
Chief Mid-Curver | Prev: @Shitcoinmaxi_ | Everything posted is NFA | DYOR Founded @SmResearchLab
Joined January 2025
- Tweets3.5K
- Following885
- Followers2.9K
- Likes3.5K
Most people don't understand this:
AI was never meant to be just ChatGPT. It was meant to be something more (explained below).
The next, and final, part of the AI trade is here...
SM retweeted
$NVDA put "accelerate drug discovery" in its earnings tonight.
Most people have no idea how connected it is to $IAM.
It helped build Iambic's core AI. NeuralPLexer, the model that predicts how a drug locks onto its target, came out of a multi-year collaboration with its researchers. Their names are on the research paper.
It co-owns the IP. One NeuralPLexer patent family is owned by Iambic, Caltech and the chip giant together.
It bought the stock. Twice. In 2023 and again in 2026, holding over 5% at both rounds.
And it all runs on their hardware. NeuralPLexer predicts a drug–protein structure in seconds on a single GPU. Enchant v2 was trained on B200s.
The biggest company in the world helped build the engine, owns part of the patents and owns part of the company.
SM retweeted
Something else to add which I find interesting.
$NVDA completely exited $RXRX, the AI drug discovery stock it backed in 2023.
A few months later, it bought more $IAM.
Now look at how small NVIDIA’s public portfolio is. Its latest 13F shows just eight US-listed stocks, led by SpaceX, Intel and CoreWeave. Iambic listed in October with pre-IPO holders locked up until April, so Iambic is the ninth.
Out of a tiny public portfolio, Iambic is very likely NVIDIA’s biggest AI drug discovery bet. A ~$700M company sitting next to SpaceX and Intel.
$NVDA put "accelerate drug discovery" in its earnings tonight.
Most people have no idea how connected it is to $IAM.
It helped build Iambic's core AI. NeuralPLexer, the model that predicts how a drug locks onto its target, came out of a multi-year collaboration with its researchers. Their names are on the research paper.
It co-owns the IP. One NeuralPLexer patent family is owned by Iambic, Caltech and the chip giant together.
It bought the stock. Twice. In 2023 and again in 2026, holding over 5% at both rounds.
And it all runs on their hardware. NeuralPLexer predicts a drug–protein structure in seconds on a single GPU. Enchant v2 was trained on B200s.
The biggest company in the world helped build the engine, owns part of the patents and owns part of the company.
SM retweeted
Another piece of outside validation for how undervalued $IAM is.
Oxford researchers mapped where AI drug discovery is heading, in npj Drug Discovery (Nature Portfolio).
They call the next era TechBio 3.0: AI designs the drug, a robot lab tests it, and every result goes straight back into the model.
Their example:
“One actionable example of the TechBio 3.0 thesis is Iambic Therapeutics.”
Only three companies made their comparison table:
Iambic: yes on all 4 criteria
Insilico: yes on all 4
Isomorphic: partial, pipeline undisclosed
What the market pays for each:
Isomorphic: ~$50B (reported)
Insilico: ~$2.8B
$IAM: ~$0.7B, with ~$455M of that in cash
Same table. Isomorphic is valued at ~70x Iambic.
Published 9 Oct, six days before Iambic’s IPO.
Nice to see Oxford researchers back what I’ve thought for a while: Iambic is level with or ahead of its comps, at a fraction of the price.
NFA, DYOR
SM retweeted
A lot surfaced on $IAM today. Here’s all of it in one place.
1. New data, out this morning
Iambic unveiled its next two AI-discovered drugs at the EORTC-NCI-AACR symposium in Barcelona. Both wholly owned.
IAM-C1 goes after the drug class behind Ibrance, Verzenio and Kisqali, the standard first treatment for the most common type of advanced breast cancer.
• Tumour already resistant to Ibrance-type drugs: 78% vs Ibrance’s 24%
• Vs the lead drug in Roche’s $850M CDK deal: 97% vs 76% in CDK2-driven tumours, 94% vs 83% in CDK2/4-driven
• White blood cell loss, this class’s biggest side effect: 13% vs Ibrance’s 69%
One pill matching what Pfizer needs two drugs to do.
