New AI model, co-developed by Evozyne and NVIDIA, can create novel therapeutic proteins for a new era of drug development
Evozyne today revealed a breakthrough AI model, developed in collaboration with NVIDIA, that the pharmaceutical industry can use to design therapeutic proteins.
Presented by NVIDIA at the J.P. Morgan Healthcare Conference in San Francisco, the novel approach can exponentially increase the number and quality of synthetic protein designs, which opens up the possibility to design new therapeutic targets for untreatable diseases and greatly reduces development time for new treatments.
Evozyne began a collaboration with NVIDIA in 2022 to develop a new deep learning model that can learn the rules of protein function and use these rules to design new proteins with improved functions. The model, known as the Protein Transformer Variational AutoEncoder (ProT-VAE), is built on NVIDIA BioNeMo, a framework for efficiently training and deploying large language models for biology. The resulting model represents a new deep learning technology that enables advances for data-driven design of synthetic proteins with engineered functions. More information on the project is available here.
“This result achieves the long-standing goal of creating new protein sequences that perform better than proteins currently found in nature. For therapeutics in particular, the discovery opens our imagination to creating new therapeutics for untreatable diseases and it can compress research and development time because it uses machine learning to identify the most likely therapeutic candidates,” said Andrew Ferguson, Evozyne scientific co-founder.
“The field of biology is quickly fusing science and engineering using the latest breakthroughs in generative AI,” said Kimberly Powell, vice president of healthcare at NVIDIA. “NVIDIA BioNeMo is a fundamental part of the ProT-VAE large language model and Evozyne’s platform, which is paving the way for machine learning-guided protein engineering resulting in synthetic functional proteins that can be used in new therapies, energy sources, materials and beyond.”
Evozyne scientists initially targeted the PAH gene, which provides instructions for making an enzyme called phenylalanine hydroxylase. Deficiencies in this gene can lead to a genetic disorder.
Using the Evozyne ProT-VAE model and NVIDIA BioNeMo framework, the team designed new synthetic PAH variants predicted to have “supernatural” function. Subsequent laboratory testing revealed that several of these variants perform up to 2.5-times better than the natural enzyme. These proteins represent potential gene therapies to replace the missing digestive enzyme.
Evozyne was founded in 2020 by Paragon Biosciences to develop novel proteins for critical applications in therapeutics and sustainability. It is creating technology platforms for gene optimization, antibody discovery and gene editing, along with potential applications in sustainability. In its first 18 months, Evozyne engineered proteins that significantly outperformed design specifications for its customers.
Visit AITechPark for cutting-edge Tech Trends around AI, ML, Cybersecurity, along with AITech News, and timely updates from industry professionals!