Governments and Startups Bet Big on Personalized Medicine
The day's boldest move came from Washington. ARPA-H, the U.S. government's health-research agency, is committing $160 million to develop "bespoke" drug therapies — custom gene-editing treatments engineered to correct extremely rare genetic mutations in individual patients. The program, reported by BioPharma Dive, points toward a future where a therapy might be built for a single person's DNA rather than a broad population.
The theme of second chances ran through the day, too. According to Endpoints News, Cellular Intelligence has picked up a Parkinson's disease stem cell therapy that one of the world's biggest drugmakers, Novo, had shelved. A program abandoned by a giant has found a new home — a reminder that promising science doesn't always die when a big sponsor walks away.
Setbacks: Roche Retreats, and Regulators Push Back
Not every long bet pays off. Roche is walking away from tominersen, its experimental Huntington's disease treatment, after pursuing it for more than a decade. Endpoints News reports the drug simply didn't perform well enough to justify continuing — a sobering close to a marathon effort against a devastating illness.
Regulators were busy on both sides of the Atlantic. In the UK, the cost watchdog NICE recommended against continuing to reimburse Amgen's lung cancer drug Lumakras, a decision that could cut off access for future patients. And in the U.S., the FDA publicly rebuked Lundbeck over inflated online marketing claims for one of its migraine treatments — a pointed signal on how drugmakers promote their products.
AI Moves From Hype to Workflow
Artificial intelligence threaded through nearly every corner of the day's news. Two fresh collaborations — one linking Evogene with Tel Aviv University, alongside a deal involving Certara — show the industry leaning harder on AI to speed the hunt for new medicines, per Yahoo Finance.
The ambitions are getting more sweeping. A system called Biomni is being positioned to handle biomedical research across the full arc of inquiry — from searching published literature all the way to designing experiments. Meanwhile, Nature spotlighted "neuro-symbolic AI," a hybrid approach drawing interest for medical diagnosis, and IGC Pharma said it will bring its AI toolkit for Alzheimer's research to the AAIC 2026 conference.
The Money Follows
Forecasters see a boom ahead. One report circulated via GlobeNewswire and picked up by Yahoo Finance projects the market for generative AI in pharma to reach $9.36 billion by 2030. A separate forecast predicts "explosive growth" in in silico protein design software — designing proteins on computers rather than at the lab bench — with AI and automation named as the drivers.
The throughline is clear: whether it's custom cures for one patient, revived therapies, or AI systems that research on their own, the industry is placing its chips on precision and speed. The wins and the walk-aways landed on the same day — proof that in pharma, ambition and hard limits still travel together.