AI Drug Discovery and Protein Computation
Analyzing bottlenecks and real-world deployment in binding prediction and closed-loop drug discovery.
3 sources · ↑1.1×Voices in this topic
All paragraphs
Oct 26, 2026
- Ian Quigley Leash Bio makes scientific-agent decisions auditable by premeasuring the full landscape. Agents get a target, a catalog, and a budget of 100 molecules to find the best binder; metabolism and other drug-like properties are outside …
- Better models’ clearest behavioral edge is leaving unproductive chemistry sooner. Agents may repeat tests, choose molecules at odds with their hypotheses, or keep modifying poor binders; stronger models abandon “crappy …
- Astra 6 leads, while expensive models can still underperform. Quigley also rates Sonnet 5.5 and Sol 6.1 well, and says DeepSeek V4 Flash was “pretty good.” Sonnet 5 is “both very expensive and not very …
- Improved chemistry fluency changed what Leash could attempt with general models. After poor SMILES handling last summer, models began working meaningfully with chemistry around February. Removing molecular information …
- Leash’s scaling plan reuses measured experiments across vast numbers of agent rollouts. Quigley calls it “a lab in the loop with amortized latency,” with choices restricted to measured molecules. Leash has screened about 1,000 …
- Transfer to new biology remains unproven, and a proposed baseline is untested. Quigley does not know whether reinforcement-learning progress will transfer. Asked about a Bayesian hyperparameter algorithm as a …
Oct 25, 2026
Oct 23, 2026
Oct 20, 2026
Oct 19, 2026
- 王晟 王晟的核心论点是，AI 制药的瓶颈不在智能，而在验证——“我们真正关心的问题都没办法用 Benchmark 衡量，也没有答案”。 Astra 在 Terminal-Bench Science 只拿到 60 多分，但王晟认为它做得比自己好；AI 在 Coding 等 Dry Lab 任务上进步很大，但实验方案设计的完成度可能只有约 30%。老鼠实验需数周、猴子实验通常需数月，现实反馈周期远慢于模型迭代；这也是 …
- 湿实验数据可能成为 AI 时代的高价值资产，也可能催生新生意：一条关于“分子在老鼠身上到底能不能杀掉癌细胞”的数据，可能价值几万美元到几十万美元，在大模型预训练语料里看不到。 Anthropic 与 Adaptyv Bio 合作，以 MCP 形式把预测接入湿实验验证；王晟还提到 Anthropic 正在布局湿实验团队。药物公司做完猴子实验后可以把项目授权给大药厂，授权费通常在 $10M 以上。
Oct 15, 2026
Oct 14, 2026
- Latent Space Leash’s reported advantage is inexpensive, carefully characterized experimental data rather than an elaborate model architecture. RJ Honicky describes the shop as a ten-person operation; Ian Quigley says Leash has developed about 2,000 protein constructs and run roughly …
- Ian Quigley Leash’s Kaggle competition exposed how quickly predictive performance disappears when chemistry genuinely changes. Its roughly 4-billion-interaction release dwarfed the then-largest aggregated binding database, but none of approximately 2,000 competing …
- Chemical benchmarks may reward recognition of scientists’ habits rather than understanding of physics. A classifier identified the correct chemist among its top five guesses with 60% accuracy across 1,815 chemists; representing molecules only …
- Hermes demonstrates some external transfer, while its failure modes still shape Leash’s experiments. Trained exclusively on internally generated data, the simple embedding-and-cross-attention model performed about as well as Boltz on binding …
- RJ Honicky The newer opportunity is a physically grounded reasoning sandbox operating at database speed. Ian describes agents starting with ten binders and ten nearby non-binders, then choosing molecules to test their hypotheses. RJ Honicky says …
- Ian Quigley Binding predictions remain constrained by chemical coverage, DNA effects and incomplete potency calibration. Off-DNA resynthesis works for about 90% of easier hits and 50% of harder ones in Leash’s hands, while sequencing readouts support relative …
Oct 12, 2026
Oct 10, 2026
- Pushmeet Kohli AlphaFold 2 reproduced a valuable experimental target; it did not explain protein behavior in full. Pushmeet Kohli says the model learned to recreate structures deposited in the PDB. The discussion leaves the true ground state, structural …
- Sal Candido The relevant scaling law depends on having data with the right statistics, not merely more observations. Sal Candido’s counterexample is metagenomic sequence data: much of it may not represent complete proteins, yet it improves models for …
- Data generation and modeling are complements whose allocation should follow the problem. Sal Candido criticizes method-first identities—“I am a modeling specialist” or “I am a data-generation specialist”—as almost religious. He …
- Brandon Anderson Scientific priors remain valuable, particularly when data is scarce, but incorrect inductive bias can eventually become a constraint. Pushmeet says AlphaFold’s design used biophysical intuition that residues influence one another, giving it an “unfair advantage” in data …
- Latent Space The next frontier is richer biological context, not another isolated structural snapshot. Kohli’s “magic wand” would model raw cryo-EM micrographs directly to recover distributions and dynamics that processed structures may …
- Pushmeet Kohli Model reliability matters more immediately than complete human-readable mechanistic interpretation. AlphaFold 2 achieved about 90 GDT on the cited dataset, but Kohli argues even 95 GDT would be dangerous without calibrated pLDDT: …
