artificial intelligence
Eli Lilly and NVIDIA team up to build supercomputer
Eli Lilly said yesterday that it will partner with chipmaker NVIDIA to build what it claims will be the “most powerful supercomputer owned and operated by a pharmaceutical company.”
The vast majority of the supercomputer's capacity will be going toward using AI for discovery of small molecules, large molecules, and genetic medicines. Some of the computing power will also go toward projects in clinical trials, manufacturing, and quality assurance processes.
It's yet to be seen, however, if the technology will actually yield effective medicines. Just having raw supercomputer power may not be enough.
Read more from STAT's Brittany Trang.
drug pricing
Bausch's withdrawal from Medicaid may be a sign of departures to come
Bausch Health this month suddenly stopped participating in Medicaid and a federal government drug discount scheme known as 340B. The move signals that other drugmakers may also be rethinking their decisions to provide medicines to these widely used programs, experts said.
The pharmaceutical industry has been under increased pressure to lower prices due to policy changes that affect the calculus used to set prices. Notably, drug companies face penalties they must pay to Medicaid when they raise the price of a medicine above the inflation rate, a consequence of a provision in a federal law that went into effect in early 2024.
Additionally, drugmakers are required to participate in the 340B program to have their drugs covered by Medicaid. The discount program, which was created to help hospitals and clinics care for low-income and rural patients, has been ballooning.
Read more from STAT's Ed Silverman.
liquid biopsy
Natera, known for spotting cancer recurrence, wades into early detection
From my colleague Jonathan Wosen: Natera, a Texas-based genetic testing company, has long focused on detecting cancer cells that linger in treated patients, stubborn holdouts that can become deadly if allowed to grow unchecked. But data reported yesterday show how Natera is looking to expand into a new area: cancer prevention and early detection.
The company, which has a $26 billion market cap, shared results from an analysis of 1,400 participants at average risk of colon cancer who gave blood before receiving a colonoscopy. Researchers found that Natera’s blood-based test detected 22.5% of advanced adenomas, abnormal growths that often presage cancer but which can easily be cut out to prevent disease. The biotech’s pre-cancer detection rate is higher than that of a blood-based test sold by Guardant Health and a test being developed by Exact Sciences and its partner Freenome.
Market analysts were cautiously optimistic, noting that the results seemed encouraging but that they’re still awaiting results from Natera’s pivotal study, FIND-CRC, which data expected in 2027.
artificial intelligence
A new protein model said to outperform AlphaFold3
From my colleague Britany Trang: Genesis Therapeutics, the AI-focused biotech based on technology from former a16z partner Vijay Pande’s Stanford lab, announced a new protein model yesterday at chipmaker NVIDIA’s conference.
Pearl, a “generative foundation model” for biomolecular structure prediction, outperforms existing models such as AlphaFold3 in predicting how small molecules bind to proteins, Genesis said.
The company is also changing its name to Genesis Molecular AI to better emphasize that it’s “pioneering foundation models for molecular AI” and to help in recruiting AI talent, said a spokesperson. The company also said the move doesn’t signal any changes to its pipeline, which focuses on small molecules for I/O generated and optimized with the help of its AI models.