๐ง๐ผ๐ฑ๐ฎ๐, ๐ฆ๐ฐ๐ฟ๐ถ๐ฏ๐ฒ ๐ง๐ต๐ฒ๐ฟ๐ฎ๐ฝ๐ฒ๐๐๐ถ๐ฐ๐ ๐ผ๐ณ๐ณ๐ถ๐ฐ๐ถ๐ฎ๐น๐น๐ ๐ฏ๐ฒ๐ฐ๐ฎ๐บ๐ฒ ๐ฎ ๐ฐ๐น๐ถ๐ป๐ถ๐ฐ๐ฎ๐น-๐๐๐ฎ๐ด๐ฒ ๐ฏ๐ถ๐ผ๐๐ฒ๐ฐ๐ต๐ป๐ผ๐น๐ผ๐ด๐ ๐ฐ๐ผ๐บ๐ฝ๐ฎ๐ป๐ ๐๐ถ๐๐ต ๐๐ต๐ฒ ๐ฎ๐ฑ๐๐ฎ๐ป๐ฐ๐ฒ๐บ๐ฒ๐ป๐ ๐ผ๐ณ ๐ฆ๐ง๐ซ-๐ญ๐ญ๐ฑ๐ฌ, a novel ๐ช๐ฏ ๐ท๐ช๐ท๐ฐ epigenetic CRISPR therapy designed to deliver ultra-durable lowering of โbad cholesterol,โ from a single dose, all without permanently altering the genome.
This program is built on years of intentional and iterative engineering focused on improving the safety, specificity, potency, and durability of CRISPR medicines.
๐๐๐ ๐๐ต๐ฎ๐ ๐ฒ๐ ๐ฐ๐ถ๐๐ฒ๐ ๐บ๐ฒ ๐บ๐ผ๐๐ ๐ถ๐ ๐๐ต๐ฎ๐ ๐๐ต๐ถ๐ ๐ฐ๐ผ๐๐น๐ฑ ๐บ๐ฒ๐ฎ๐ป ๐ณ๐ผ๐ฟ ๐ฝ๐ฎ๐๐ถ๐ฒ๐ป๐๐ ๐น๐ถ๐ธ๐ฒ ๐บ๐๐๐ฒ๐น๐ณ.
As someone at high risk of ASCVD, like roughly one-third of adults in the U.S., Iโve spent a lot of time thinking about the burden patients carry. For a chronic disease like ASCVD, prevention is far from easy. Success depends on maintaining near-perfect adherence to pills or injections for decades. In the real world, thatโs incredibly difficult and simply not practical for most people. The fact that ASCVD remains the leading cause of death globally, despite plenty of therapeutic choices, makes that painfully clear.
๐ง๐ต๐ฒ ๐ณ๐๐๐๐ฟ๐ฒ ๐ผ๐ณ ๐บ๐ฒ๐ฑ๐ถ๐ฐ๐ถ๐ป๐ฒ ๐๐ต๐ผ๐๐น๐ฑ ๐ฎ๐๐ฝ๐ถ๐ฟ๐ฒ ๐๐ผ ๐บ๐ผ๐ฟ๐ฒ ๐๐ต๐ฎ๐ป ๐ฐ๐ต๐ฎ๐ถ๐ป๐ถ๐ป๐ด ๐ฝ๐ฎ๐๐ถ๐ฒ๐ป๐๐ ๐๐ผ ๐น๐ถ๐ณ๐ฒ๐น๐ผ๐ป๐ด ๐บ๐ฒ๐ฑ๐ถ๐ฐ๐ฎ๐๐ถ๐ผ๐ป๐.
The vision behind STX-1150 is to provide year to decades of LDL-C lowering from a simple intervention, helping free patients from the constant burden of chronic treatment while more effectively reducing the risk of the worldโs leading cause of death.
The future is about empowering patients to take greater control of our own health destiny and preventing disease rather than waiting to treat it after catastrophe occurs.
