New AI tool reshapes proteins, quantum spin qubits improve, and long COVID drugs show promise

Researchers unveiled an AI system called Raygun that can resize proteins without losing function, while quantum teams reported spin-qubit devices with error rates below 0.01%, and clinical studies suggest paxlovid and metformin may cut long-COVID risk.

Four separate research teams have pushed spin-qubit quantum hardware forward, delivering devices with between five and 18 qubits and lowering error rates to less than 0.01%, a milestone for scalable quantum computing. In parallel, an artificial-intelligence system named Raygun can redesign natural proteins—either shrinking or enlarging them—by mimicking evolutionary steps such as subunit addition, deletion, or substitution, while preserving structural integrity, according to computational biologist Fajie Yuan.

Clinical investigations reported in The Lancet Infectious Diseases and Clinical Infectious Diseases show that the COVID-19 antiviral paxlovid cuts the chance of developing long-COVID after three months by 40%, and the diabetes medication metformin reduces six-month long-COVID risk by roughly 50%. These findings highlight rapid progress across quantum technology, protein engineering, and post-viral treatment strategies. Additional briefings noted the expanding range of disease-carrying mosquitoes due to climate shifts and highlighted public-health challenges in Ebola-affected regions, but the core story centers on the three scientific breakthroughs.