De novo design of high-affinity protein binders to bioactive helical peptides
Problem
Designing proteins that bind short helical peptides (such as peptide hormones) with high affinity was an unmet challenge in de novo protein design.
Why it matters
Sensitive detection methods for peptide hormones could improve clinical management of patients.
What the AI did
The RF diffusion generative model designed protein binders either de novo from random noise or by noising/denoising lower-affinity designs, producing picomolar-affinity binders.
Results
Picomolar-affinity binders to helical peptide targets; claimed highest-affinity binders generated purely by computation without experimental optimization.
Impact
Demonstrates generative AI can computationally design high-affinity binders to previously intractable helical peptide targets.
Record
| Venue | bioRxiv (Cold Spring Harbor Laboratory) |
|---|---|
| Published | 2022 |
| Work type | preprint |
| Authors | Susana Vázquez Torres, Philip J. Y. Leung, Isaac D. Lutz, Preetham Venkatesh, Joseph L. Watson, Fabian Hink, Huu‐Hien Huynh, Andy Hsien-Wei Yeh et al. |
| Citations | 22 (OpenAlex, at import time) |
| References | 31 |
| Model or method | RF diffusion |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1101/2022.12.10.519862 |
| Open access | green |
| Keywords | protein design, RF diffusion, deep learning, helical peptides, picomolar binders |
Importance rating
significant — thousands to millions affected
De novo design of high-affinity binders to helical peptides is a genuinely new capability in protein design with potential clinical diagnostic and therapeutic use.