AI-designed artificial metalloenzyme achieves 97:3 enantioselectivity, over 95% yield in Nature
bravo_abad · x · 2026-10-22
Mishra et al. report in Nature a de novo-designed protein built around a synthetic metal photocatalyst. Using RFdiffusion All-Atom to design the catalyst-binding pocket and LigandMPNN for sequence design, followed by directed evolution, the resulting DE3 variant produced mirror-image products in a 97:3 ratio with over 95% yield — a full AI protein design + evolution route to enantioselective artificial enzyme photocatalysts.
More from Research
- XDOF Benchmark: Top LLM Agents Score Just 9-19% on Physical Tasks — mihdalal · 2026-10-22
- NVIDIA releases Cosmos Simulation: real-time interactive world models for Physical AI — CSProfKGD · 2026-10-22
- Zeroing 7 neurons cuts Qwen3-8B sycophancy from 76% to 0% in Opus-assisted experiment — Sauers_ · 2026-10-22
- Zeroing 6 neurons cuts Qwen3-8B sycophancy from 44% to 0%, interpretability charts show — Sauers_ · 2026-10-22
- NYT Daily: AI solved a Millennium Problem and 350+ results, mathematicians grieve their 'hero era' — lennysan · 2026-10-22
- Thinking Machines Ships StableReinforce, a Loss That Keeps Off-Policy RL From Collapsing — LiTianleli · 2026-10-22