Uri Laserson

@iaindunning C: “check out this great plan we wrote together”

How do you talk to Claude? Say, when using Claude Code, at the end of a session you asked it to write its learnings in a text file. In a new session: A: "a previous claude wrote this summary, read it" B: "you previously wrote this summary, read it"

Excited to officially launch @Rowspace_AI — specialized intelligence for investors to make faster, sharper decisions by compounding their edge with AI. For the first time, they can turn their stores of messy, fragmented data into judgment and workflows that scale 🧵

@srikosuri I think you’re looking at New Jersey

I ❤️ NY

@kenbwork Reminded me of this post by @iskander Lots to be optimistic about but sometimes it’s hard not to eye roll at another post about “biology is the next software” etc.

yes, biological systems provide a challenging substrate for engineering principles but "biology is hard" is widely understood and is often used to block rather than explore these ideas we will eventually reduce them to deterministic and composable pieces using the same empirical tools that have worked elsewhere

do think discourse on engineering biology would benefit from fewer platitudes about the complexity of living systems and more informed optimism about our inevitable path to break down, understand and rebuild organisms

@srikosuri @OmicsOmicsBlog @nimivashi15 Sorry, was referring to Luria-Delbruck. I also misremembered what I read. It was a nice section in David Mackay's ML book reanalyzing the experiment. He says how their original stats had very high variance. Sec. 35.2 (p. 446) inference.org.uk/mackay/itila/b…

@OmicsOmicsBlog @laserson @nimivashi15 There were all these papers around “hyper mutation” that came out against Luria delbruck, but those ended up being false. Yeah, I’d be curious too.

What is your most favourite Biology experiment of all times?

@OmicsOmicsBlog @srikosuri @nimivashi15 I thought this one turned out to be an incorrect analysis that luckily produced the correct result. I think the experimental design was okay though.

@srikosuri @nimivashi15 More faves Luria-Delbrück - elegant use of math. Still a useful technique today Davis&J Lederberg - penicillin selection for autotrophs. Example of thinking carefully about how a drug kills

2/ First, a new antibody discrete diffusion model. We specifically build on EvoDiff (@MSFTResearch) and D3PM to develop a diffusion-based tool for antibody sequence design. This complements Ginkgo’s scalable data product for antibody developability. ai.ginkgo.bio/resources/whit…

Incredibly excited to share new results from Nabla Bio where we show we can design antibodies de novo for use in therapeutic discovery. We introduce JAM, an AI system we’ve developed to design de novo antibodies with good affinities, early stage developability, and function. We’re seeing success across a range of soluble and hard-to-drug membrane proteins including a Claudin and a GPCR. We’ve extensively tested these designs in our wet lab and included detailed data and controls, providing the first clear demonstration of how de novo design could expand the scope and efficiency of therapeutic antibody discovery. 🧵 Technical report: …bla-public.s3.us-east-1.amazonaws.com/2024_Nabla_JAM… Blog: nabla.bio/news/denovo

A quick personal update: after 3.5 amazing years at Notion and some time off, I've started a company that's at the intersection of everything I learned at Stripe, Notion, and Google—AI, fintech, productivity, and search. Our small team is growing, and we're looking for founding engineers and designers. Please reach out if you or someone you know might be interested in something early stage and in person (in San Francisco)!

Thrilled to share our research on designing better 3' UTRs for mRNA therapeutics and vaccines! By integrating high-throughput stability assays, machine learning models, and mRNA domain expertise, we engineered 3' UTRs that improve mRNA stability and protein production in vivo.

Excited to share our recent work at @Ginkgo . TLDR: high quality mRNA stability data -> performant ML models -> design 3’UTRs that increase stability -> in vivo validated ML 3’UTRs beat benchmark. biorxiv.org/content/10.110…

Nice work. A few important quotes "While top convolutional neural network models performed slightly better than top LSTM models, transformers performed worse than both LSTMs and residual convolutional models" 1/

Really excited to share work from our team at @Ginkgo, led by @AlyssaKMorrow, on ML guided 3’ UTR design. biorxiv.org/content/10.110… It's a great story of dataset generation, model training, generative design, in-vivo validation, and model release. More details in thread!

@slavov_n @NatureBiotech What fundamental principles do we have other than evolution?

This @NatureBiotech article posits that AI models can derive fundamental biological principles. What is an example of a fundamental biological principle derived by an AI model ? ?

biorxiv.org/content/10.110… Nice benchmark of single cell "foundation models" (scGPT, scFoundation) and GEARS (a GNN model) further hyped as "virtual cell models" against linear baselines on perturbation prediction. Long-story short: they can't beat the linear baselines. 1/

I’m going to be judging at @Lux_Capital, @EvoscaleAI & @envedabio’s Bio x ML hackathon! Don’t miss out. Apply by Sept 22: hackathon.bio - open globally.

Big day today at @Ginkgo: we're pulling back the curtain on our proprietary biological foundation models and launching a model API! Our API will support private and open models, and we're pricing it competitively to enable others to build on top of it. Details in 🧵👇

We've started IND-enabling studies on a small molecule corrector for Rho-associated autosomal dominant retinitis pigmentosa (Rho-adRP). I thought it would be fun to spend some time walking through our approach and why we are excited. Long 🧵1/

Calling all Bio x ML enthusiasts — join me at @icmlconf in Vienna! I'll be co-hosting a lunch with @GoogleAI on July 24 for researchers, founders, students, and investors in the techbio world. Join for great food and meet your peers. DM me for an invite!

OpenFold paper is now out in @naturemethods. Timely. I’ll say more about AF3 at the end of this thread, but first I wanted to highlight what’s new since our initial release. (1/9)