Today we're announcing a second, larger collaboration with @TakedaPharma — deploying our integrated AI + wet-lab platform broadly across their portfolio to design antibodies, multispecifics, and other next-generation biologics. It’s been incredibly fun working with the Takeda

"software investors" were obsessed with the idea of the idea of a tiny team reaching a billion+ run rate in enterprise saas because of "AI" this is a silly idea- gtm, sales, etc all track linearly with headcount. but it will become common in biology, nabla is ~20 ppl

Small focused teams are the way

another billion dollars goes to @SurgeBiswas @nablabio is a very, very cool company and one that i am permanently bullish on. i wrote about them back in janurary and am constantly happy i picked this specific ai antibody startup to write about:

Great write up in @endpts by @AndrewE_Dunn covering our second collaboration with Takeda, the drugs our JAM platform is unlocking, and where we go from here on our quest to make drug development designable

Great memories in the @geochurch lab pushing the limits of proteins, multiplexing , and AI with @glebkuz, @PierceOgdenJ, @ekelsic, now founders of @ManifoldBio and @Dyno_Tx @ElliotHershberg has a special ability to capture history and paint the future at the same time.

New blogpost on the latest in AI antibody design. Including some code to easily run Germinal and IgGM on modal! https://t.co/G2ZC6fzUxq

Earlier this year, I learned I have a pathogenic variant in BRCA2, which gives me a very high lifetime risk of breast cancer (55-69%), as well as an increased risk of ovarian cancer (12-29%) and pancreatic cancer (5-10%). 🧵1/ …

For a platform therapeutics company as technologically rich and fast-moving as Manifold Bio, a website is always just a snapshot in time. But I’m especially excited about this one. Our goal was to capture the scale of the tissue-targeted medicines opportunity and the uniquely

We’re launching AAV Apex Suite today — a new product line to support gene therapy drug developers and CDMOs. AAV gene therapies have incredible potential to cure disease, but cost and scale limitations have held them back. At @64xbio, we’re on a mission to solve these challenges

Excited to unveil Boltz-2, our new model capable not only of predicting structures but also binding affinities! Boltz-2 is the first AI model to approach the performance of FEP simulations while being more than 1000x faster! All open-sourced under MIT license! A thread… 🤗🚀

AI hype or reality - can we design antibodies from scratch? In my latest article for @GENbio, I unpack how arguments over terminology have clouded the industry’s shared mission of bringing better therapeutics to the clinic faster. https://t.co/mYxjnpL7qw

Scratch That? De Novo Antibody Design Enters the AI Drug Discovery Toolbox Debates over terminology clouded the mission of bringing better therapeutics to the clinic faster @nablabio @HHMINEWS @BakerLaboratory @abscibio @generate_biomed @Caltech https://t.co/WdWMqb9xuI

Today, we’re announcing the first major discovery made by our AI Scientist with the lab in the loop: a promising new treatment for dry AMD, a major cause of blindness. Our agents generated the hypotheses, designed the experiments, analyzed the data, iterated, even made figures

ICYMI we @Dyno_Tx unveiled 3 new capsids (delivering payloads to the Eye, CNS, and Neuromuscular) presented as oral spotlights in #ASGCT2025 this week! We also announced the first edition of the #GATC conference and our new frontiers program (if you are a startup in this space,

Nice piece from @xiaofei_lin @GENbio on our recent advance in de novo antibody design for GPCRs

Great write-up by @RobertFService at @ScienceMagazine on this work. Quotes from Wei Wang, David Baker, and Andrew Bradbury check it out: https://t.co/3wCH8jREON

Much more in our paper: https://t.co/QXt90xTK1I Our blog:

We're excited by these results and have more in development. Just as test-time reasoning is transforming language model capabilities, test-time scaling will be a fundamental scaling law for biomolecular design.

Activating antibodies are exceedingly rare. Yet by prompting JAM with just one de novo activator, we steered our system to create 300+ diverse new activators with substantially improved properties. One design, created in just weeks, activates the receptor with potency matching