Nature study finds AI-redesigned proteins evolve into better enzymes than natural starting points
lucapinello · x · 2026-07-23
AI-redesigned protein starting points improve enzyme evolution
A Nature paper reports a workflow that combines ProteinMPNN sequence design with autonomous laboratory evolution to engineer enzymes with better properties than those evolved from natural proteins.
- The team redesigned three distinct botulinum neurotoxin proteases using AI-guided protein sequence design.
- Those redesigned variants showed improved stability and full catalytic efficiency.
- In side-by-side continuous evolution campaigns, the redesigned proteins consistently produced variants with higher activity than wild-type starting points.
- Across four evolution campaigns, the redesign opened access to otherwise unreachable functional sequences and sped up adaptation.
- In a final demonstration, the AI-redesigned protease achieved higher catalytic efficiency and stability while reducing native substrate cleavage, including 79× greater selected specificity for ataxin-2 than the best WT-derived variant.
The study argues that AI-redesigned starting points can expand the reachable protein fitness landscape and improve downstream enzyme engineering results.
More from Research
- 3D ResNet Paper Crosses 3,000 Citations Eight Years After CVPR 2018 — HirokatuKataoka · 2026-09-11
- Sample selection and ordering matter a lot in LLM training: DataFlex makes data scheduling dynamic — Puzzleheaded_Box2842 · 2026-09-11
- Jeff Heaton's Intro to the Math of Neural Networks eBook Is Free to Download — blaizedsouza · 2026-09-11
- Mathematician Daniel Litt Launches Problem Repo to Track Human vs AI Progress: 15 Problems, 1 Solved — littmath · 2026-09-11
- Open ECDSA.fail challenge uses AI agents to shrink Shor's-algorithm quantum circuits for Bitcoin keys — StefanoGogioso · 2026-09-11
- Alex Townsend posts 200 open problems in numerical linear algebra for humans and AI agents — IgorCarron · 2026-09-11