Top Story Today
Big Pharma's Compute Arms Race Just Got Its F1 Engine
Bristol Myers Squibb is now the proud owner of an NVIDIA Vera Rubin DGX SuperPOD, the most powerful AI system in any life sciences company. The purchase makes BMS the latest top-five pharma giant to buy a supercomputer, following Lilly, but the first to skip ahead to NVIDIA's newest architecture, which delivers up to 10x more computing power per megawatt. BMS plans to train proprietary foundation models on decades of internal data, with the goal of significantly accelerating clinical trial timelines. One sickle cell candidate already in trials reportedly wouldn't exist without AI.
Why it matters: When one pharma giant buys an AI supercomputer, it's a strategy choice. When three do it in under a year, it's a new competitive requirement. The era of AI infrastructure as an optional R&D experiment is over; it's becoming table stakes for drug discovery.
Read more →Global Health and Access
India Just Got Its First Dengue Vaccine. It Only Took Forever.
After decades with zero approved dengue vaccines, India cleared Takeda's QDENGA for ages 4 to 60. The critical distinction from the Dengvaxia disaster: no blood test required before vaccination. In the TIDES trial, QDENGA prevented 80% of dengue cases and 90% of hospitalizations. India's true dengue burden may be 282 times higher than official counts suggest, making this approval potentially transformative for 1.4 billion people.
Read more →China's Insurance List Cracks Open for Two GLP-1 Drugs
Pfizer's ecnoglutide and Innovent's mazdutide passed preliminary review for China's national insurance formulary, putting them on track for price negotiations this fall. Getting on the list means access to hundreds of millions of patients, but historically requires 50-70% price cuts. In a head-to-head trial, mazdutide delivered 10.3% weight loss versus 6.0% for semaglutide.
Read more →The Pentagon Dropped Its Flu Shot Rule. Then 300 Troops Got Sick.
After Defense Secretary Hegseth made flu shots optional in April, vaccination rates at Lackland Air Force Base plunged from near-100% to roughly 40%. By June, nearly 300 recruits were sick with influenza, one trainee died under circumstances still being investigated, and the Pentagon quietly reinstated the mandate for boot camps across all branches. Officials called the timing "purely coincidental."
Read more →Science and Discovery
Gene Therapy Via Ultrasound and Bubbles: Stunning, or Too Good to Be True?
SonoThera claims it can deliver full-length dystrophin (the gene behind Duchenne muscular dystrophy) using microbubbles and ultrasound instead of viruses. Primate data showed expression hitting 50% of normal levels in skeletal muscle, a number so high that top scientists say they "find it hard to believe." The $186 million-backed startup hasn't dosed a human yet, but the approach could sidestep every major limitation of viral gene therapy.
Read more →This Obesity Startup Thinks the GLP-1 Gold Rush Is Already Over
While dozens of companies chase GLP-1 drugs, newly launched Mwyngil Therapeutics raised $60 million to bet on completely different biology. Its lead target, GPR75, produced 25% body weight reduction in animals while preserving muscle mass (the thing GLP-1 critics keep hammering). The company is backed by OrbiMed and plans first-in-human trials around 2027.
Read more →One CRISPR Shot to Kill Your Cholesterol Forever? Scribe Files for IPO.
Co-founded by Nobel laureate Jennifer Doudna, Scribe Therapeutics filed to raise up to $75 million for a CRISPR system that silences cholesterol genes without cutting DNA. In monkeys, a single dose dropped LDL cholesterol by 50% or more for over 515 days. The Phase 1 trial starts mid-2026, but Scribe enters a crowded field where Verve already has human data.
Read more →Regulatory and Policy
The FDA Liked the Drug. It Hated the Factory.
Lantheus' cancer imaging kit got rejected by the FDA with zero concerns about safety or efficacy. The sole problem: unresolved inspection issues at a third-party manufacturer. It's part of a troubling pattern; roughly 74% of FDA Complete Response Letters since 2020 have been driven by manufacturing deficiencies, not science failures. Good drugs are dying on the factory floor.
Read more →A Senator Thinks the FDA Is Failing Rare Disease Patients
Sen. Ron Johnson launched a formal investigation into FDA rare disease rejections, calling its demands for new trials "bureaucratic idiocy." The agency has rejected 23 rare disease therapies since 2025, and families have staged symbolic funerals outside FDA headquarters. The probe could reshape approval standards for ultra-rare conditions just as user-fee reauthorization negotiations give Congress real leverage.
Read more →New Drug Prices Dropped 40% Last Year. Don't Get Excited.
Median launch prices for new drugs fell to $216,000 in 2025, down from over $370,000 the year before. Sounds like progress, but the real reason is simpler: fewer multi-million-dollar gene therapies happened to launch that year. The underlying pricing behavior hasn't budged. Back in 2008, the median was just $2,000.
Read more →$330 Million Says Eyeballs Are the Next Big Thing in Biotech
Ollin Biosciences closed one of the largest private biotech rounds in two years, raising $330 million to take its eye disease drug OLN324 into Phase 3 trials against Roche's $5 billion Vabysmo. The kicker: Ollin's CEO helped develop Vabysmo at Genentech before leaving to build what he believes is a better version. Investors including ARCH, a16z, and Blackstone agree.
Read more →