Emmanuel Noutahi (@ENoutahi) profile picture

Eternal skeptic, occasional optimist. AI, Bio, Chem, and side quests.

Montréal, Québec
Joined January 2015
Photo by Emmanuel Noutahi (@ENoutahi)Photo by Emmanuel Noutahi (@ENoutahi)Photo by Emmanuel Noutahi (@ENoutahi)Photo by Emmanuel Noutahi (@ENoutahi)Photo by Emmanuel Noutahi (@ENoutahi)Photo by Emmanuel Noutahi (@ENoutahi)Photo by Emmanuel Noutahi (@ENoutahi)Photo by Emmanuel Noutahi (@ENoutahi)Photo by Emmanuel Noutahi (@ENoutahi)Photo by Emmanuel Noutahi (@ENoutahi)
Ever heard of Spasfon, the pink pill? It has been prescribed for more than 60 years, especially to women. The catch? Clinical evidence for many of its common uses is remarkably weak. Ep 2 of Beyond the Assay is out: enoutahi.com/blog/spasfon-fa…
4
3
28
103,138
Emmanuel Noutahi retweeted
Folks often point to the speed of clinical trials as a reason why AI won’t result in more drugs. We’re building the foundational predictive capabilities that will resolve this bottleneck, using actual biology. Advances in biostats and trial regulation allow us to directly convert predictive signal into smaller, cheaper, faster trials without compromising safety — more on this to come.
Today, we are excited to introduce Ladon, a state-of-the-art biological foundation model that predicts disease trajectory and treatment response in patients. There is immense complexity in the individual biology of every patient. Two patients can have the same diagnosis and receive the same treatment, yet have very different outcomes: one responds well and lives for years, while the other’s disease continues to progress. Understanding and resolving this heterogeneity to predict outcomes is how we accelerate our delivery of medicines for patients and unlock clinical trial abundance. Ladon is our first step towards this future: one where we can learn more from every patient enrolled so that we can run smaller, faster trials with higher success rates. We built it to give drug developers quantitative support on billion-dollar clinical trial decisions. Today we're excited to share Ladon's capabilities in our blog and technical report. Ladon is a 3B parameter model trained on millions of RNA-seq samples spanning decades of technical and biological variation across tissues and species. Ladon beats gene expression baselines, established biomarkers, and existing RNA foundation models at predicting patient prognosis and therapeutic response in 140+ oncology tasks, including external validation on past clinical trials. This is significant because clinical trials are fundamentally generalization problems. We show that Ladon-derived predictions are robust across indications, treatments, and sequencing platforms. Reach out if you’d like to get started with the model today! Read more about Ladon in the blog: blank.bio/introducing-ladon Find more technical details in the white paper: blank.bio/introducing-ladon
3
6
2
48
2,341
Building RNA-seq models that generalize across patient cohorts is really hard. @blankbio_ has some really cool results with Ladon. Congrats to the team!
Today, we are excited to introduce Ladon, a state-of-the-art biological foundation model that predicts disease trajectory and treatment response in patients. There is immense complexity in the individual biology of every patient. Two patients can have the same diagnosis and receive the same treatment, yet have very different outcomes: one responds well and lives for years, while the other’s disease continues to progress. Understanding and resolving this heterogeneity to predict outcomes is how we accelerate our delivery of medicines for patients and unlock clinical trial abundance. Ladon is our first step towards this future: one where we can learn more from every patient enrolled so that we can run smaller, faster trials with higher success rates. We built it to give drug developers quantitative support on billion-dollar clinical trial decisions. Today we're excited to share Ladon's capabilities in our blog and technical report. Ladon is a 3B parameter model trained on millions of RNA-seq samples spanning decades of technical and biological variation across tissues and species. Ladon beats gene expression baselines, established biomarkers, and existing RNA foundation models at predicting patient prognosis and therapeutic response in 140+ oncology tasks, including external validation on past clinical trials. This is significant because clinical trials are fundamentally generalization problems. We show that Ladon-derived predictions are robust across indications, treatments, and sequencing platforms. Reach out if you’d like to get started with the model today! Read more about Ladon in the blog: blank.bio/introducing-ladon Find more technical details in the white paper: blank.bio/introducing-ladon
1
19
720
You can see writing examples and a description of the method in our blog: advances.edisonscientific.co… And read the paper here: arxiv.org/abs/2610.38851 5/5
6
5
1
127
10,197
This weekend, I used frontier LLMs to screen the math claims in ~4,800 accepted NeurIPS 2026 papers, over half across all tracks. Agents read the claims, and when something doesn’t hold, they explain why, check whether it was already addressed elsewhere, and propose a repair when they can. A second set of agents then checks the criticism and refine it. 14% of the assessed papers were flagged for major or critical issues. proofnote.enoutahi.com/
3
10
215
This is less exciting than solving Millennium problems, but speeding up science a little still seems fun. Maybe @NeurIPSConf could run automated checks on mathematical claims during review ? And frontier labs could sponsor it 😁 ?
1
1
41
Explore: proofnote.enoutahi.com/ Code: github.com/maclandrol/proofn… Authors of flagged papers: if a report gets something wrong, please reach out. If it’s right, don’t feel bad, small errors happen to the best of us. That’s part of science.
32
Emmanuel Noutahi retweeted
The Singularity is coming for Physics and Biology. RIP Math.
105
652
113
6,869
642,974
Emmanuel Noutahi retweeted
Lots of discussion at WorldADC about needing better biomarkers. Measuring target expression is not enough - we need better signatures of response, resistance, and toxicity, that map to the biological mechanism of action for ADCs
3
12
813
Emmanuel Noutahi retweeted
Seeking someone to fly some prototypes from London to the US on October 18th. Too important to trust FedEx… Who is in London and wants to work on something fun, get a free flight to/from the US and a fee to make it worthwhile? Need to be referencable in my network… DM me if interested.
