Noor’s Newsletter — Issue #11
This isn’t a recap of the weeks—it’s an attempt to understand the forces reshaping how we live, govern, and evolve.
We’re already halfway through November — and what a year it’s been! From LLMs becoming mainstream to AI-generated and AI-discovered drugs reaching patients, 2025 will be a year to remember. But it’s not over yet — exciting news continues to land.
The number of collaborations between big pharma and techbio companies keeps rising. Eli Lilly has been one of the most active players over the past few months, and this time it announced a $345 M partnership with XtalPi for AI-antibody design — the largest collaboration I’ve seen between a pharma and an AI-platform company. AI-antibody design is getting its own spotlight, especially after the success of AlphaFold, with numerous techbio and pharma players expanding in the space (Takeda × Nabla Bio, Generate:Biomedicines × Novartis).
Lilly, again proving how active it is in the AI space, also announced an AI drug-discovery licensing deal with Insilico, gaining access to the latter’s Pharma.AI platform to support target identification and molecule design, with upfronts, milestones, and royalties.
There’s also growing interest in CNS therapeutics, where one of the toughest challenges remains crossing the blood–brain barrier. Roche just announced a partnership with Manifold Bio to develop AI-guided shuttles for brain-drug delivery — a ~$55 M deal that highlights how AI is moving into tissue-specific delivery.
Samsung Epis Holdings, the Korean giant, launched a new biotech subsidiary, Epis NexLab, to develop next-generation biotech platforms bridging electronics and biopharma. The announcement doesn’t mention AI explicitly — but I’d bet it’s high on their agenda.
On the investment front, while venture funding has lagged this year, a recent analysis shows that big pharma has stepped in. Novo Holdings leads the way, followed by (not surprisingly) Lilly Ventures and Sanofi Ventures — the three now dominate private biotech rounds amid a slower VC market, each backing a dozen-plus deals in 2025.
Interesting Things in Research and Beyond
David Baker Lab’s recent Nature paper showcases RFdiffusion models designing antibodies with 0.3 Å accuracy — fully de novo scaffolds, no templates needed. A glimpse of AI as molecular engineer, not just pattern recognizer.