IAM217 is the brain drug.
• ~90% shrinkage of breast tumours growing inside the brain
• Volastra’s rival, already in the clinic, at twice the dose: tumours kept growing
2. Oxford researchers
A review in npj Drug Discovery (Nature Portfolio) mapped the next era of AI drug discovery, TechBio 3.0: AI designs the drug, robots test it, results go straight back into the model.
Their example: “One actionable example of the TechBio 3.0 thesis is Iambic Therapeutics.”
Only three companies made their comparison table. Iambic got a yes on all four criteria. Isomorphic, reportedly worth ~$50B, got a partial and no disclosed pipeline.
3. UK government compute
Isambard-AI is Britain’s £225M national AI supercomputer, running 5,448 NVIDIA chips.
Iambic was awarded time on it to build NeuralPLexer4, the next version of its model that predicts how a drug locks onto its target. NeuralPLexer3 was published at NeurIPS.
At this year’s Isambard Summit, the UK’s AI minister opened proceedings. Next keynote on the programme: Iambic’s co-founder.
4. $NVDA
In yesterday’s earnings, NVIDIA said it’s helping “accelerate drug discovery.”
Look at its drug discovery bets. It sold every $RXRX share. It’s been an Iambic shareholder since 2023, co-owns a NeuralPLexer patent family, and its Iambic stake is bigger than its only other biotech holding.
NFA, DYOR
$IAM THESE RESULTS ARE INSANE!
Conclusions from today's data:
1. IAM217 is a potent and differentiated KIF18A inhibitor with selective activity in tumor cells with chromosomal instability
2. IAM217 is well tolerated and exhibits dose-dependent plasma and tumor levels that correlate well with tumor regression
3. IAM217 demonstrates robust anti-tumor activity across a broad spectrum of breast and ovarian xenograft models
4. IAM217 reduces tumor burden in an intracranial mouse model of brain metastasis
5. IAM217 in combination with an approved microtubule inhibitor demonstrates synergy in ovarian and TNBC cell lines in vitro
6. These data support the continued development of IAM217 as a therapeutic strategy for CIN-high cancers, including in patients with brain metastases. IND submission accepted.
BREAKING: $IAM just released new data on its second cancer drug, IAM217.
Data shows that IAM217 reached tumors in the brain and shrank them by about 90% in mice. A competing drug from another company did not shrink them at all in the same test.
When cancer spreads to the brain, many drugs stop working. Most medicines cannot get past the brain's natural barrier. That is why brain tumors are one of the hardest problems in cancer, and where patients most need new options.
What the data shows
1. It gets into the brain. The rival drug does not.
Both drugs shrank tumors in the body. Only IAM217 shrank the tumors in the brain. Same type of drug, very different result. Iambic's drug simply reaches places the competitor cannot.
2. It stays there all day.
Drug levels in the brain fluid stayed high enough to work for more than 24 hours. That points to one pill a day.
3. It goes after cancer cells, not healthy ones.
IAM217 hit cancer cells at a very low dose. It took roughly 300 times more to harm the cells that make blood. That matters, because low blood counts are a common side effect of cancer drugs.
4. The body absorbs all of it.
In mice, 100% of the pill made it into the blood. It also scored best in its group on how easily it passes into the body.
5. It mixes well with other drugs.
Cancer patients often take several medicines at once. IAM217 showed fewer warning signs for clashing with other drugs than all three rivals in Iambic's comparison. It also worked better together with a common chemotherapy than either drug alone, in lab tests.
6. It worked across many cancers.
Tumors shrank in breast, ovarian, colon and bone cancer models.
Why this is bullish:
This is not a one-off. Iambic's lead drug, IAM1363, is already in patients and is shrinking or holding brain tumors in 21 of 22 people who had them. Now its second drug shows the same strength in the brain.
That looks like a pattern. Iambic's AI platform keeps producing drugs that reach the brain, which is exactly where most competitors fall short.
