以计算速度实现实验室闭环——Leash Bio 的 Ian Quigley

几年前，我偶然发现了 Leash Bio 的 Substack，立刻被吸引住了。无论是他们举办的 Kaggle 竞赛——它表明，不，我们确实还没有解决蛋白质—配体结合问题——还是他们疯狂的发现：要预测一个分子有多好，或许只需要知道它的作者是谁。对于任何试图用 AI 做化学研究的科学家而言，他们的 Substack——或者至少是 Owl Posting 对其进行的深度解析——都应该是必读内容。

因此，我终于能和 Leash Bio 的 Ian Quigley 坐下来深入交流时，感到非常兴奋。Leash 专注于 DNA 编码化合物库（DNA Encoded Libraries，简称“DELs”），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。在发现 Leash 之前，我听到的全是 DEL 用户的恐怖故事。然而，Ian 和他的同事们不知为何找到了驯服这头野兽的方法，提取出干净、可复现的数据，极其适合大规模建模。先不剧透太多，答案归根结底在于认知上的严谨，以及一种非凡的信念：你就是必须把这件事做成。

事情还会变得更精彩！在本期节目接近尾声时，Ian 预告了录制期间其团队刚刚开始探索的一项东西：一个让 Agent 进行可验证科学推理的 AI 试验场。他们发现，过去几年收集的数据可以提供数以万计的“实验室闭环”（lab-in-the-loop）模拟，这正是 Agent 进行科学强化学习所需要的东西。两个月后，它已发展为 Minos；如今，他们发布了完整的预印本论文、博客文章和开放数据集。几天后，我们还将发布一期闪电播客，带来最新进展。敬请关注！

我们与 Ian 聊得非常尽兴，这大概是迄今最有趣的节目之一。无论你是完全不了解化学的新手，还是 AI x Bio 领域的资深实践者，我们相信你都能从中有所收获。

尽情享受！

我们讨论了：

• 为什么 DEL 会让每一个接触它的人都深受创伤，以及 Leash 如何在每个分子上加入数十个条形码，以识别那些结合 DNA 而非药物的蛋白质。

• Ian 的一条原则：生物学规模化只有两种方式——拍照和 DNA 测序。

• Kaggle 竞赛：在地下室里完成约 40 亿次测量，吸引约 2,000 支团队参加，却没有人能在未见过的化学库上取得超越随机猜测的表现。以及 Leash 为什么曾暗自担心，共折叠（co-folding）会令他们失去竞争力。

• ChEMBL 中的聪明汉斯现象：模型可以从分子结构猜出其化学家，而“是谁做出来的”对活性的预测能力几乎与结构本身一样强。

• Hermes 仅使用 Leash 自己的数据训练，却能在公开基准上与 Boltz 齐头并进。此外，还有 Boltz 早已在 PDB 中见过的 BRD4 分子。

• 打破口袋：在发现 Hermes 记住的是蛋白质的“昵称”而不是结合位点后，如何大规模构建蛋白质突变体。

• 类似 Lumacaftor 的命中分子，如何意外成为错误折叠蛋白质的检测器。

• Agent 沙盒：LLM 用自然语言推理构效关系（SAR），在 650 万个分子中寻找最佳结合物，而预算仅为 100 次尝试。

• Recursion 帮，以及为什么其校友创办的公司名称都很押韵。

• Kosuri 悖论：如果你能收集数据，为什么还需要模型？

• 在一个炎症项目中，如何在六个月内从 DEL 命中物推进到具有良好小鼠 PKPD 的候选物；以及在细胞因子靶点上，相比抗体，口服药的理由。

• 什么能阻止资金更充足的实验室以更大规模超越他们，以及 Ian 希望通过“行政命令”实现的大规模并行纯化。

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’s 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. 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.