In new research from the University of Chicago Pritzker School of Molecular Engineering, Argonne National Laboratory and Purdue University, the AI-driven robotic lab Polybot designed and built new polymers that hit precise, targeted shades of orange and green – a task that takes researchers decades of experience to learn – in just 72 hours.

Neuroimaging analysis in brain disorders faces a persistent challenge: brain signals are complex and high-dimensional, while high-quality labeled datasets remain limited. This review article systematically examines how self-supervised learning can help address that gap by learning meaningful representations directly from unlabeled neuroimaging data. It covers major methodological families, including contrastive, generative, and hybrid generative-contrastive approaches, and discusses their applications in functional MRI, EEG, and multimodal brain network analysis.

The review argues that self-supervised learning offers more than annotation efficiency. It may enable more transferable and clinically useful representations for disease screening, diagnosis, and prognosis across heterogeneous datasets and disorders. At the same time, interpretability, data heterogeneity, missing modalities, and clinical validation remain major barriers. Future work will likely focus on stronger multimodal fusion, better cross-site generalization, and more clinically adaptable model design.

Researchers at the University of Missouri have discovered certain proteins may be the key to saving plants’ lives when multiple stressors hit at the same time. This knowledge may one day lead to crops that are more resistant to harsh conditions brought on by multiple stressors during the same growing seasons.

In a recent study, Mizzou scientists found that Arabidopsis thaliana, a plant that serves as a popular model organism for biology research, needs a specific protein to protect itself when exposed to simultaneous stress from excessive heat, sunlight and salty soil. The findings pave the way for scientists to better understand the underlying cellular biology that allows plants to survive even when hit by multiple stressors.

Segregated conductive polymer composites are attracting growing attention as electromagnetic shielding and thermal interface materials. However, lack of theoretical models and micro-void formation during fabrication hinders their applicability. In a new study, researchers have developed a new fabrication strategy that minimizes micro-void formation while significantly improving electrical and thermal properties. Additionally, they have proposed new performance prediction models that could accelerate design of customized materials across industries.

Today, most of the salmon consumed in Japan is imported from countries like Chile and Norway, according to the Ministry of Agriculture, Forestry and Fisheries. But just two decades ago, Japanese chum salmon made up a much larger share of domestic salmon consumption. Their numbers have declined sharply in recent years, and new research from Hokkaido University suggests that this decline may be linked to the loss of their natural habitats along their migratory routes.

A new study published in Big Earth Data introduces Geo-Disasters, a database that geocodes 9,217 climate-related disasters reported by EM-DAT from 1990 to 2023. The database enhances disaster risk analysis by providing accurate geocoded locations and offering an open, reproducible framework for integrating spatial climate and socioeconomic data. This framework supports machine learning applications for climate risk, vulnerability, and adaptation studies. The Geo-Disasters database aims to bridge critical data gaps in global climate-hazard risk assessment and inform more equitable adaptation planning. The link to Geo-Disasters Database is https://doi.org/10.5281/zenodo.15487667.