Today, in collaboration with Anthropic, we’re excited to share rapid initial progress towards an antibody treatment for food allergy, which affects over 200 million people worldwide, discovered with the help of Claude.
At Divergence Labs, we believe that AI can unlock the next generation of antibody therapeutics with more effective mechanisms than before. Such mechanisms are urgently needed in food allergy, which involve overcoming one of the tightest affinity protein pairs in immunology, IgE and its receptor FcεRI, on mast cells.
Treatment with Xolair—the only approved antibody for food allergy—sequesters circulating IgE, but mast cells remain sensitized for weeks to months due to the extraordinarily slow dissociation of IgE from FcεRI.
We envision a treatment that actively pulls IgE off of FcεRI, without waiting for natural dissociation. Like picking a lock, this requires the right combination of geometry and force.
Our research team, working with Claude, ideated on mechanisms to accomplish this disruption, and converged on an antibody format with two arms with different functions and epitopes. Within one month, we designed, synthesized, and experimentally confirmed the biophysical function of the two arms, which we are now assembling into full therapeutic formats.
Claude was trained on protein sequences, at scale.
We gave 20 language models the start of little-studied bacterial proteins and asked for the next amino acids. Only the latest Claude models predict them. Not GPT, not Gemini, not seven open-weight models.
The fall of “what you can do” and the rise of “who you know”
Traditionally, many of us believe in a world where how hard you work is more important than who you know. I used to subscribe to this idea as both a scientist, who believed that facts about our world are agnostic to who says them, and as an American, who grew up believing that success was a function of how hard you worked, not who you knew.
However I am now feeling a substantial change in AI, specifically in the past five years, where who you know is becoming more important, and what you can do is becoming less important. I am obviously not saying that technical skills don’t matter, or that you cannot be successful without being well connected. Technical skills are still important and make or break your career. But I am noticing a trend. To assign arbitrary numbers, I would say that five years ago, success was 75% your skills and 25% who you knew. Now, it is maybe 55% skills and 45% who you know.
There are three technical reasons for this trend. The first is the commoditization of intelligence—coding agents have basically democratized research execution. In the past you could outwork competitors by staying up late to debug code or babysitting your experiments with higher frequency. Now, many research tasks can be completed by anyone who knows what to prompt. Research taste obviously still matters, but the playing field is definitely more equal. As technical execution becomes less of a differentiator, other advantages naturally become more important.
The second reason for this trend is that with the success of scaling language models, it is simply a fact that being resource rich makes you so much more likely to succeed. Ten years ago, an academic lab with eight GPUs had a real chance of doing great research. Now, a single RL run can require hundreds (if not thousands) of GPUs, as well as substantial data and infra. So now, two equally capable researchers can have very different outcomes depending what resources they have. And access to these resources is not a function of technical skill alone. It also depends on reputation, trust built with decision-makers, and professional relationships. These things have always mattered, but as the resource requirements of frontier AI have grown, so too has their importance.
Third, when companies stopped publishing research and moved to secretive, closed-source models, AI became a field of poor observability. In domains of perfect observability like chess or soccer, it is very easy to tell who is doing well and understand their contributions. But now in AI we have fewer public mechanisms for attributing credit, making it much harder for outsiders to know who did what for model launches. How does this get resolved? When it's difficult to independently verify someone's technical contributions, people increasingly rely on indirect signals such as reputation, personal endorsements, and who will vouch for them. Consequently, who you know plays a much bigger role in hiring, recruiting, and fundraising.
We often talk about how to improve technical skills, but nobody really talks about how to get better at "who you know". One observation is that many well-connected people mix their work and personal lives. The traditional wisdom is to keep a firm boundary between the two, which in the past was a very reasonable policy. But the returns of being social have gone up, which means that anyone still holding to that boundary is paying a much steeper price for it than they would have before.
Sam Altman is the prime example of a well-connected person who mixes work and personal. There is so much to learn from Sam. For example, back when OpenAI was smaller, Sam invited almost all the employees to his house for social hours. He hosts poker nights and dinners at his house for recruiting purposes. Sam has a reputation of being generous with his time and treating his friends well and I think that has yielded a disproportionate return for him. Obviously not all interpersonal relationships at OpenAI went perfectly but I think a large part of OpenAI’s success is from Sam’s connections.
The interesting thing about this is that intentionally trying hard at getting to know people is probably the suboptimal strategy. I don't think Sam made his connections because he knew they would pay off in his career, but rather because he is a generous person who is innately curious about people and technology. Perhaps the best strategy is to be your best and truest self, which is easier said than done, especially in a changing technological world.
