Listen to this article Pro Gen Creates Original Proteins Using Artificial Intelligence
Researchers have developed an AI system called Pro Gen that can generate artificial enzymes from scratch with efficacy comparable to natural enzymes. Using the Pro Gen AI system, the experiment shows that the field of biology can be understood through natural language processing, paving the way for a more potent protein engineering technology surpassing directed evolution. By accelerating the development of new proteins, it will invigorate the field of protein engineering, enabling their use in a wide range of applications, from therapeutics to plastic degradation.
Pro Gen Ignites Protein Design Using Natural Language Model
A natural language model has jumpstarted the process of protein design by creating active enzymes. Salesforce Research developed an AI program called Pro Gen that constructs artificial proteins from amino acid sequences using next-token prediction. Although originally designed for language text, the language model can now generate functional proteins.
AI System Creates Artificial Enzymes from Scratch
In laboratory experiments, Pro Gen generated enzymes that demonstrated efficacy comparable to natural enzymes. Moreover, these artificially created enzymes had amino acid sequences that greatly deviated from any known natural protein. This suggests that Pro Gen has the potential to create entirely novel proteins with unique functions. To create the model, scientists fed the amino acid sequences of 280 million different proteins of all kinds into the machine learning model and let it digest the information for a couple of weeks. Then, they fine-tuned the model by priming it with 56,000 sequences from five lysozyme families, along with some contextual information about these proteins.
AI System Grasps Fundamental Concepts of Biology
The experiment proves that natural language processing can comprehend specific fundamental concepts of biology, even though the technology was initially designed for language tasks. The AI, known as Pro Gen, could learn how to shape enzymes by studying raw sequence data. X-ray crystallography revealed that the atomic structures of the artificial proteins generated by Pro Gen appeared normal, despite their unique amino acid sequences never seen before.
AI system creates limitless functional proteins
The AI system has the remarkable ability to create artificial proteins with endless possibilities. Lysozymes contain up to 300 amino acids, making them relatively small compared to other proteins, but the staggering number of possible combinations resulting from the 20 available amino acids (20^300) is still significant. Despite this, the model can effortlessly generate functional enzymes. It’s quite remarkable that a technology designed for language tasks can accomplish such a feat.
New protein engineering technology, Pro Gen, is set to revolutionize the field and surpass directed evolution. It accelerates the development of proteins for diverse applications, from therapeutics to plastic degradation. Pro Gen’s unique capability to generate functional proteins from scratch is ushering in a new era of protein design. It is a versatile new tool available to protein engineers, and researchers are looking forward to seeing the therapeutic applications.