IAM217 is already cleared by the FDA to start human trials. A third drug is close behind. And the company has about $455M of cash after its IPO.
Source: cdn.prod.website-files.com/6…
SM retweeted
Something else to add which I find interesting.
$NVDA completely exited $RXRX, the AI drug discovery stock it backed in 2023.
A few months later, it bought more $IAM.
Now look at how small NVIDIA’s public portfolio is. Its latest 13F shows just eight US-listed stocks, led by SpaceX, Intel and CoreWeave. Iambic listed in October with pre-IPO holders locked up until April, so Iambic is the ninth.
Out of a tiny public portfolio, Iambic is very likely NVIDIA’s biggest AI drug discovery bet. A ~$700M company sitting next to SpaceX and Intel.
$IAM these are the results from a company valued at 1/82th of its closest competitor, with a 12x faster AI model, 100% of OPEX already covered by big pharma payments, a 2nd drug advancing to Phase 1, and over 75% of market cap in cash. $PDSB $NAUT
$IAM earnings are out of this world 🔥🚀
We have reached the pivotal point where partners are already funding the OPEX.
Last 9 months: $99.8M billed vs. $107.7M of OPEX.
Q3: $53.9M billed vs. $41.5M of OPEX.
Revenue up 142% YoY. $455M cash. Market cap $580M.
SM retweeted
🚨 𝗕𝗥𝗘𝗔𝗞𝗜𝗡𝗚: Michael Carrick has published a very LONG and EMOTIONAL letter to Manchester United fans! ❤️
"Dear United fans,
I've had the immense privilege of leading Manchester United, the football club that I love.
There are very few who get to live that and feel it, and for that I'm incredibly grateful.
With that comes huge responsibility, and to be at Manchester United, you need to understand exactly what that means.
It's different to anywhere else.
Ultimately, you need to push to be the best, but there is a culture, there are values and there are behaviours that you have to live by and respect, which have been built over a long period of time.
It's a football club that is unique, with substance and history that cannot be ignored.
This is what inspired me every single day.
I want to say a big thank you to the supporters. I must say, you are a special, rare breed.
We've had some great times this year and moments we'll all remember. The best thing about being in my position was seeing so many of you lose yourselves in the emotion of those moments, leaving the stadium desperate to come back again next time.
I just wanted to give you something back.
That love never dies.
I know we've not had the results lately, but you've shown me nothing but love and support in times when the easiest thing, and what most end up doing, would be to turn the other way.
That means more than a win or two.
That's what makes you who you are.
That's why I fell in love with this club all those years ago.
A special thank you to my coaching staff. You gave me everything you had, every single day, and I couldn't have asked for any more. I'm incredibly grateful for your loyalty, hard work and support throughout.
To all the staff behind the scenes, thank you for everything you do for this football club. So much of your work goes unseen, but it never goes unappreciated.
And to the players, thank you for your efforts and the moments we've shared. I wish you all the very best for what's ahead.
I want nothing but success for Manchester United. This club has been such a big part of my life and that will never change. I'll always be proud to have played my part and I'll always want the very best for it.
I'll be right there with you from now on, watching from the sidelines, cheering the lads on."
Michael.
SM retweeted
I spent the last week going through $IAM line by line. S-1 exhibits, job boards, trademark filings, trial registries, the actual partner contracts. Here's what I found that I haven't seen anyone talk about.
First, the setup.
Iambic is an AI drug discovery company with a real drug in the clinic: IAM1363, a brain-penetrant HER2 pill, 122 patients dosed. It trades around $14–15, roughly a $700M company.
Isomorphic (Google DeepMind's drug spinout) is reportedly raising at ~$50B. It has no drug in human trials yet.
1. They're quietly building a GPCR shop
GLP-1, GIP and amylin all work through GPCRs. Iambic has been staffing GPCR biology for years:
• Head of GPCR Biology since 2024: Will Barnes. Postdoc under Nobel laureate Bob Lefkowitz (the GPCR guy), then Director of Renal & Metabolic Drug Discovery. His public bio says he now leads Iambic's non-oncology programs "from target identification through entry into clinical development."