While these technical trends have made the field less dependent on technical merit, I don’t think this is necessarily a bad thing, and it is not the first time this has happened in our field. For example, abstractions introduced by Python also made coding less technical and enabled more people to code. And with some skillsets, there is more opportunity now than ever. For instance, people without programming backgrounds can now create great technology, with product or research taste as more salient levers. So while what we can do is commoditized, what we choose to do, and who we do it with will matter more than ever.
Thanks Edward Hu for feedback on this essay.
After almost four months, Mol-1 has officially dethroned OpenDDE to become the #1 open-weight antibody foundation model on cofoldarena.ai. Congrats @jiarlu and @MolenceBio team!
All predictions are available to view and download.
Introducing Mol-1: Open Frontier Biomolecular Intelligence 🚀🚀🚀
📊 75.0% on FoldBench-AB, near IsoDDE's reported 75.58%!!!
🧬 60k-token assemblies on one single H20, outperforming Anthropic 8/10.
🎯 VHHs binding hit rate as high as 63% with best SPR KD at 43 pM.
GitHub: github.com/molence/Mol-1
Try it: lab.molence.com
Language models can learn new tasks from just a few examples in their prompt, with no retraining. This is called in-context learning (ICL). Now genome language models can do it too.
With a handful of examples, Evo 2 ICL generates proteins, de novo antibodies, CRISPR-Cas9 loci, mitochondrial genomes and entire bacteriophages.
We can even use the synthetic sequences iteratively to construct a chain-of-thought (CoT) and achieve harder design goals.
New work with Yukun A. Hao, @BrianHie and team 🧵
🚀 Excited to share BindClaw! 🦀
Starting from a prompt or research paper, BindClaw connects biological mechanisms to protein design: identify targets and binding sites → build mechanism-specific computational assays → generate and assess binders → refine designs using the results.
We use agents to evaluate whether designs fit the intended biological objective, including competition with a native partner and preference for a particular molecular state. 🧬
In a human-guided calmodulin campaign, 15 of 18 designs showed binding to calcium-loaded CaM, including subnanomolar binders supported by both BLI and SPR.
This work is an incredible team effort together with @DidiKieran , @saint__ai, @ Kyle Gion and @sacdallago and many others!
📄paper/project: research.nvidia.com/labs/dbr…
💻code: will be released in a few weeks!
Our recent work on Claude-designed anti-IgE antibodies for food allergy was spearheaded by the brilliant @EErives40101 and @ferrara_em76515, follow them for potential future updates!
In collaboration with Anthropic, we recently announced promising initial progress toward the computational design of a biparatopic disruptive IgE inhibitor for food allergy, a condition affecting upwards of 200 million people worldwide.
The current standard of care, omalizumab (Xolair), acts by sequestering free immunoglobulin E (IgE), preventing it from binding its high-affinity receptor FcεRI on mast cells and basophils, cross-linking in the presence of allergen, and triggering anaphylaxis. However, omalizumab’s inability to disrupt pre-existing receptor-bound IgE leaves many patients highly reactive and presents an urgent and unmet need.
Significant effort has therefore been directed toward the development and characterization of disruptive IgE inhibitors, which additionally drive the dissociation of receptor-bound IgE. The most potent such disruptive inhibitors are biparatopics: two “arm" molecules linked together into a single construct whose avid binding to the bound complex realizes the desired disruptive effect at significantly lower, and ideally therapeutically viable, concentrations.
We reasoned that the precise geometric specifications required of such a construct made it a promising target for computational design, and with the help of Claude, our research team at Divergence Labs undertook a computational design campaign toward this end, designing and experimentally validating several viable arms that we are presently combining into a full biparatopic.
While this work is neither a complete discovery campaign nor even the full extent of the proposed computational campaign, we believe it highlights the strength of molecular design in enabling the rapid discovery of binders with diverse geometric and kinetic profiles, both key to the function of derived multispecifics. The ability to do so at low cost in time and capital, and in a geometrically aware fashion, stands as a valuable complement to immunization, high-throughput display, and other means of screening at scale.
For more, check out Anthropic’s blog post below, and our technical report here: divergencelabs.com/publicati…. (1/2)
Public preview of BioEmu2! @MSFTResearch AI for Science’s new foundation model for biomolecular structure, dynamics and function. Open weights, MIT license.
github.com/microsoft/bioemu2
Movie: BioEmu2-simulated binding of intrinsically disordered peptide PUMA to MCL-1 receptor. 🧵⬇️
Super exciting to see the $1.8B commitment to data collection towards the virtual cell led by @biohub@alexrives ! The best direction to take this investment IMO is to scale up (no pun intended) Thousandfold Expansion Microscopy from @eboyden3@AliHShaib, enabling in situ, nm-resolution imaging of individual molecules in intact cells with simple light microscopy.
