DeepAFM: A deep learning method to decode protein motion
Peer-Reviewed Publication
Updates every hour. Last Updated: 8-Jun-2026 05:16 ET (8-Jun-2026 09:16 GMT/UTC)
Proteins change shape as they function, and these changes are essential for processes such as drug interactions and cellular activity. Researchers from the Tokyo University of Science developed an AI-based method called DeepAFM that is trained on millions of simulated images representing different protein states, accurately identifying transitions between closed and open states in a protein called SecA. This approach highlights the growing potential of AI to solve complex challenges in biology and medicine.
The function of transposable elements (TEs), during the conversion of stem cells into specialized cell types, remains poorly understood. A study shows that TEs act as binding platforms for key proteins involved in brain development, with many becoming active during neuronal differentiation. The study traces the evolutionary origins of these regulatory elements and demonstrates their role in gene regulation. These findings indicate that TEs may have expanded and refined gene regulatory networks in the brain.
- Achieving an international conservation target to protect almost a third of the world’s land and sea in the next four years could directly affect the lives of almost half the people on the planet, finds a new report.
- The study is the first to consider the social implications of the target at the global level.
- Supporting the people who live near areas that could be designated for nature - financially and otherwise - is vital for success.
CSHL Associate Professor Saket Navlakha and former graduate student Cici Zheng have discovered a naturally occurring Voronoi diagram in Chinese money plants’ leaves. Their research answers a longstanding question in biology regarding the mathematics of looping vein structures and could help explain how plants solve complex problems in nature.
Targeted protein degradation has become one of the most promising strategies in modern drug discovery, enabling scientists to eliminate disease-causing proteins instead of merely blocking them. Now, researchers at CeMM, AITHYRA (both Institutes of the Austrian Academy of Sciences), and CeTPD have discovered that a single small molecule can recruit not one, but two independent protein disposal systems at the same time. This dual mechanism introduces a built-in redundancy that could make future degrader therapies more robust and less vulnerable to resistance. The findings, reported in Nature Chemical Biology (DOI: 10.1038/s41589-026-02224-y), expand the design principles of targeted protein degradation and open new avenues for more resilient medicines.
Researchers from The University of Osaka found that the zinc finger proteins RLF and ZFP292 play redundant roles in stabilizing the CoREST corepressor complex at gene promoters in embryonic stem cells. Deleting both proteins allowed the cells to differentiate, suggesting that they could be useful targets for maintaining stem cell quality.
The Nuffield Council on Bioethics (NCOB) says there is a clear need for guidance on how to conduct ethical research using neural organoids, warning that governance gaps reflect wider weaknesses in the UK’s outdated regulatory ecosystem which could hinder scientific progress.