Bibliographic Details
| Title: |
De novo design of high-affinity protein binders with AlphaProteo |
| Authors: |
Zambaldi, Vinicius, La, David, Chu, Alexander E., Patani, Harshnira, Danson, Amy E., Kwan, Tristan O. C., Frerix, Thomas, Schneider, Rosalia G., Saxton, David, Thillaisundaram, Ashok, Wu, Zachary, Moraes, Isabel, Lange, Oskar, Papa, Eliseo, Stanton, Gabriella, Martin, Victor, Singh, Sukhdeep, Wong, Lai H., Bates, Russ, Kohl, Simon A., Abramson, Josh, Senior, Andrew W., Alguel, Yilmaz, Wu, Mary Y., Aspalter, Irene M., Bentley, Katie, Bauer, David L. V., Cherepanov, Peter, Hassabis, Demis, Kohli, Pushmeet, Fergus, Rob, Wang, Jue |
| Publication Year: |
2024 |
| Collection: |
Quantitative Biology |
| Subject Terms: |
Quantitative Biology - Biomolecules |
| More Details: |
Computational design of protein-binding proteins is a fundamental capability with broad utility in biomedical research and biotechnology. Recent methods have made strides against some target proteins, but on-demand creation of high-affinity binders without multiple rounds of experimental testing remains an unsolved challenge. This technical report introduces AlphaProteo, a family of machine learning models for protein design, and details its performance on the de novo binder design problem. With AlphaProteo, we achieve 3- to 300-fold better binding affinities and higher experimental success rates than the best existing methods on seven target proteins. Our results suggest that AlphaProteo can generate binders "ready-to-use" for many research applications using only one round of medium-throughput screening and no further optimization.
Comment: 45 pages, 17 figures |
| Document Type: |
Working Paper |
| Access URL: |
http://arxiv.org/abs/2409.08022 |
| Accession Number: |
edsarx.2409.08022 |
| Database: |
arXiv |
