Andrew Hui

Testing a longer-form research format away from the usual slide decks on bio-data for model training. Putting some highlights in the thread. demandrew.com/publish/post/1…

@ShahramSN but the whole business is built on retail hype??

From hype to humdrum assets? :) "Recursion’s comprehensive platform stands out among the surging crowd of other AI biotechs, Khan said, because the focus is on products, not new models. Recursion was the first biopharma to build a supercomputer with Nvidia, she noted; she now wants to be the first to use AI to put a new drug candidate into a partner’s pipeline." After a decade of hype, Najat Khan is bringing Recursion back down to earth fiercebiotech.com/biotech/after-…

chat is this chatgpt?

3 years ago... how we feeling about proteomics?

@AppleHelix is it corporate strategy or marketing strategy? I would argue rxrx is retail-forward and gets rewarded with a better valuation while sdgr is perceived as boring legacy-saas both businesses arguably have very similar corporate strategies and have pivoted to asset development

I think the fact that $SDGR's market cap is 50% of $RXRX is embarrassing. There needs to be a change in corporate strategy

underrated point of the ajax story is that schrodinger owned a sizable equity stake - as with structure therapeutics (3.2B), Morphic (acquired 3.2B), among others... If they were a VC fund they would be taking victory laps, unfortunately they're a SaaS business...

@LifeSciLad offer still drastically under cash on balance sheet though, would be brutal if it goes through

Radoff-JEC Group upped their bid to buyout $SEER to $2.35. The letter has some painful truths in it (stock down 90% since IPO, reported losses of $465M). Surprised the board hasn't even responded to them investor.seer.bio/static-files/e…

@LifeSciLad >1b market cap, big day for life science tooling?

$ALMR opens around $23 😮

@two_natural lots of people chasing this opportunity so you would have to assume so, there are preclinical "techbio" competitors like Recursion and Genesis Therapeutics going after the same target as well

Is part of the bet here that PI3Ka inhibitors can move beyond HR positive HER2 negative patients? Even if you can deliver much better tolerability, seems hard to make sense of paying 2 billion upfront for a candidate in phase 1/2 trials if the addressable patient population is the same as what Itovebi is addressing. Especially so given Eli Lilly & Relay's efforts. What am I getting wrong?

@JulianKlymochko That’s why they have a pretty website 🙂

First time I've seen an "activist" campaign on a stock with 0 votes

@buccocapital Peloton up because Anthropic enables better workouts..?

“The Market” is just a guy staring a two screens. One has Truth Social. The other is Anthropic's blog In front of him are 5 buttons that say: Software, Semis, European Defense, Energy, and Gold. Trump or Anthropic post and he hits a button to make those stocks move +5% or -5%

@mkoeris Similar findings from this recent preprint as well arxiv.org/pdf/2603.05532

I trust what DL writes

@MelindaBChu1 the ones that do pivot to drug development end up crowding the same targets anyways (e.g. PI3Ka mutant inhibitors)

This blew my mind. 🤯 I always thought based on media / hype that with new AI /Tech Bio companies were that you have idea, start the project, get list of compounds, synthesize them, ➡️ automated lab ➡️ Phase 1 human study, and then “cure ALL diseases in 5-10 years.” That’s what people say. 🚨 So I figured most TechBio would have to pivot to drug development b/c that’s wheee th real value in Biotech is and there’s not enough customers (other than Big Pharma) for all these SaaS. 🚨BUT looking into this, I learned that most companies only do 1 small step in drug design, so in theory you would need 6-10 of them to actually get to compound synthesis. 🚨This means 2 things: 1.). Means they can’t really pivot to developing drug assets b/c it’s only 1 small step of process. 2.) Reinforces my view that there aren’t enough customers for each SaaS 🌟 Check out this technical paper where I go into this in more detail and demonstrate my multi-model orchestration approach.

@packyM had a sick lab tour last year from them, very cool work

I am so clucking eggcited that Neion Bio is finally coming out of its shell. 🐣 Today, pharma uses Chinese hamster ovary (CHO) cells to produce biologics like Keytruda and Humira in huge stainless steel bioreactors. Merck spent $1B on a single Keytruda facility. In early 2024, Elliot Hershberg wrote that chicken eggs are much more efficient bioreactors. They run on grain and water, produce six grams of protein per unit, and we already farm them at massive scale. Sam Levin, who we previously backed at Melonfrost, and Dimi Kellari explored the frontiers of this research, which they could get their arms around because the cutting edge stuff is happening in a very small number of labs around the world. They realized that the time was right to build a company that uses nature's bioreactors to produce drugs at a fraction of the cost. In today's NYT article, Sam predicts that the cost can be 1/10th or even 1/100th of the current cost, and that just 3,900 hens could meet global Humira demand. I'm proud to back Sam, Dimi, and the Neion Bio team as they work to hatch the balk of the world's biologics and dramatically lower the cost to produce critical drugs, right here in NYC. And I'm sure they're happy that I can share my chicken puns with all of you instead of just replying to investor updates with them. Check out what they're up to in the NYT article below.

Biotech funding (US/China/Global)

China biotech funding

TL;DR AI performance is tied to scaling laws, which means more data going in, hence, "data problems". The next wave of value creation will be tied to new sources of data that explore new biological space. But the issue is that this data isn't readily available for the industry.

8/ Scaling generative models drives the cost of hypothesis generation to zero. However, the key is still wet lab validation. The gap between hypothesis generation and wet lab validation creates "speculation debt" that still needs to be bridged

Testing a longer-form research format away from the usual slide decks on bio-data for model training. Putting some highlights in the thread. demandrew.com/publish/post/1…