• Oct 2025: they posted a GPCR Biology Director role ($180–240K) to "define and own GPCR biology strategy across one or more programs."
• An older GPCR posting described the internal pipeline as "oncology, metabolic, and CNS."
• The S-1 lists discovery programs in metabolic disease and endocrinology.
• In April, Barnes presented at a GPCR drug discovery summit on using NeuralPLexer + Enchant for "GPCR assets."
No metabolic program is named yet. But the people, the hiring and the filings all point the same direction.
2. The partner contracts are attached to the S-1. I read them.
• Takeda deal: carves out "Enablement Licenses," letting third parties run Iambic's AI models through a hosted SaaS/API. Their NeuralPLexer trademark is filed for platform-as-a-service. Nobody is modeling a software line.
• The Takeda deal also defines "Proximity Inducing Proteins" and "Multi-Protein Targets." That's molecular glue / degrader language.
• Lundbeck deal: $10M upfront, Lundbeck pays for the team's work, and Iambic keeps all the collaboration IP plus the right to train its models on the data. Exclusivity covers one target only, so every other GPCR stays open.
3. The board and advisors
• Director Shalini Sharp also sits on the boards of Neurocrine and Septerna, two GPCR-focused companies.
• Director Kurt von Emster sits on Abingworth/Carlyle's clinical co-development committee. That's capital that funds late-stage trials.
• Advisors: Nobel laureate Frances Arnold, Amgen's former head of research Ray Deshaies, NVIDIA's former AI director Anima Anandkumar, and Derek Lowe, pharma's best-known skeptic.
4. The HER2 trial keeps growing
The registry target went from 243 patients at 28 sites to 383 patients at 48 sites. You don't expand a trial on a drug you're losing faith in.
5. The money is behind it
• Partners: Takeda, AbbVie, Bayer, Lundbeck, Jazz, RevMed. The AbbVie deal announcement called Enchant v3's prediction accuracy "industry-leading."
• A director's fund bought $11.5M in the IPO.
• ARK bought every dip from Nov 2–6.
Why the GPCR angle matters
When Septerna signed its oral GPCR obesity deal with Novo ($200M+ upfront and near-term), the stock ran 69% in a day. One deal that size would be $4+ a share in cash for $IAM.
What I'm watching
• Fri: Russell preliminary additions
• Nov 18: CEO on Jefferies' AI drug discovery panel
• Dec 9: SABCS (zongertinib breast data, the competitor to watch)
• First 10-Q in the next few weeks
• Patent publications and new GPCR job posts
What could go wrong
The confirmed response rate is still maturing (23% in the last cut). There's a big lock-up expiry in April. They'll likely need to raise again by 2027–28. And none of the GPCR work is disclosed yet. This is a read of public breadcrumbs, not a guarantee.
But at ~$700M you're getting a clinical HER2 drug, six pharma partners, a 41B-parameter model, and a GPCR team run by a Lefkowitz alum, for under 2% of what Isomorphic is asking with nothing in the clinic.
NFA. Do your own work.
Meet $IAM's Head of GPCR Biology 👇
🧬 Trained under Nobel laureate Bob Lefkowitz, the GPCR legend
💊 Before Iambic: Director, Renal & Metabolic Drug Discovery
🎯 Now leads ALL non-oncology programs "through entry into clinical development"
🎤 April talk: NeuralPLexer + Enchant for "GPCR assets"
GLP-1's receptor is a GPCR. 👀
NFA
SM retweeted
$IAM's hidden metabolic bet: what the filings show, what they don't, and what it's worth
Disclosure: I am long $IAM at an average of $9.95. Research, not advice.
The short version
• Iambic told investors it works on metabolic disease, and its CEO repeated it on the record. That line was added to the IPO filing late, between the July and August drafts.
• The clue is GPCRs, the receptor family that GLP-1 and every other big obesity target belongs to. Iambic has a head of GPCR biology, and its audit chair sits on the board of a company with a $2.2B Novo Nordisk obesity deal.
