👋 Hey, let’s get into it. Anthropic bought a biotech whose all-hands meeting had exactly two hands. Both belonged to the founder.

This week, Anthropic disclosed its July acquisition of Mithril Medicine, whose only full-time employee was founder Cody Tranbarger. He used Claude to coordinate operations with part-time consultants and 11 outside labs.

Working through that network, Mithril moved a potential muscular-dystrophy treatment from idea to promising mouse results in under 18 months, according to Anthropic. Human trials could begin as early as late 2028 if further work succeeds.

💉 Lilly’s Retatrutide Beats Novo’s Ozempic on Blood Sugar and Weight Loss
Lilly’s latest retatrutide results give Novo another reason to wince. In a Phase 3 trial in adults with type 2 diabetes, patients taking 12 mg lost 20.9% of their weight over 80 weeks, versus 9.4% with Ozempic 2 mg. Blood sugar also improved more. Those estimates assume patients stayed on treatment. The catch: 7.9% stopped retatrutide because of side effects, versus just 2.7% on Ozempic. Retatrutide remains experimental, and full results are pending. Still, beating an established rival head-to-head carries weight. Novo may want to consider muting Lilly’s press alerts.

💰 Ferring Agrees to Buy Fertility Biotech Gameto for Up to $1.38B
Ferring agreed to acquire Gameto for up to $1.38 billion, including conditional development, regulatory, and commercial milestones. The upfront payment wasn’t disclosed. Gameto’s Fertilo uses stem cell derived ovarian support cells to help retrieved eggs mature outside the body, aiming to simplify fertility treatment. Pairing that approach with Ferring’s established reproductive medicine business could help it reach more clinics if development succeeds. Fertilo is currently in Phase 3 testing and remains investigational.

💵 Astellas Takes an Audax Cancer Drug Option Worth Up to $1.34B
Astellas secured an exclusive option to license Audax’s preclinical DAX-044, which combines an antibody with the immune-stimulating protein IL-12. The aim is to concentrate activity in tumors while limiting the systemic toxicity that has dogged IL-12 development. Upfront and potential option-exercise payments total $285 million, but another $1.05 billion in milestones, plus royalties, could follow. Audax will advance the candidate into Phase 1, expected in 2027, with Astellas taking over if it exercises the option. That’s a sizeable RSVP before the first human readout.

🧬 FDA Proposes One Application for Multiple Personalized Gene Editing Variants
The FDA wants personalized gene editing to spend less time reinventing its paperwork. New draft guidance would let multiple patient-specific variants sit under one biologics license application, provided they stay within predefined boundaries. Later variants could draw on the original clinical evidence plus their own laboratory safety and activity data, avoiding a separate trial for every mutation. That could improve the economics of treating ultra-rare diseases, where conventional trial populations may barely exist. The initial application still needs evidence of safety and effectiveness. This remains a nonbinding proposal, with comments due November 27.

🏭 Teva Plans Layoffs as It Closes a New Jersey Plant
Teva plans to close its underutilized Fairfield plant at the end of January, with 61 workers affected and layoffs effective February 1, 2027.

🙌 Genmab’s Drug Combo Cuts Death Risk 50% in Head and Neck Cancer
Investigational petosemtamab plus Keytruda beat Keytruda alone in Phase 3 first-line PD-L1-positive recurrent/metastatic head and neck squamous cell carcinoma.

❌ Artiva Ends Cell Therapy and Plans 94% Staff Cut
Artiva discontinued its autoimmune cell therapy after durability and safety concerns, canceled planned Phase 3 development, and is evaluating alternatives, including liquidation.

🎯 Merck’s Sac-TMT Cuts Death Risk 32% in Endometrial Cancer
Median survival reached 19.0 versus 12.1 months with chemotherapy in advanced or recurrent endometrial cancer previously treated with platinum chemotherapy and immunotherapy.

🧪 Abbott’s Upgraded Cancer Test Tracks Up to 5,000 Variants
Abbott’s upgraded Oncodetect blood test tracks up to 5,000 tumor-specific variants, versus 200 previously, to monitor residual cancer.

🎯 GSK’s B7-H3 ADC Cuts Progression or Death Risk 67% in Relapsed Osteosarcoma
China’s Phase 3 trial showed 5.5 versus 1.4 months’ median progression-free survival against chemotherapy, although severe treatment-related adverse events affected 93% versus 83%.

📈 WuXi AppTec Raises Guidance Despite U.S. Policy Uncertainty
Nine month revenue rose 38.6% to approximately $6.83 billion, prompting WuXi to lift its full-year revenue forecast to roughly $9.15–9.29 billion as U.S. policy risks persist.

🤢 Vanda’s Nereus Reduces GLP-1 Induced Vomiting in Phase 3
Nereus lowered vomiting rates to 43%, versus 58.5% with placebo, after participants received one semaglutide dose without the usual gradual dose increases.

Here we are, a biotech and pharma newsletter, giving Anthropic two stories in one issue. That probably tells you something about how aggressively AI companies are entering the biotech race. Anthropic, or as I like to call it, “the Claude Father,” is building quite the scientific family.

