A perspective published by researchers at the University of Science and Technology Beijing discusses research by Stoddart et al. in Nature Chemistry that demonstrates that engineered supramolecular crystals can optimize hydrogen storage performance.

A comprehensive review reveals how large language models (LLMs) are transforming bioinformatics by enabling unprecedented analysis of biological sequences, protein structures, and multi-omics data. The study highlights LLMs’ capabilities in accelerating drug discovery, improving disease diagnostics, and mining biomedical literature—while addressing critical challenges like interpretability and data bias. Future directions emphasize multimodal AI integration and ethical frameworks to advance precision medicine.

A team of international researchers has published a comprehensive review in Advanced Science detailing how combining protein three-dimensional spatial structures with artificial intelligence is reshaping the way scientists predict and design drug combinations. By integrating high-resolution protein modeling with advanced AI algorithms, the approach offers unprecedented accuracy in identifying synergistic and antagonistic effects, paving the way for safer and more effective treatments for cancer, infectious diseases, and metabolic disorders.

Team Atlanta, led by Professor Insu Yun of the Department of Electrical and Electronic Engineering at KAIST and Tae-soo Kim, an executive from Samsung Research, along with researchers from POSTECH and Georgia Tech, won the final championship at the AI Cyber Challenge (AIxCC) hosted by the Defense Advanced Research Projects Agency (DARPA). The final was held at the world's largest hacking conference, DEF CON 33, in Las Vegas on August 8 (local time).

Professor Seungbum Koo’s research team received the Clinical Biomechanics Award at the 30th International Society of Biomechanics (ISB) Conference, held in July 2025 in Stockholm, Sweden. The Plenary Lecture was delivered by first author and Ph.D. candidate Jeongseok Oh. This research was conducted in collaboration with Professor Joon-Ho Wang’s team at Samsung Medical Center.

 

Folding structures are widely used in robot design as an intuitive and efficient shape-morphing mechanism, with applications explored in space and aerospace robots, soft robots, and foldable grippers (hands). However, existing folding mechanisms have fixed hinges and folding directions, requiring redesign and reconstruction every time the environment or task changes. A Korean research team has now developed a “field-programmable robotic folding sheet” that can be programmed in real time according to its surroundings, significantly enhancing robots’ shape-morphing capabilities and opening new possibilities in robotics.

KAIST (President Kwang Hyung Lee) announced on the 6th that Professors Jung Kim and Inkyu Park of the Department of Mechanical Engineering have developed the foundational technology for a “field-programmable robotic folding sheet” that enables real-time shape programming.