• Peptides are a model capability, not a program. The prospectus never uses the word.
• The obesity pill race has moved on. The liver problem that killed Pfizer's pills has been solved by at least four groups. The open lanes are oral amylin, next-gen receptor targets and longer-lasting peptides.
• One typical early obesity deal pays about $80M to $110M upfront. That is a quarter to a third of Iambic's enterprise value. Nothing is priced in today.
1. Iambic said "metabolic". Then it said it again.
The confidential IPO draft of 24 July 2026 does not mention metabolic disease. The next draft, on 28 August, adds this sentence, which survives unchanged into the final prospectus:
"we are engaged in a number of discovery stage programs in indications, including oncology, metabolic disease and endocrinology, which we estimate represent an over $70 billion in aggregate global commercial opportunity."
The same draft adds that Iambic has eight discovery programs, wholly owned and partnered, and expects up to 25 in 2027. It also says partnerships are expanding the platform to new target classes, "for example, G protein-coupled receptors".
On 16 October the CEO, Tom Miller, was asked about moving beyond cancer. About the wholly owned pipeline, he said: "It includes metabolic disease. It includes endocrinology." (BioPharma Dive)
The company's own documents do not all agree:
• The Enchant v3 page (21 September) lists the pipeline areas as oncology, neuroscience and inflammation and immunology. Metabolic is not listed.
• The risk factor on moving beyond cancer names neurology, gastrointestinal and inflammation only.
• There is no use-of-proceeds line for metabolic work and no metabolic row on the pipeline page.
My read: something metabolic existed or was planned by late August, and Iambic chose to name it before the IPO. It has not said what.
2. Why GPCRs are the tell
A GPCR (G protein-coupled receptor) is a switch on the surface of a cell. The big obesity targets are GPCRs: the GLP-1, GIP, glucagon, amylin and MC4R receptors. Lilly's new obesity pill, Foundayo, is a small molecule that flips the GLP-1 switch.
What Iambic has:
• A GPCR team. An "Executive Director, Head of GPCR Biology" has appeared on Iambic's LinkedIn page since at least February 2025.
• GPCR agonist work. From March 2025 to March 2026 the pipeline page showed a partnered "GPCR target / allosteric agonist / neurological" program. That is very likely Lundbeck (my inference). It is the only disclosed work in which Iambic designs a pill that switches a GPCR on, which is the same job an oral GLP-1 does. It is in brain disease, not obesity.
• A boardroom view of the playbook. Audit chair Shalini Sharp has been a director of Septerna since January 2024. In May 2025 Septerna signed an obesity deal with Novo Nordisk worth up to about $2.2B, with more than $200M upfront and near-term. It covers oral small molecules on GPCR targets "including GLP-1, GIP, and glucagon receptors".
None of this proves a program. It shows the skills, the people and the template are in the building.
3. Peptides: the capability is real, the program isn't
• Enchant v3 (September 2026) lists peptides as a new capability, citing "the success of GLP-1s". It publishes no peptide results.
• In NeuralPLexer3's paper, protein–peptide docking was its strongest category: 85.1% against 76.6% for AlphaFold-Multimer. That is 47 cases with overlapping uncertainty ranges, and AlphaFold 3 was not compared.
• The prospectus never uses the word "peptide". I found no peptide patent, no peptide program and no peptide job post in the live board or 65 archived ones.
So "dark horse in peptides" is a stretch today. "Dark horse in oral metabolic pills" fits the evidence better.
4. Where the obesity market actually is
The market is enormous:
• Lilly's tirzepatide (Mounjaro and Zepbound) sold about $36.5B in 2025 and $43.8B in the first nine months of 2026.
• Novo's semaglutide franchise sold about DKK 228B (roughly $35B) in 2025.
Pills are coming, but the peptide pill is winning so far. In the first nine months of 2026 the Wegovy pill (oral semaglutide, a peptide) sold DKK 10.9B, roughly $1.7B. Foundayo (a small molecule) sold $281M. Part of that is timing: the Wegovy pill launched in January, Foundayo in April.
