计算速度下的实验室在环——Leash Bio 的 Ian Quigley
Latent Space 对话 Leash Bio 的 Ian Quigley，谈 DNA 编码库、Hermes 模型、ChEMBL 中的“聪明汉斯”现象，以及面向科学推理的智能体沙盒 Minos。
中文处理结果
几年前我偶然发现了 Leash Bio 的 Substack，然后就被吸引住了。无论是他们的 Kaggle 竞赛——它表明我们其实并没有解决蛋白质-配体结合问题——还是他们那个疯狂的发现：知道一个分子的作者，也许就足以预测它有多好。他们的 Substack——或者至少 Owl Posting[1] 对他们的深度解读——应该是任何试图用 AI 做化学的科学家的必读材料。
所以，当我终于能坐下来与 Leash Bio 的 Ian Quigley 交流、向他请教时，我非常兴奋。Leash 专注于 DNA 编码库（即“DEL”），这是一种收集蛋白质与小分子之间超高通量相互作用数据的方法。[2] 在我发现 Leash 之前，我从 DEL 用户那里听到的全是恐怖故事。然而，Ian 和他的朋友们不知怎么找到了驯服这头野兽的方法，提取出干净、可复现、非常适合大规模建模的数据。不剧透太多，答案归结为认识论上的严谨，以及一种令人难以置信的信念：你必须去做那件事。
更棒的是！在这期节目接近尾声时，Ian 透露了他们团队在我们录制时刚刚在玩的东西：一个让智能体进行可验证科学推理的 AI 游乐场。他们发现，过去几年收集的数据可以提供数万个“实验室在环”（lab-in-the-loop）模拟，而这正是智能体做科学强化学习所需要的。两个月后，它已经成熟为 Minos，今天他们发布了完整的预印本、博客和开放数据集。我们将在几天后发布一期闪电播客来更新进展。敬请期待！
我们和 Ian 聊得非常开心，这可能是我们迄今为止最有趣的一期节目之一。无论你是完全的化学新手，还是 AI x Bio 的专家实践者，我们觉得你都能从中有所收获。
享受吧！
我们讨论了：
- 为什么 DEL 会让每个接触它的人留下心理阴影，以及 Leash 如何在每个分子上放置几十个条形码，以捕获那些结合 DNA 而非药物的蛋白质。
- Ian 的法则：生物学只有两种扩展方式——拍照和测序 DNA。
- 那场 Kaggle 竞赛：来自一个地下室的大约 40 亿次测量、约 2000 支队伍，而在一个未见过的化学库上没有人能超过随机水平。还有为什么 Leash 曾暗自担心共折叠会让它们过时。
- ChEMBL 中的“聪明汉斯”：模型可以从分子结构猜出它的化学家是谁，而“是谁做的”对活性的预测能力几乎与结构本身一样好。
- Hermes，仅用 Leash 自己的数据训练，在公开基准上与 Boltz 并驾齐驱。还有那个 Boltz 已经在 PDB 中见过的 BRD4 分子。
- 打破口袋：在发现 Hermes 记住的是蛋白质的“绰号”而非结合位点之后，大规模工程化蛋白质突变体。
- 类 Lumacaftor 的命中物，意外成为错误折叠蛋白质的检测器。
- 智能体沙盒：用平实的英语推理 SAR 的 LLM，在 100 的预算内从 650 万个分子中寻找最佳结合物。
原始正文
Lab-in-the-Loop at Compute Speed — Ian Quigley, Leash Bio
A few years ago I stumbled upon Leash Bio’s Substack, and was hooked. Be it their Kaggle competition that showed that no, we really haven’t solved protein-ligand binding, or their crazy discovery that knowing a molecule’s author is perhaps all you need to know to predict how good it is. Their substack — or at least Owl Posting’s1 deep dive on them — should be required reading for any scientist trying to do chemistry with AI.
So I was super excited to finally be able to sit down with Ian Quigley from Leash Bio, and pick his brain. Leash specializes in DNA Encoded Libraries (aka “DELs”), a way of collecting ridiculously high-throughput interaction data between proteins and small molecules.2 Before I discovered Leash, I had heard nothing but horror stories from DEL users. And yet, Ian and friends somehow found a way to tame the beast, extracting clean and reproducible data that is just perfect for large scale modeling. Without teasing too much, the answer comes down to epistemic rigor, and an incredible conviction that you just have to Do The Thing.
It gets even better! Near the end of this episode, Ian teases something his team was just playing with when we recorded: an AI playground for agents to do verifiable scientific reasoning. They discovered the data they had spent the last few years collecting could provide tens of thousands of “lab-in-the-loop” simulations, just what an agent needs for scientific RL. Two months later, it’s matured into Minos, and today they’re releasing a full preprint, blog, and open dataset. We’ll be releasing a lightning podcast in a few days with an update. Stay tuned!
We had a blast with Ian, probably one of our most fun episodes to date. For both complete chemistry novices and expert AI x Bio practitioners, we think you’ll get something out of this.
Enjoy!
We covered:
- Why DELs traumatize everyone who touches them, and how Leash puts dozens of barcodes on each molecule to catch proteins that bind the DNA instead of the drug.
- Ian’s rule that biology only scales two ways: taking pictures and sequencing DNA.
- The Kaggle competition: about 4B measurements from a basement and about 2,000 teams, with nobody beating chance on an unseen chemical library. Also why Leash quietly feared co-folding would make them obsolete.
- Clever Hans in ChEMBL: models can guess a molecule’s chemist from its structure, and “who made it” predicts activity about as well as the structure does.
- Hermes, trained only on Leash’s own data, keeping pace with Boltz on public benchmarks. Plus the BRD4 molecule Boltz had already seen in the PDB.
- Breaking the pocket: engineering protein mutants at scale after catching Hermes memorizing “nicknames” for proteins instead of binding sites.
- Lumacaftor-like hits as an accidental detector for misfolded proteins.
- The agent sandbox: LLMs that reason about SAR in plain English, hunting for the best binders among 6.5M molecules on a budget of 100.
- The Recursion mafia, and why its alumni companies all rhyme.
- The Kosuri paradox: if you can collect the data, why do you need the model?
- Going from a DEL hit to good mouse PKPD in six months on an inflammation program, and the case for pills over antibodies on cytokine targets.
- What stops a better-funded lab from outscaling them, and Ian’s by-fiat wish for massively parallel purification.
1
The hands-down best bio blogger on the internet.
2
As always, the secret here is to map your problem to a combinatorial library and a readout using Next Generation Sequencing. We talk about this in depth in the podcast!