Topics

Researchers at the University of Washington have computationally designed synthetic proteases - enzymes that precisely cut other proteins at chosen sequences. Two studies published in Nature describe zinc-based metalloproteases and cysteine proteases created entirely from scratch using deep learning. These designer enzymes show rate enhancements millions to hundreds of millions of times faster than uncatalyzed reactions, and they work inside living cells. The work opens paths to destroying disease proteins, clearing misfolded proteins from neurodegenerative diseases, and activating drugs only in targeted tissues.

Google Research has developed a framework called VERID that connects AI agents to physical laboratory robots, enabling closed-loop molecular design cycles. The system combines a digital twin of laboratory equipment with machine learning to run experiments, learn from results, and self-improve.

ORCA Computing unveils its PT-3 photonic quantum computer, reporting gains in AI-driven protein discovery and industrial optimization through hybrid quantum-classical computing.

Researchers at Chalmers University of Technology have developed an AI system that performs the full cycle of scientific research on its own. The AI scientist generates hypotheses, designs experiments, and interprets results with minimal human input. The system was tested on brewer's yeast, demonstrating a closed-loop approach to biological discovery.

Scientists’ judgments could help automated microscopes pursue discoveries that conventional numerical targets overlook.

A collaboration between Anthropic's Claude and mathematician Youness Lamzouri has produced a new unconditional proof that more than 67% of non-trivial zeros of the Riemann zeta function lie on the critical line, with additional results.

A new framework embeds physical laws directly into AI systems for materials discovery, addressing key weaknesses in purely data-driven approaches: opaque predictions, poor performance outside training data, and violations of known scientific principles.