beam instrument. The group used this new method to carve helical-shaped devices from a topological magnet composed of cobalt, tin, and sulfur.

In a multi-institutional study published today in Nature Nanotechnology, researchers from The University of Texas MD Anderson Cancer Center reported that engineered bispecific nanobodies successfully strengthened mucosal defenses in the respiratory tract, improving protection against influenza infection and reducing SARS-CoV-2 transmission in vivo.

Researchers report a room-temperature organic microcavity where two different forms of spin–orbit coupling act together to produce the optical spin Hall effect (OSHE), a way to route light based on its polarization “spin”. By tuning photon momentum (viewing angle), they observe two coexisting spin textures in one device: a quadrupole pattern at high momentum and a mirror-symmetric pattern at low momentum. The hybrid effect also sustains a long-lived polarization bias of about 300 picoseconds, pointing to robust polarization control for future spin-photonic and topological photonic technologies.

A comprehensive review published in Science China Life Sciences by a collaborative team led by Prof. Wenjie Shu (Bioinformatics Center of AMMS) et al. highlights that Protein Foundation Models (pFMs) have emerged as game-changers in life science.

These AI tools, trained on large-scale datasets, can predict protein characteristics and design new proteins with desired functions. This review explores the progress, uses, challenges, and future of pFMs. It looks at the diverse data—from genetic sequences to 3D structures and functional information—that these models learn from. It covers key AI methods and highlights real-world impacts in research, protein design, and medicine. The article also discusses major challenges, including data scarcity and the complexity of validating model outputs. Looking ahead, the review highlights promising developments, such as modeling protein interactions and building virtual cell systems, which have the potential to enpower the next generation of bioengineering. This comprehensive overview serves as both a valuable resource for computational researchers and a strategic reference for scientists using these tools in related fields.

Spin-out PeroCycle, which is developing a closed carbon loop system to decarbonise foundation industries, announces two new appointments strengthen the company’s leadership team at a pivotal stage of growth.