Researchers have developed SuSha, a multi-model ensemble learning tool that predicts a microorganism's salt tolerance directly from its genome. Trained on the amino acid composition of 123 bacterial and archaeal species and validated on 2,678 metagenomes from freshwater, marine, and hypersaline habitats worldwide, SuSha achieved over 90% accuracy for extreme halophiles. It offers a powerful way to study the more than 81% of microbes that cannot be cultured in the lab.

For the first time, researchers have shown that ketamine activates two separate mechanisms in the brain, one for pain relief and another for altering the state of consciousness.

The findings may help in developing treatments that provide the benefits of ketamine while reducing its psychological side effects.

Carbonate cerium and neodymium isotope records from South China show that low-oxygen waters invaded shallow habitats twice during the Late Ordovician Mass Extinction, with a brief recovery between the two pulses. Earth-system models suggest that rapid climate swings reorganized continental weathering and nutrient delivery, amplifying marine oxygen loss and ecological stress.

The research team from National University of Defense Technology systematically described the memory efficiency issues in the training process of large language models (LLMs) with Transformer architecture for scientific fields. As the scale of models continues to expand, memory requirements have become a key bottleneck restricting their applications in scientific fields such as biology, medicine, chemistry, and meteorology. The article systematically reviewed memory-efficient pre-training technologies at the algorithm level, system level, and software-hardware collaborative optimization, and combined with the AlphaFold 2 case, demonstrated how customized memory optimization methods can reduce storage overhead while maintaining prediction accuracy.