Its IPO filing says nine of the ten largest pharmaceutical companies were customers as of September 30. That’s a substantial foothold inside drug development. The opportunity extends across discovery, experiment planning, manufacturing documentation, and regulatory work. Anyone who has watched a program stall between departments can appreciate the appeal of connecting those pieces.

The ambitions also extend into the laboratory. Anthropic’s Model Hardware Standard, currently in research preview, is intended to connect AI agents with microscopes, liquid handlers, and robotic arms. Alongside partnerships and internal drug programs, that brings the company closer to conducting the work itself.

Which raises an interesting business question: how long does pharma’s helpful AI partner stay just a partner? I can imagine these companies eventually developing medicines that compete with their customers’ pipelines. Another possibility is an early-stage discovery business, producing promising candidates and licensing them to whichever pharma company brings the best offer. Today’s software vendor could become tomorrow’s partnering meeting.

Either route requires much more than impressive model performance. The filing acknowledges the responsibilities that come with operating labs, handling biological materials, and navigating lengthy, uncertain drug development.

Strong software adoption gives Anthropic distribution, but reproducible experiments and clinical evidence will determine how far its scientific ambitions travel.

This is how much Novartis recorded as a partial impairment on del-desiran, acquired through its $12 billion Avidity deal, after September’s Phase 3 setback in myotonic dystrophy type 1. Despite this expensive bump in the road, the company claims to be optimistic about the acquisition’s longer-term prospects.

Here’s what biotech Redditors are talking about:

💰 Biotech Workers Share Their Pay History, and Everyone’s Listening
Everyone wants to know what everyone else makes. Fact. One Redditor shared their journey from $50,000 in academic lab work to a $225,000 base salary in a machine learning and product role. Other biotech and pharma workers chimed in with their own career moves and pay histories. Plenty to browse before asking your boss for a raise on the grounds that “Reddit user BioBro69 says I’m underpaid.”

🎓 Why Does This Job Need a PhD?
Ah, the age-old question: why does this job require a PhD? A thread about non-bench roles sparked the familiar debate over scientific training, credential filters, and whether relevant experience gets enough credit. Can’t we all be friends? Meanwhile, somewhere in Silicon Valley, a college dropout is retiring before their biology friends finish their dissertations. One group is choosing a yacht name. The other is being asked for “just one more experiment.”

🦖 Nanotyrannus Skeleton Sells for $47.3M at Sotheby’s
The skeleton nicknamed Jodi fetched $47.3 million, becoming the second most expensive dinosaur sold at auction. The buyer’s identity remains unknown.

📺 Lilly Defeats Novo’s Request to Temporarily Block Disputed GLP-1 Ads
A judge denied Novo’s request to halt Lilly’s ads while the lawsuit proceeds, allowing them to keep running as the companies fight in court.

🦠 Regeneron Adds $270M to Its U.S. Ebola Stockpile Agreement
Regeneron expects approximately $270 million for additional deliveries of Inmazeb, its approved antibody treatment for Zaire Ebola virus infection, through 2031.

🧫 Syntax Moves Stem Cells Toward Insulin Producing Cells in Four Days
Syntax Bio says its Cellgorithm platform guided human stem cells toward becoming insulin-producing pancreatic cells in four days. It uses timed CRISPR activation to switch on the cells’ own genes without cutting or editing the genome. The goal: faster, more controllable cell production for regenerative medicine. Next comes proving these cells can do the job, including reliably responding to glucose. Biology has accepted the itinerary, but we still need to check where the cells actually arrived.

🦕 October 29, 1831: A Future “Bone Wars” Rival Is Born
Othniel Charles Marsh was born near Lockport, New York, on this day in 1831. The future Yale paleontologist, working with collectors and field teams, helped assemble major fossil collections and named dinosaurs including Stegosaurus and Brontosaurus. His bitter competition with Edward Drinker Cope became the Bone Wars, expanding paleontological knowledge while turning scientific priority into a personal feud. Modern researchers will recognize the ingredients: prestige, publication credit, and someone absolutely convinced the other lab is getting too much attention. The specimens were considerably larger.

I keep thinking about Mithril’s one person payroll. Getting acquired by Anthropic deserves an office party. Catering, a band, an open bar. The works.

I’m picturing the vendors arriving, setting everything up, then finally asking, “So, when will the rest of the employees get here?”

“They’re here. Hey, I’m Cody.”

See you Tuesday for the next issue. Please enjoy the remaining 47 crab cakes.

✍️ Today’s email was brought to you by Josh Martin. Connect with me on LinkedIn.

Put your brand in front of scientists, founders, investors, operators, and other professionals across biotech and pharma.

BioNucleus Briefing has 2,500+ combined subscriptions across our email and LinkedIn newsletters, with one exclusive sponsor per issue.

Think your company could be a fit? Learn more about our audience and get in touch about an upcoming campaign.

If you’re enjoying BioNucleus, share it with a friend or coworker who’d get value from it. We’re a small operation, and every share makes a bigger difference than you think.

Was this email forwarded to you? Sign up here 👇

Subscribe for free

Follow us on social and stay one step ahead