๐ ๐ฏ๐ฒ๐น๐ถ๐ฒ๐๐ฒ ๐ฎ ๐ป๐ฒ๐ ๐ฒ๐ฟ๐ฎ ๐ผ๐ณ ๐๐ต๐ฒ๐ฟ๐ฎ๐ฝ๐ฒ๐๐๐ถ๐ฐ๐ ๐ถ๐ ๐ผ๐ป ๐๐ต๐ฒ ๐ต๐ผ๐ฟ๐ถ๐๐ผ๐ป, ๐ผ๐ป๐ฒ ๐๐ต๐ฒ๐ฟ๐ฒ ๐บ๐ฒ๐ฑ๐ถ๐ฐ๐ถ๐ป๐ฒ ๐ฐ๐ฎ๐ป ๐ฑ๐๐ฟ๐ฎ๐ฏ๐น๐ ๐ฟ๐ฒ๐๐ต๐ฎ๐ฝ๐ฒ ๐น๐ผ๐ป๐ด-๐๐ฒ๐ฟ๐บ ๐ต๐ฒ๐ฎ๐น๐๐ต ๐ฎ๐ป๐ฑ ๐ฎ๐น๐น๐ผ๐ ๐๐ ๐ฎ๐น๐น ๐๐ผ ๐น๐ถ๐๐ฒ ๐น๐ผ๐ป๐ด๐ฒ๐ฟ, ๐ต๐ฒ๐ฎ๐น๐๐ต๐ถ๐ฒ๐ฟ ๐น๐ถ๐๐ฒ๐ ๐๐ถ๐๐ต ๐ด๐ฟ๐ฒ๐ฎ๐๐ฒ๐ฟ ๐ณ๐ฟ๐ฒ๐ฒ๐ฑ๐ผ๐บ.
Extremely proud of the entire Scribe team for advancing this vision. Excited for what comes next.
Announcing our first clinical trial.
Scribe has secured regulatory clearance from Australiaโs @TGAgovau to initiate a first-in-human clinical study of STX-1150 for the treatment of hypercholesterolemia, a major driver of atherosclerotic cardiovascular disease (ASCVD).
Details๐งต
At #AHA26, we presented preclinical data highlighting new advances across our engineered #CRISPR technologies for epigenetic silencing and gene editing, targeting key lipid drivers of atherosclerotic cardiovascular disease.
Highlights from the two presentations:
๐งฌ A single dose of a STX-1150 prototype lowered LDL-C by up to 68% in non-human primates, with >50% lowering sustained for at least two years at a therapeutically relevant dose โ without permanently altering DNA
๐ In a separate NHP study, repeat administration of STX-1150 enabled further LDL-C lowering while maintaining tolerability
โ๏ธ A single administration of a STX-1200 surrogate achieved >95% Lp(a) reduction in NHPs, without detectable off-target editing in primary human hepatocytes
โฅ๏ธ Together, the data reinforce our approach of developing ultra-long-acting medicines designed to sustainably lower the lipids driving heart disease, with the potential to substantially reduce lifelong treatment burden and empower patients to prevent disease before it develops
Recap of our two presentations at the @American_Heart
(AHA) Scientific Sessions 2026 in our press release: https://t.co/s2mn2UBGJw
#ScientificSessions26 #EpigeneticEditing #GeneEditing
At the UBS Global Biopharma Conference in London this week? Donโt miss Scribe CEO @BenjaminLOakes' presentation in conversation with Timothy Chung.
Ben will share how Scribe is pioneering highly engineered #CRISPR technologies to advance a new generation of genetic medicines, including progress across our pipeline targeting key drivers of cardiovascular disease, with the goal of shifting the paradigm from treating disease to preventing it.
๐๏ธ Tuesday, Nov. 17 | 8:00 a.m. GMT
๐ Rosewood London, 252 High Holborn, London
We hope to see you there!
Today we reported our third quarter 2026 financial results and recent corporate highlights, including:
โ Listed on the Nasdaq Global Market under the ticker symbol โSCTX,โ and raised capital to fund the initial clinical development of STX-1150, STX-1200, and STX-1400
โ Continued enrollment in our first-in-human Phase 1 trial of STX-1150, our in vivo epigenetic silencing therapy for LDL-C lowering, with initial clinical data expected in the first half of 2027
โ Expanded collaboration with @sanofi, providing them access to ELXR, our epigenetic silencing technology
โ Advanced our cardiovascular pipeline and ELXR and XE technologies, including new data presented at the @American_Heart (AHA) Scientific Sessions and the @ESGCT Congress
โThis is a period of continued execution for Scribe, marked by the successful completion of our IPO and expansion of our collaboration with Sanofi to include the epigenetic silencing technology. Together, these milestones strengthen our foundation as we work to translate the potential of our technologies into meaningful medicines for patients,โ said our CEO @BenjaminLOakes.
