Generative AI designs thiolation domains that triple NRPS peptide yields
Researchers built and tested 578 recombinant synthetase variants in vivo, with one design showing a 12 °C higher melting temperature and better soluble expression.
PR
By Priya Raman · Science Reporter
Sep 22 13:48 ETSource: Nature
The wire
- 01A Nature paper combined ESM3, ProteinMPNN and EvoDiff with design-build-test-learn cycles to generate 76 de novo thiolation domains, tested across 578 recombinant non-ribosomal peptide synthetase variants.
- 02Non-ribosomal peptide synthetases make many clinically used peptide therapeutics, so reprogrammable assembly lines could widen manufacturing routes for novel peptide drugs.
- 03Work is engineering-stage in vivo expression data, with yield gains up to roughly 3-fold; no therapeutic candidate or clinical application is reported.
From the source
Read at nature.comGenerative AI designs functional thiolation domains for reprogramming non-ribosomal peptide synthetases
Large language models and generative protein design promise to accelerate biotechnology, but it remains unclear whether they can engineer dynamic megasynth(et)ases whose activity depends on transient, context-specific domain interfaces. Non-ribosomal peptide synthetases (NRPSs) exemplify this challenge and produce many clinically used therapeutics. Here we integrate pretrained generative models (E
