Вештачка интелигенција

AI Enhances Protein Design through Functionally Directed Engineering

Summary

Protein design advancements bring new hope for the production of efficient protein-peptide lines and the prediction of their potential functionalities. The team at Generate Biomedicines in Massachusetts has developed artificial intelligence (AI) that can generate feasible protein structures and predict […]

AI Enhances Protein Design through Functionally Directed Engineering

Protein design advancements bring new hope for the production of efficient protein-peptide lines and the prediction of their potential functionalities. The team at Generate Biomedicines in Massachusetts has developed artificial intelligence (AI) that can generate feasible protein structures and predict their potential functionalities.

In their work titled “Illuminating Protein Space with a Programmable Generative Model,” published in the journal Nature, the team introduces Chroma, an intelligent and versatile AI capable of producing diverse proteins with specific properties.

Proteins are amino acid compounds that make up all living organisms and perform most of the essential functions necessary for life. Nearly all pharmaceutical drugs target proteins in the body, and many drugs are either proteins themselves or are produced with the help of proteins.

With proteins as the foundation of human health and pharmaceutical solutions for diseases, the ability to create new proteins with specific properties could unlock a range of new drug targets. More flexible control over proteins could improve the safety of existing drugs and enable treatments for diseases that currently have no cure.

Chroma’s uniqueness lies in its programmability, which allows users to specify various characteristics, from the distance between amino acids to semantic specifications through classifiers.

The team conducted experimental validation tests to challenge Chroma’s effectiveness in generating well-expressed protein designs that exhibit stable folding and structural alignment with the intended design.

Based on the most concrete assessment, Chroma has shown a success rate of approximately 3% in the successful characterization and purification of proteins. Chroma has demonstrated the ability to generate proteins with diverse structures and properties, handle complex shapes, and showcase effective designs, highlighting its potential for customized protein engineering.

Many incurable diseases have protein targets in the body that are too complex or challenging for existing pharmaceutical proteins to bind to. By reconstructing the properties necessary for binding targeted protein targets as a starting point for generative protein synthesis, researchers could discover treatments for thousands of diseases. Efforts are currently underway to generate sequences of feasible protein structures and then search for potential target matches.

The possibilities offered by Chroma can shift the focus from generating feasible protein structures to emphasizing the intended functionality of proteins and forcing structural formation to follow that intended function.

FAQ