Oral peptides are wasteful. The FDA label puts oral semaglutide's absorption at about 0.4% to 1%. The Wegovy pill is 25 mg a day against 2.4 mg a week injected.
The liver problem is mostly solved. Pfizer stopped lotiglipron in 2023 and danuglipron in April 2025 over liver safety. Since then, orforglipron (Foundayo), Structure's aleniglipron, AstraZeneca's elecoglipron and MindRank's MDR-001 have reported no liver signal. MDR-001 was designed with AI property prediction and has been in Phase 3 in China since January 2026. "AI fixes the GLP-1 pill" is already a crowded story.
What is still open:
• Weight loss. Pills take off about 11% to 15%, while injections reach 20% to 29% (retatrutide 28.7%).
• Oral amylin. Amylin is the hottest next-generation target. The only oral small-molecule amylin drug in human trials is Structure's ACCG-2671, in Phase 1 since December 2025.
• Longer-lasting peptides. Monthly shots and better half-lives. Half-life prediction is exactly what Enchant's first benchmark measured.
That is where an AI chemistry platform with a GPCR group could matter.
5. What it could be worth
What pharma paid upfront for early-stage (discovery or preclinical) obesity assets, 2023 to 2026:
• Novo – Lexicon (2025), preclinical, oral: $45M
• Lilly – Nimbus (Jan 2026), discovery, oral: $55M incl. near-term
• Merck – Hansoh (Dec 2024), preclinical, oral: $112M
• Novo – Valo (Jan 2025), discovery, AI: $190M incl. equity and near-term
• Novo – Septerna (May 2025), discovery, oral: >$200M incl. near-term
• Novo – Hengrui (Sep 2026), Phase 1-ready, oral: $300M
The median across seven early deals is about $112M. Counting only clean upfronts (four deals), it is about $79M.
For IAM: one such deal is $1.65 to $2.35 a share, about 25% to 36% of today's roughly $313M enterprise value. The market pays nothing for it today.
The reality check: Structure Therapeutics. Structure has a Phase 3-ready oral GLP-1 and the only oral amylin pill in the clinic. Its enterprise value on 11 November was about $276M ($1.51B market value less $1.24B of cash), below Iambic's. The market is not handing out multiples for oral obesity pipelines. For Iambic, the value of a metabolic program arrives as partner cash, not as a re-rating.
6. What to watch
• Patents. International applications publish about 18 months after filing, so anything filed after about May 2025 is still invisible. A GPCR or incretin family appearing in 2027 would be the first hard proof.
• SEC comment letters. Once released, they may show whether the SEC asked about the August metabolic addition.
• The first 10-Q (around 30 November) and the J.P. Morgan conference (January): any named metabolic target.
• The job board. In vivo obesity pharmacology or GPCR metabolic roles.
• Bayer. The June 2026 deal's therapeutic area is still undisclosed. Bayer's focus is cardio-renal, oncology and women's health.
Bottom line
There is no named metabolic program, no peptide program and no metabolic partner. There is a late-added prospectus line, a CEO who said it twice in one breath, a GPCR team, a director who has seen a $2.2B Novo GPCR deal from the inside, and a deal market that pays about $80M to $110M upfront for exactly this stage.
At about $313M of enterprise value, that is a free option on top of the HER2 drug, not the thesis. Size it that way.
Long $IAM at $9.95. Not advice.
One $IAM drug that's already been synthesised could be worth more than the whole company.
In 2024 Genentech paid $850M upfront for Regor's CDK cancer drugs. The lead had been tested in just 12 patients.
pharmaphorum.com/news/genent…
Iambic has its own: IAM-C1. Same targets, CDK2 and CDK4, designed by Iambic's AI to avoid the problem that sank Pfizer's version. Heading into the clinic next, IND guided this quarter.
