De novo design of high-affinity protein binders with AlphaProteo

Vinicius Zambaldi,David La,Alexander E. Chu,Harshnira Patani,Amy E. Danson,Tristan O. C. Kwan,Thomas Frerix,Rosalia G. Schneider,David Saxton,Ashok Thillaisundaram,Zachary Wu,Isabel Moraes,Oskar Lange,Eliseo Papa,Gabriella Stanton,Victor Martin,Sukhdeep Singh,Lai H. Wong,Russ Bates,Simon A. Kohl,Josh Abramson,Andrew W. Senior,Yilmaz Alguel,Mary Y. Wu,Irene M. Aspalter,Katie Bentley,David L.V. Bauer,Peter Cherepanov,Demis Hassabis,Pushmeet Kohli,Rob Fergus,Jue Wang

2024-09-12

Abstract: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.

Biomolecules

What problem does this paper attempt to address?

The paper aims to address the problem of designing high-affinity protein binders in protein-protein interactions. Specifically, the paper introduces a machine learning model system called AlphaProteo for the de novo design of protein binders. While current methods have made some progress on certain target proteins, they still face challenges, especially in designing high-affinity binders without requiring multiple rounds of experimental optimization. Through the AlphaProteo system, researchers are able to achieve the following: 1. **High Success Rate**: Screening stable, highly expressed, and highly specific binders from dozens of design candidates, thereby reducing the need for high-throughput screening methods. 2. **High Affinity**: For most tested targets, the best binders achieve affinity levels in the picomolar (pM) to low nanomolar (nM) range, reducing the labor required for subsequent affinity optimization. 3. **Versatility**: The method is applicable to a variety of target proteins with different structures and biochemical properties, achieving successful binders with just one round of design without the need for complex manual intervention. The paper details the performance of AlphaProteo on 8 structurally diverse target proteins and compares it with existing best methods, showing higher experimental success rates and better binding affinities. Additionally, the paper validates the functionality and structural accuracy of these binders.

De novo design of high-affinity protein binders with AlphaProteo

Improving de novo protein binder design with deep learning

Robust de novo design of protein binding proteins from target structural information alone

De novo design of high-affinity binders of bioactive helical peptides

BindCraft: one-shot design of functional protein binders

ProteinInvBench: Benchmarking Protein Inverse Folding on Diverse Tasks, Models, and Metrics.

De novo design of protein interactions with learned surface fingerprints

Design of High Affinity Binders to Convex Protein Target Sites

Improved protein binder design using ꞵ-pairing targeted RFdiffusion

AlphaDesign: A de novo protein design framework based on AlphaFold

De novo design of drug-binding proteins with predictable binding energy and specificity

De Novo Design of Peptide Binders to Conformationally Diverse Targets with Contrastive Language Modeling

AlphaBind, a Domain-Specific Model to Predict and Optimize Antibody-Antigen Binding Affinity

Peptide binder design with inverse folding and protein structure prediction

Accurate de novo design of high-affinity protein binding macrocycles using deep learning

ProBID-Net: A Deep Learning Model for Protein-Protein Binding Interface Design

De novo design of protein structure and function with RFdiffusion

Target-Specific De Novo Peptide Binder Design with DiffPepBuilder

Design of Peptide Binders to Conformationally Diverse Targets with Contrastive Language Modeling

Recent Progress Using De Novo Design to Study Protein Structure, Design and Binding Interactions

Expanding the Sequence Spaces of Synthetic Binding Protein Using Deep Learning-Based Framework ProteinMPNN