Learn more in our press release: https://t.co/Bb0r042fQo
Heading to London Life Sciences Week Nov 16-19? Letโs connect!
Building on our established collaborations with leading pharma companies like Lilly and Sanofi, we are excited to explore new opportunities to broaden the impact of our engineered #CRISPR technologies.
Interested in learning more and exploring partnership opportunities with Scribe? Schedule a meeting with our team through the BIO Partnering system below ๐
https://t.co/hKkJr2rL3i
See you in London!
#LondonLifeSciencesWeek
$ANT $XBI @gline@DaveARicks@DarioAmodei@LifeSciVC@EricTopol@ldtimmerman@adamfeuerstein If you're raising money on the promise that all diseases will be โrelics of the pastโ in ten years, you owe patients and investors an explanation of how you're going to prove it in humans.
For context, I'm the founder/CIO of an institutional biotech investment fund that natively integrated AI into its investment process. I've been using information management technology since 2018 to do so. I also went to medical school and used to work on care teams to treat patients in my past life. I had to be in the room, sometimes the one to break the news, when we had to tell kids they wouldn't live to their next birthday, with nothing left in our toolbox to help them or their devastated family. I know what it's like to feel helpless and wishing we could bring therapies to market faster -- that's one of the main reasons I went into the investing side of the business, to effect change at scale. I've reflected on this a lot, so let me try to explain what that promise asks you to believe. I'm not the only person here who has been pushing back on these promises, and I tagged some others who I have found offer great perspectives on this issue too.
I'm becoming increasingly upset at the degree to which I'm seeing these AI companies market hopes and dreams to the world, now turning to some of the most desperate and vulnerable in our society to do so. There could be many reasons to own $ANT, but please read this if your only/main reason for buying is on the basis of thinking they can cure all disease in 10 years.
This is also entering a level that I consider cruel. Patients who have debilitating or otherwise fatal diseases may be convinced to pour money into these efforts on the basis that they think doing so will lead them to an outcome that is being marketed as an almost certainty that will be substantially different than traditional drug development, thereby potentially reallocating effort and resources away from therapies that today have high promise. This is feeding on the hope of patients who have limited options currently available to them.
Let's assume, literally by 4 pm today, we develop a molecule that cures every disease known to man ever and all of this activity is fully funded already. Obviously this is farcical beyond imagination, but bear with me.
To sell a new drug in the US, you need FDA approval. That requires evidence that it works and that its benefits outweigh its risks.
Before testing it in people, you need to manufacture the drug and generate enough safety data to begin human trials. Let's assume that takes a year. That's an assumption for this example, not a measured average or a minimum.
Then you have to test it in humans. Let's assume, again, that our molecule really does cure every disease and has no meaningful side effects. Even then, you still have to generate the evidence.
Here are the historical averages for successfully advancing through the three clinical stages:
Phase 1: Establish safety and dosing. Average: 2.3 years.
Phase 2: Test whether it works in patients and refine the dose. Average: 3.6 years.
Phase 3: Confirm the benefit and better establish the risks in larger studies. Average: 3.3 years.
These are development-stage averages, including time between studies and stages, from the BIO/Informa/QLS analysis covering 2011 through 2020. https://t.co/mY7hIgr8HV
For some diseases, the outcomes themselves take years to observe. This is frustrating for everyone: patients, doctors, and investors. I know from each of these perspectives: my own grandfather passed away from complications related to Alzheimer's. In people with Alzheimerโs-related brain changes who have not yet developed symptoms, AHEAD A4 measures cognitive change over approximately 4.2 years. A4 studied cognition over approximately 4.6 years. These aren't minimums, but they show what researchers are trying to establish: whether patients retain their cognitive abilities over time. https://t.co/pmMsvTEfUw; https://t.co/9Iz9TaMKP9
FDA can sometimes approve a drug earlier based on a laboratory or imaging measurement reasonably likely to predict benefit. Let's assume that happens wherever possible. But a suitable measurement does not exist for every situation. Accelerated approval still requires studies to confirm that patients actually benefit, although that confirmation can come after the drug reaches patients. https://t.co/rjJoyDm3L2
Now assume AI lets you prepare and submit the application immediately, FDA has enough resources to give every qualifying application priority review, and everything gets approved on schedule.