Price IAM-C1 like Regor's:
~$17.90 a share
+ ~$8.60 of cash per share
= ~$26.50, vs $15.50 today
And that ignores everything else:
IAM1363, the lead drug
- Already in 122 patients across 53 sites in 8 countries
- Brain disease controlled in 21 of 22 patients
- Registrational trial guided for as early as 2027
IAM217, built to reach the brain
- Shrank brain tumours ~90% in mice where a rival drug did nothing
- Designed by Iambic's AI without a lab structure of its target
The AI platform
- $77M upfront from AbbVie, Takeda and Lundbeck, with ~$2.4B more in potential milestones
- NeuralPLexer matches AlphaFold 3's accuracy at ~15x the speed (Iambic's figures)
- Enchant v3: 41B parameters, more accurate with every version
- Google's AI drug company Isomorphic is reportedly valued at ~$50B
One drug out of three, priced like Regor's, is worth more than all of Iambic today, cash included.
NFA
"Iambic’s business is priced at ~0.6% of Isomorphic’s."
In $IAM you get superior tech with fraction of price and drug already in trial.
The valuation gap does not make sense.
Iambic is $100 stock minimum.
$IAM vs Isomorphic, side by side. What’s a fair price for the small one?
Isomorphic is $GOOGL drug company:
AlphaFold 3, built with Google DeepMind
IsoDDE, an in-house engine it says more than doubles AlphaFold 3’s accuracy on the hardest cases (its own numbers)
Lilly and Novartis: $82.5M upfront, up to ~$2.9B in milestones
$2.7B raised, ~$5B more reportedly coming
No drug in human trials yet
Valued at ~$50B. Take out roughly $5B of cash and the business is ~$45B.
Iambic has:
NeuralPLexer3: ~30 seconds per prediction vs ~6 minutes for AlphaFold 3, and more physically valid predictions (77.9% vs 73.1%, Iambic’s numbers)
Enchant v3, 41B parameters
Its own automated lab, ~1,000 molecules a week on a 7-10 day loop
AbbVie, Takeda, Lundbeck: ~$77M upfront, up to ~$2.4B in milestones, plus Bayer and RevMed
A HER2 drug in 122 patients, and two more entering the clinic
Valued at ~$665M. Take out ~$410M of cash and the business is ~$255M.
The deal books are about the same size. The market values Isomorphic’s at ~545x its upfront cash, and Iambic’s at ~3x.
Iambic’s business is priced at ~0.6% of Isomorphic’s.
Isomorphic earns a premium: Alphabet, DeepMind, the AlphaFold name, a far bigger war chest. The question is how big.
Iambic at a share of Isomorphic’s business:
1% → ~$18 a share
2% → ~$28
3% → ~$37
5% → ~$56
Wall Street’s targets ($22-35) put it at 1.4-2.8%. It trades at 0.6%.
SM retweeted
ai and biotech is the sector i think the market is mispricing the most right now. two small caps show it well.
$IAM (Iambic) -> up 31% today to $13.38, market cap $634m, cash $409m, enterprise value about $225m, funded to q4 2028
$NAUT (Nautilus) -> $2.38, market cap $303m, cash $114m, enterprise value about $189m, runway into q1 2028 with a raise planned before mid-2027
before today's move Iambic was trading close to its cash, with an enterprise value around $75m. i've already been in and out of $NAUT for a 60% profit, and it has gone from under $1 to $2.38 since early october. a price like that is usually when a company raises, so dilution is the risk to watch.
of the two i still prefer $IAM cash is 65% of its market cap against 38% for $NAUT and it is funded two years out.
which one would you rather own going into a raise?
What is new in Iambicpedia $IAM:
- Browse by topic. All DD is now categorized: 31 pages in six sections, covering the company, the AI models, the drugs and the stock.
- Recent updates. Newest first, with filters to show one kind of news at a time.
- Catalyst calendar. Every date is tagged as catalyst, filing or overhang.
- Q&A section. 14 questions with short answers.
- Glossary with 40+ terms. Sources page with links to the SEC filings.
- Random page button, if you just want to explore.
Read it here: lekiresearch.com/free/wiki

























