Jessie A. Ellis
January 14, 2025 04:31
NVIDIA has unveiled BioNeMo Blueprint, a new AI-based approach that leverages GPU-accelerated microservices to accelerate protein binder design in drug discovery.
In a groundbreaking development in drug discovery, NVIDIA has introduced BioNeMo Blueprint, a comprehensive workflow designed to accelerate the protein binder design process. According to a blog post from NVIDIA, this innovative approach leverages generative AI and GPU-accelerated microservices to greatly simplify the traditionally laborious and time-consuming process of designing therapeutic proteins.
Challenges of protein design
The design of therapeutic proteins that can specifically bind to target molecules is an important yet challenging aspect of drug discovery. Existing methods often involve extensive trial and error, requiring the synthesis and validation of thousands of candidates, which can take years to complete. Considering the complexity of human proteins, which average 430 amino acids in length, the design possibilities are virtually endless, so efficiently exploring this vast search space is a Herculean task.
Introducing NVIDIA BioNeMo Blueprints
BioNeMo Blueprint aims to revolutionize this process by providing a reference workflow for drug discovery platforms. Leverages generative AI to intelligently explore massive search spaces to guide researchers to find stable and structurally constrained protein binders. This significantly reduces the number of iterations and time required to find viable candidates.
Take advantage of advanced AI and GPU technologies
The workflow begins with the amino acid sequence of the target protein and utilizes AlphaFold2 to predict its 3D structure. MMseqs2, NVIDIA’s accelerated multi-sequence alignment (MSA) algorithm, enhances this process by providing fast and accurate alignments, allowing researchers to efficiently explore large databases. These advancements make AlphaFold2 NIM 5x faster and 17x more cost-effective than its predecessor.
Following 3D structure prediction, the RFdiffusion AI model explores optimal binding configurations, allowing users to refine search parameters for reliable interactions. RFdiffusion NIM delivers 1.9x faster speeds than the baseline model, improving the efficiency of your design process.
ProteinMPNN then generates and optimizes amino acid sequences for this configuration, ensuring the creation of stable complexes. The final step includes validation using AlphaFold2-Multimer, which ensures stable interactions between binder and target protein, minimizing the risk of experimental failure.
Accelerating Drug Discovery
This integrated approach not only accelerates the design-to-discovery cycle, but also reduces the need for costly and labor-intensive laboratory work. By prioritizing the most promising candidate designs, researchers can focus their resources more effectively and lay the foundation for a faster, more efficient drug discovery process.
For more information about BioNeMo Blueprint, visit the official NVIDIA blog.
Image source: Shutterstock