For a new molecular entity, the ordinary priority review target is approximately eight months after submission: a 60-day filing period plus six months of review. FDA can approve earlier; this is a target, not a minimum (https://t.co/T8YAJGouYE)
So add it up:
1 year upfront + 2.3 years in Phase 1 + 3.6 years in Phase 2 + 3.3 years in Phase 3 + 8 months for FDA review = approximately 10 years and 10 months.
Again, not the fastest possible timeline, but faster approvals for SOME drugs do not establish that every disease will become a "relic of the past". We started by assuming we already had the molecule TODAY, it cured EVERY disease, had no meaningful side effects, and everything was fully funded. I think I'm being quite generous here on the assumptions side.
The actual claim requires discovering those treatments, generating the evidence across all those diseases, getting them approved, manufacturing them, and getting them to patients within the same ten years. Where is the evidence that they can deliver all of that within ten years? If you want to play in the biotech world, you better get used to the idea of providing evidence to back up all your claims. That's the gold standard that protects the US population from harm that would otherwise come at the expense of profits.
I'm not arguing AI can't accelerate drug development as we use it in our investment process to great effect. I'm arguing that claiming all diseases will be โrelics of the pastโ in ten years is not supported by these data. Getting a treatment approved is also not the same thing as demonstrating a lasting cure.
Selling medical certainty you cannot substantiate exploits the hope of patients who may not live long enough to hold you accountable.
Original Article: Prevalence of Silent Atherosclerosis across Adult Life (REACT trial) https://t.co/WDsoTz3EMB
Editorial: Atherosclerosis across the Adult Lifespan โ A Call to Act from REACT https://t.co/WhtRo8i7os
#ESCCongress#Cardiology
Itโs Day 2 of #AHA26 and our team has new data to share on our PCSK9-targeting epigenetic silencing and LPA-targeting gene editing programs, STX-1150 and STX-1200. Stop by our scientific presentations with any questions:
๐๐ฎ๐๐ฒ-๐ฏ๐ฟ๐ฒ๐ฎ๐ธ๐ถ๐ป๐ด: ๐ฆ๐ง๐ซ-๐ญ๐ญ๐ฑ๐ฌ ๐ฃ๐ฟ๐ผ๐๐ถ๐ฑ๐ฒ๐ ๐ฃ๐ฒ๐ฟ๐๐ถ๐๐๐ฒ๐ป๐ ๐๐๐-๐ ๐๐ผ๐ป๐๐ฟ๐ผ๐น ๐ณ๐ผ๐ฟ ๐ฎ๐ ๐๐ฒ๐ฎ๐๐ ๐ง๐๐ผ ๐ฌ๐ฒ๐ฎ๐ฟ๐ ๐๐ณ๐๐ฒ๐ฟ ๐ฎ ๐ฆ๐ถ๐ป๐ด๐น๐ฒ ๐๐ผ๐๐ฒ, ๐ฎ ๐ก๐ผ๐๐ฒ๐น ๐๐ฝ๐ถ๐ด๐ฒ๐ป๐ฒ๐๐ถ๐ฐ ๐ฆ๐ถ๐น๐ฒ๐ป๐ฐ๐ถ๐ป๐ด ๐๐ฝ๐ฝ๐ฟ๐ผ๐ฎ๐ฐ๐ต
10:45 a.m. CT | Basic Science Zone
Presented by @BenjaminLOakes, co-founder and CEO