2
9
1,822
If you (or your AI agents 😉) use the Datamol ecosystem, I updated datamol, molfeat, medchem and SAFE a few weeks ago: bug fixes, new featurizers, and support for latest Python and RDKit versions. SAFE also fixes some edge cases where SMILES ↔ SAFE conversion lost tetrahedral or E/Z stereo. github.com/datamol-io/
1
1
5
149
Emmanuel Noutahi retweeted
Building predictive biology models is one of the most important scientific challenges of our time, but solving this requires new experimental biological data at an unprecedented scale. I’m pleased to announce that @GoogleDeepmind is partnering with @ChanZuckerberg’s Virtual Biology Initiative to create the multi-modal datasets needed for improved predictive biology models. I wrote recently about the need to discover how life works at many different scales, from DNA and individual proteins to living cells and whole-body health. As we learnt from AlphaFold and the Protein Data Bank, solving grand biological challenges will not be possible without high-quality open data and efforts like the @czbiohub initiative. Read the announcement here: biohub.org/news/virtual-biol… And my previous essay here: linkedin.com/pulse/understan… @AllenInstitute, @BroadInstitute, @GladstoneInst, @humancellatlas, @ProteinAtlas, @sangerinstitute, @IsomorphicLabs, @Meta
9
43
6
421
35,054
Emmanuel Noutahi retweeted
In Boston for WorldCDx this week! Excited to share early looks into our transcriptomic foundation model with a few partners from pre-clinical to clinical. Many issues with signature robustness and cross-species transfer can be solved with AI. We're working on this.
1
2
13
485
The submissions per month to arxiv has reached lift-off. One view is research has been meaningfully accelerated by AI and this is the effect. A cynical explanation is it's low-effort slop and preprints have lost all meaning. I really wonder how we can measure which is true.
8
21
2
130
10,430
Emmanuel Noutahi retweeted
Choosing a drug target is one of the most consequential decisions in drug discovery. Yet the evidence is vast, fragmented, and constantly changing: making systematic assessment across thousands of targets nearly impossible to do manually. Introducing Target Atlas, agentic target intelligence powered by Biomni agents. Hundreds of AI agents search scientific literature and databases in parallel, extract evidence against defined target-assessment criteria, and synthesize it into structured, source-linked assessments. But the Atlas isn't static. Scientists can deploy Biomni agents directly to interrogate evidence, integrate private data, compare and prioritize targets, investigate new hypotheses, and carry out downstream work: from running analyses to designing binders and planning experiments. Our first disease-specific application was on Parkinson's disease, developed with The Michael J. Fox Foundation's Targets to Therapies Initiative @MichaelJFoxOrg, combining an expert-defined target evaluation framework with additional criteria and evidence collected and analyzed by AI agents. We've also extended the approach to an ever-growing set of 20+ diseases spanning oncology, autoimmune, cardiovascular, and neurological conditions, creating a systematic view of the target landscape at a scale that would be difficult to build and maintain manually. Target Atlas is available in beta today at biomni.phylo.bio — we'd love your feedback as we continue to improve it! Read more: phylo.bio/blog/target-atlas-…
12
41
3
280
24,036
Emmanuel Noutahi retweeted
We've rebuilt the entire @AsimovPress website. We experimented with dozens of fonts and type sizes, column widths, and color palettes to find options that felt comfortable to read. JavaScript is also much easier to embed; many of our articles will include interactive elements. Footnotes, too, are now (finally!) visible in the margins.
9
21
2
282
10,999
Emmanuel Noutahi retweeted
A generative model can look like it's working and still fail to output the disease-relevant cell groups or small molecules you care most about. Fortunately, we have techniques to diagnose and correct these failures. In our most recent Valence at the Whiteboard, Liam explains how to evaluate generative models using Precision, Recall, Density and Coverage (PRDC) to spot subtle failure modes and know what to do about them. Full video: youtu.be/teIoP92Aeow
2
5
14
1,228
Emmanuel Noutahi retweeted
I am very proud of what the team at @LawZero_ is building and deeply grateful for the landmark support from the Governments of Canada and Germany that will power our next phase of growth!
We are happy to share that Canada and Germany have announced funding of up to $300 million for LawZero! This will power the next phase of our technical roadmap towards developing a fundamentally new form of advanced, safe, and capable AI. lawzero.org/en/news/lawzero-…
11
32
4
199
16,661
Emmanuel Noutahi retweeted
Today, there's a strong focus on fast-follows of existing drugs. The drug programs that inspire me look different – the target is known, but we don't know how to build the drug, often requiring the dev of new tech that paves the path for more drugs.
6
31
14
237
66,207
Emmanuel Noutahi retweeted
Reminder that the winner will be selected this Friday! The prize is 1 free year of ChatGPT Pro ($200 per month tier) Submissions must use the Rosalind Workbench suite of plugins (can also find them under the "Scientific Research" section of the plugin marketplace)
The Rosalind Workbench has been out for a week now - show me the coolest thing you've built with it! Coolest project gets 1 free year of ChatGPT Pro! Either comment here or quote tweet me to submit. Provide screenshots and the Rosalind plugins you used to make it happen!
2
1
13
2,192
"The Machine Stops" by Forster is one of those stories that feels more contemporary every time you return to it cs.ucdavis.edu/~koehl/Teachi… Written in 1909, but it feels so current....
1
90