Imagine blowing up a cell to the size of a marble and studying the identity, interactions, and even conformations of every individual macromolecule in a diseased tissue. This kind of full state observability will go a long way to closing the gap between observational and causal understanding (not to mention generating petabytes of high-SNR data for AI models).
Hello world, meet 1,000× Expansion Microscopy.
1,000,000,000× expansion by volume! A gel that starts at a few centimeters will then expand to the volume of an Olympic swimming pool. biorxiv.org/content/10.64898…
In our new bioRxiv preprint, work carried out between MIT and UMG, led by Helena Hu in collaboration with scientists from the labs of @eboyden3 Ed Boyden, Silvio Rizzoli, and myself, we present Thousandfold Expansion Microscopy.
By enlarging biological specimens across multiple rounds of expansion, molecular-scale features, as small as the distances between adjacent amino acids, can be visualized with conventional optical microscopes.
Democratizing super-resolution microscopy.
When do biological reasoning models use their biological inputs?
High performance can create a mirage of reliance on biological representations. Reasoning traces don’t always reflect the evidence.
Our paper examines these limits and paths to better biological reasoning models.
Excited to share a preprint of my main PhD work in @nickpolizzi_'s lab - introducing LUCI-tag, a new orthogonal self-labeling protein tag based on K-Ras(G12C) and its covalent inhibitors! biorxiv.org/content/10.64898… (1/16)
Today, we’re announcing mBER-2, our latest AI protein design system, and sharing a bit of how we use it to explore biology in vivo at scale.
For a long time, we’ve been working on what I think of as “frontier problems” in medicine and biology. We know a lot about the targets involved in disease, but there is still an enormous amount of biology we haven’t explored, and potential uses for that biology we haven’t discovered.
One of those problems is drug delivery. There are thousands of potential targets, and most receptors remain unexplored as routes for delivering medicines. We want to understand which ones we can use, where they can take a drug, and what kinds of molecules make that possible.
Ultimately, we need to test these molecules in vivo to understand what they actually do. We’ve spent the last six years building measurement technologies that let us generate millions of measurements in living systems.
mBER-2 is an AI protein design system built for that scale of in vivo measurement. Our goal is to generate binders across the full range of targets we care about, while systematically exploring the possible binding sites on each one. It’s not just what you bind, but where and how you bind that determines whether a molecule performs it's function. mBER-2 lets us probe those differences at massive scale.
We’re also sharing a look at some of our in vivo data. We’ve designed more than 50 million molecules across over 4,000 targets, and screened millions of molecules in living systems. By probing hundreds of receptors, we’ve found new pathways for delivering genetic medicines into fat that outperform industry benchmarks by a large margin.
This is just the beginning. We call the space of all bindable sites across proteomes the EpiTome. As we explore it, we’re building the data to connect AI design, receptor binding, and what a molecule actually does in a living organism.
Much like the idea of a virtual cell, we envision a Virtual Organism that helps us design medicines with specific properties in mind. The foundation is our own in vivo data, connecting molecular design to outcomes measured in living systems.
The endgame is to use AI to explore more biology, measure what happens, and use what we learn to design better medicines. Every round should deepen our understanding of biology and improve our ability to build molecules that do what we need them to do.
Genomes are full of dark matter of unknown functions.
We present Minerva, a new method for discovery guided by genome language models. Minerva reveals the interactions hidden in non-coding DNA, pointing to hundreds of new putative RNAs and repetitive elements per bacterial genome. Minerva allows us to find and study elements invisible to traditional methods at orders of magnitude greater scale than before.
Our wet-lab validation campaign for Proteina-Complexa is now on bioRxiv! It includes some new exciting experimental results, from large protein structures (fully codesigned!) to functional carbohydrate binders.
biorxiv.org/content/10.64898…
Thread with some of the additions 🧵1/n
I’m so excited to report that pLM-designed peptides work in vivo!! 🧬🐭 In our new @ScienceAdvances paper, a SaLT&PepPr guide directs selective degradation of pathogenic β-catenin in mice! A huge milestone for sequence-based binder design. 🌟
📜: science.org/doi/10.1126/scia…
🧵👇🏾