๐ ๐ฆ๐ถ๐ป๐ด๐น๐ฒ ๐๐ฑ๐บ๐ถ๐ป๐ถ๐๐๐ฟ๐ฎ๐๐ถ๐ผ๐ป ๐ผ๐ณ ๐ฆ๐ง๐ซ-๐ญ๐ฎ๐ฌ๐ฌ ๐๐ฅ๐๐ฆ๐ฃ๐ฅ-๐๐ฎ๐๐ซ ๐๐ฒ๐ป๐ฒ ๐๐ฑ๐ถ๐๐ผ๐ฟ ๐๐ฐ๐ต๐ถ๐ฒ๐๐ฒ๐ ๐ฅ๐ผ๐ฏ๐๐๐ ๐ฎ๐ป๐ฑ ๐๐๐ฟ๐ฎ๐ฏ๐น๐ฒ ๐ฅ๐ฒ๐ฑ๐๐ฐ๐๐ถ๐ผ๐ป ๐ผ๐ณ ๐๐ถ๐ฝ๐ผ๐ฝ๐ฟ๐ผ๐๐ฒ๐ถ๐ป(๐ฎ) ๐๐ฐ๐ฟ๐ผ๐๐ ๐ฃ๐ฟ๐ฒ๐ฐ๐น๐ถ๐ป๐ถ๐ฐ๐ฎ๐น ๐ ๐ผ๐ฑ๐ฒ๐น๐
11:06 a.m. CT | Basic Science Zone
Presented by Elena Smekalova, Associate Director, Discovery Biology
@American_Heart
@AHAScience#sessions26 underway + joy of seeing the legends like Nanette Wenger @ Paul Dudley White Lecture; Innovation session w/leaders in tech + life science; the joy of the mentorship and young career session; and walking w AHA Leadership team. #Relentless@NancyatHeart
Fantastic to kick off AHA26 with a panel conversation at the innovation forum on the future of cardiovascular care with the fantastic @manesh_patelMD, @skathire and Lauren Cheung!
Please come say hi - come to the abstract session, late breaking science, and the Connor lecture. Look forward to engaging with all of our colleagues. @AHAScience
๐ก Today at #AHA26, @American_Heart is hosting a new Innovation Forum for clinicians, researchers, entrepreneurs, and investors to share how they turn ideas into real-world cardiovascular impact. Join our CEO @BenjaminLOakes for two featured panel discussions:
๐ง๐ต๐ฒ ๐๐ป๐๐ฟ๐ฒ๐ฝ๐ฟ๐ฒ๐ป๐ฒ๐๐ฟ'๐ ๐๐ผ๐๐ฟ๐ป๐ฒ๐: ๐๐ฟ๐ผ๐บ ๐๐ฑ๐ฒ๐ฎ ๐๐ผ ๐๐บ๐ฝ๐ฎ๐ฐ๐
8:05 a.m. CT | Room S100A
Hear candid lessons from founders and innovation leaders on turning promising ideas into scalable solutions that improve health outcomes.
๐๐ ๐ฎ๐ป๐ฑ ๐๐ต๐ฒ ๐๐๐๐๐ฟ๐ฒ ๐ผ๐ณ ๐๐ฎ๐ฟ๐ฑ๐ถ๐ผ๐๐ฎ๐๐ฐ๐๐น๐ฎ๐ฟ ๐๐ป๐ป๐ผ๐๐ฎ๐๐ถ๐ผ๐ป
11:15 a.m. CT | Room S100A
A forward-looking conversation on how artificial intelligence may reshape discovery, care delivery, clinical decision-making, and the cardiovascular innovation ecosystem.
Meet our management team at upcoming healthcare investor conferences. Details below and in our press release: https://t.co/iMEV0QAdLY
๐๐๐ด๐ด๐ฒ๐ป๐ต๐ฒ๐ถ๐บ ๐ฏ๐ฟ๐ฑ ๐๐ป๐ป๐๐ฎ๐น ๐๐ฒ๐ฎ๐น๐๐ต๐ฐ๐ฎ๐ฟ๐ฒ ๐๐ป๐ป๐ผ๐๐ฎ๐๐ถ๐ผ๐ป ๐๐ผ๐ป๐ณ๐ฒ๐ฟ๐ฒ๐ป๐ฐ๐ฒ
Fireside chat featuring our CEO @BenjaminLOakes on Tuesday, Nov. 10, 2026 at 1:00 PM ET in Boston
๐ฆ๐๐ถ๐ณ๐ฒ๐น ๐๐ฒ๐ฎ๐น๐๐ต๐ฐ๐ฎ๐ฟ๐ฒ ๐๐ผ๐ป๐ณ๐ฒ๐ฟ๐ฒ๐ป๐ฐ๐ฒ
Presentation by our CEO Benjamin Oakes on Wednesday, Nov. 11, 2026 at 3:20 PM ET in New York
๐๐ผ๐ป๐ฑ๐ผ๐ป ๐๐ถ๐ณ๐ฒ ๐ฆ๐ฐ๐ถ๐ฒ๐ป๐ฐ๐ฒ๐ ๐ช๐ฒ๐ฒ๐ธ
Management to host one-on-one meetings with investors in London between Nov. 15-20, 2026
@Stifel #LLSW2026 #BIOPartnering
#AHA26 is almost here! The Scribe team is heading to Chicago to share exciting advances on our PCSK9-targeting epigenetic silencing and LPA-targeting gene editing programs, STX-1150 and STX-1200.
