Claude autonomously designs protein binders with 93% success rate
ycombinator · x · 2026-08-19
Anthropic research shows Claude 3.5 Sonnet successfully designed novel protein binders for 14 out of 15 targets (93% success rate) in a de novo design task. Guided by expert prompts, the designs were independently validated by wet labs (Adaptyv Bio & Twist Bioscience) for high affinity and structural accuracy.
Observers highlight the democratization of protein design: combining open-source tools like ESMFold2 with capable LLMs like Claude allows for deep, complex biological workflows previously restricted to human experts.
Related event: Anthropic Says Claude Designed Protein Binders for 14 of 15 Targets(45 posts)→
More from Research
- One year of Recursive Language Models: GPT-5-mini beats GPT-5 by 110% on 132k-token tasks, cheaper too — a1zhang · 2026-10-16
- MIT PhD Sharut Gupta joins DeepMind; Isola hails her representational alignment defense talk — phillip_isola · 2026-10-16
- 36 primitives, just 10 parameters each: a remarkably precise T-pose body fit — letandrewcook · 2026-10-16
- Interp researcher pushes back: if you can't simulate a neuron, you haven't explained it — nabla_theta · 2026-10-16
- MLP-Splatting (ECCV 2026) code released: MLPs as light-field primitives instead of Gaussians — AjdDavison · 2026-10-16
- Four-page solution to subcubic APSP builds intuition for Anthropic-AI-aided breakthrough — gautamcgoel · 2026-10-16