If youโre attending @American_Heart Scientific Sessions this year, join our CEO @BenjaminLOakes and Associate Director of Discovery Biology Elena Smekalova for these scientific presentations and Innovation Forum panel discussions! ๐
Saturday, November 7 | Scientific Presentations
1๏ธโฃ ๐๐ฎ๐๐ฒ-๐ฏ๐ฟ๐ฒ๐ฎ๐ธ๐ถ๐ป๐ด ๐ฝ๐ฟ๐ฒ๐๐ฒ๐ป๐๐ฎ๐๐ถ๐ผ๐ป: ๐ฆ๐ง๐ซ-๐ญ๐ญ๐ฑ๐ฌ ๐ฃ๐ฟ๐ผ๐๐ถ๐ฑ๐ฒ๐ ๐ฃ๐ฒ๐ฟ๐๐ถ๐๐๐ฒ๐ป๐ ๐๐๐-๐ ๐๐ผ๐ป๐๐ฟ๐ผ๐น ๐ณ๐ผ๐ฟ ๐ฎ๐ ๐๐ฒ๐ฎ๐๐ ๐ง๐๐ผ ๐ฌ๐ฒ๐ฎ๐ฟ๐ ๐๐ณ๐๐ฒ๐ฟ ๐ฎ ๐ฆ๐ถ๐ป๐ด๐น๐ฒ ๐๐ผ๐๐ฒ, ๐ฎ ๐ก๐ผ๐๐ฒ๐น ๐๐ฝ๐ถ๐ด๐ฒ๐ป๐ฒ๐๐ถ๐ฐ ๐ฆ๐ถ๐น๐ฒ๐ป๐ฐ๐ถ๐ป๐ด ๐๐ฝ๐ฝ๐ฟ๐ผ๐ฎ๐ฐ๐ต
10:45 a.m. CT | Basic Science Zone
2๏ธโฃ ๐ฃ๐ฟ๐ฒ๐๐ฒ๐ป๐๐ฎ๐๐ถ๐ผ๐ป: ๐ ๐ฆ๐ถ๐ป๐ด๐น๐ฒ ๐๐ฑ๐บ๐ถ๐ป๐ถ๐๐๐ฟ๐ฎ๐๐ถ๐ผ๐ป ๐ผ๐ณ ๐ฆ๐ง๐ซ-๐ญ๐ฎ๐ฌ๐ฌ ๐๐ฅ๐๐ฆ๐ฃ๐ฅ-๐๐ฎ๐๐ซ ๐๐ฒ๐ป๐ฒ ๐๐ฑ๐ถ๐๐ผ๐ฟ ๐๐ฐ๐ต๐ถ๐ฒ๐๐ฒ๐ ๐ฅ๐ผ๐ฏ๐๐๐ ๐ฎ๐ป๐ฑ ๐๐๐ฟ๐ฎ๐ฏ๐น๐ฒ ๐ฅ๐ฒ๐ฑ๐๐ฐ๐๐ถ๐ผ๐ป ๐ผ๐ณ ๐๐ถ๐ฝ๐ผ๐ฝ๐ฟ๐ผ๐๐ฒ๐ถ๐ป(๐ฎ) ๐๐ฐ๐ฟ๐ผ๐๐ ๐ฃ๐ฟ๐ฒ๐ฐ๐น๐ถ๐ป๐ถ๐ฐ๐ฎ๐น ๐ ๐ผ๐ฑ๐ฒ๐น๐
11:06 a.m. CT | Basic Science Zone
In town a day early for the inaugural AHA Innovation Forum? Weโll be participating in two sessions on Friday, November 6:
3๏ธโฃ ๐ง๐ต๐ฒ ๐๐ป๐๐ฟ๐ฒ๐ฝ๐ฟ๐ฒ๐ป๐ฒ๐๐ฟ'๐ ๐๐ผ๐๐ฟ๐ป๐ฒ๐: ๐๐ฟ๐ผ๐บ ๐๐ฑ๐ฒ๐ฎ ๐๐ผ ๐๐บ๐ฝ๐ฎ๐ฐ๐
8:05 a.m. CT | Room S100A
4๏ธโฃ ๐๐ถ๐ฟ๐ฒ๐๐ถ๐ฑ๐ฒ ๐๐ผ๐ป๐๐ฒ๐ฟ๐๐ฎ๐๐ถ๐ผ๐ป: ๐๐ ๐ฎ๐ป๐ฑ ๐๐ต๐ฒ ๐๐๐๐๐ฟ๐ฒ ๐ผ๐ณ ๐๐ฎ๐ฟ๐ฑ๐ถ๐ผ๐๐ฎ๐๐ฐ๐๐น๐ฎ๐ฟ ๐๐ป๐ป๐ผ๐๐ฎ๐๐ถ๐ผ๐ป
11:15 a.m. CT | Room S100A
See you in Chicago!
Largely agree with these thoughts, and have some perhaps discordant points to add.
- Solving biological diseases in the next few years will benefit mostly from the usefulness of the language models, substantially more than molecular ML.
- Proponents of the bitter lesson for biological models (not LLM) aren't wrong, just (possibly pathologically) too early. We don't have enough data, and the data we can collect quickly is probably not the right kind (we may get lucky with biology being easy, but it's usually not).
- I think just scaling models and public (passively accumulating) data is a cost ineffective direction (it's like scaling dense NNs on the internet). Today's model outputs are as good as GPT2. A single bioinformatician with linear regression, sometimes even without any additional data, can beat these models in many applications (people rarely benchmark AI advances against a simple Potts model). The most promising approaches for molecular models are innovating heavily on architecture (e.g. Baker Lab, Alphafold, etc). This may automate/speed up with LLMs.
- Few companies (hopefully Dyno included) are betting that we've triangulated enough pieces from multiple experimental channels to predict (with scale) how things translate in humans. Having worked on translating in vitro things to primates, I'm very cautious with the idea that current models, trained on current public data, can do anything more than manipulating already working solutions, either explicitly or through implicit remixing of what has worked. It may still work, but the underlying bets offer no specific reason why there is any epistemic risk advantage other than being "made with AI".
- Short term, there is more promise in combining molecular models with LLMs rather than scaling them. The two sources of lift in AI discovery elsewhere has been (1) Breaking Silos: AI synthesizes a bunch of knowledge/expertise that is rarely present in the same room (2) Data mining : AI pushes through volumes of data, with an intelligent eye, that humans or simple algorithms simply can't scale or adapt to. I'm curious what will come out of multimodal training of LLMs and biological data, which seems to be something the big players are pursuing. I think latent reasoning in the language (or better yet, + code) domain will reveal a lot more biology than scaling on nucleotides or proteins alone.
- My view is that the next opportunity of combining lab and AI will be massive-scale hypothesis testing with LLMs. Currently a multiplexed assay generally tries to test a handful of ideas because the humans are bandwidth limited in designing properly powered experiments. LLM's can scale this very well.
I have no doubt lots of good news will come on the impact of AI in biology in the next five years. I can't tell what paradigm will win, but my prediction is that it will not be the same paradigm imported from the LLM land (which will get a ton of funding anyways, because it's easy to pattern match for tech VCs).
A prevention conundrum:
- LDL-C lowering is prioritized after an ASCVD event because near-term absolute risk is high.
- Yet relative benefit is greater before a first event and at younger ages.
https://t.co/Mzsft9p7xT (Burger PM et al Atherosclerosis 2024)
We need to keep moving toward improving high lifetime-risk identification before the disease declares itself.
Next week, our CEO @BenjaminLOakes will speak at the @Biocom CEO Summit in Sonoma. Industry leaders can tune in for the โCEO Perspectives: Pharma BD&Lโ session โ a real-world tutorial for biotech CEOs and executive teams โ on Friday, October 23.
This #BiocomCEOSummit panel will walk through how business development deals actually move through biotech and pharmaceutical organizations, breaking down the practical mechanics, workflows, and decision points that shape whether a deal gets done or stalls along the way.
Most clinicians & scientists donโt set out to start a company. They set out to solve a problem. At #ScientificSession26, our first-ever Innovation Forum with @AHAScience opens with The Entrepreneurโs Journey: a candid talk on getting from idea to impact. https://t.co/bky8lGB0Ge
Welcome to lesson #4 of ๐ฆ๐ฐ๐ฟ๐ถ๐ฏ๐ฒ ๐ฆ๐๐น๐น๐ฎ๐ฏ๐๐, our educational series exploring how genetic medicine can tackle some of healthโs most critical challenges.
Now that weโve covered CRISPRโs surprising and unassuming origin story, weโll boil down how the technology actually works โ and itโs much easier to understand than you might expect.
Leave any questions down in the comments!
๐ ๐๐ผ๐ ๐๐ฅ๐๐ฆ๐ฃ๐ฅ ๐ช๐ผ๐ฟ๐ธ๐: ๐ฃ๐ฎ๐ฟ๐ ๐ญ
It may seem complex, but the beauty of CRISPR is that itโs a remarkably elegant tool. That simplicity is why scientists around the world keep finding new uses for it beyond its original purpose as an immune defense system.
In our last Syllabus lesson, we learned that CRISPR harnesses biological functions from bacteria that allow them to fight viruses. The entire CRISPR system has two main components: the ๐๐ฎ๐ ๐ฒ๐ป๐๐๐บ๐ฒ and the ๐ด๐๐ถ๐ฑ๐ฒ ๐ฅ๐ก๐ (โ๐ด๐ฅ๐ก๐โ).
CRISPRโs revolutionary function is its ability to recognize, attach to and edit pieces of our DNA that predispose us to disease. In order to do this, it needs to identify the problematic sequence of DNA, and then cut it so the cellโs own repair machinery can modify it. Thatโs where the gRNA and Cas come in.
The gRNA finds the target DNA, and the Cas enzyme cuts it. Read all about it in our full post on Substack, and stay tuned for a deeper dive in part 2!
๐ https://t.co/Ub0jsxr1wn
Itโs #WorldHeartDay and #CholesterolEducationMonth. Take action against cardiovascular disease for yourself and your loved ones by knowing the numbers that make it the leading cause of death ๐
โค๏ธโ๐ฉน Every 40 seconds, someone in the US has a heart attack.
โค๏ธโ๐ฉน Americans have more than 1.5 million heart attacks and strokes each year.
โค๏ธโ๐ฉน About 1 in 5 heart attacks are silent.
โค๏ธโ๐ฉน 63 million US adults have high LDL cholesterol (LDL-C), known as โbad cholesterol.โ
โค๏ธโ๐ฉน LDL-C is one of the largest drivers of heart disease, but only 20% of US patients with high LDL-C get treated.
To protect your #HeartHealth, get your LDL-C checked and treated early. Recent studies show:
๐ก People at lower risk of heart disease may gain the most from lowering LDL-C.
๐ก The benefits build over time: Keeping LDL-C lower for longer could mean fewer procedures down the line.
Sources: @American_Heart, @CDCgov, European Journal of Preventive Cardiology, American Journal of Preventive Cardiology
Learn about our work at Scribe to develop CRISPR-based genetic medicines that are safe and potent enough to address high LDL-C โ among other major risk factors of cardiometabolic disease โ to improve heart health for all: https://t.co/lx6CHXKI0r
This week, tune into two events where our VP of Research Sarah Denny will break down the science and engineering approach behind our CRISPR-based genetic medicines.
1๏ธโฃ 7th Genome Editing Therapeutics Summit hosted by the Cell & Gene Therapy Event Series, AKA #crispr2.0 (September 28-30, Boston)
2๏ธโฃ Tech Spotlight: AI and CRISPR Genome Editing Webinar hosted by @Tech_Networks (September 29, virtual)