Scientists from the University of Wisconsin-Madison have developed a robotic platform managed by artificial intelligence that significantly speeds up the process of creating new proteins. This discovery could have wide applications in medicine, chemistry, material science, and biotechnology (source).
The technology, called Self-Driving Autonomous Machines for Protein Landscape Exploration (Sample), works by teaching AI models to understand the connections between protein sequence and function. These models design new proteins and send them to be tested by a robotic system. The data obtained from the tests are then used to refine the AI algorithm.
In a recent study, the team tested the ability of four Sample agents to improve the heat resistance of glycosyl-hydrolasis enzymes. Despite their different behaviors, all agents were able to quickly create heat-resistant enzymes.
The researchers highlight that automated laboratories have the potential to accelerate scientific discovery in protein engineering and synthetic biology. They estimate that with the help of Sample, protein engineering could be achieved in just a few weeks, compared to the 6 to 12 months it would take for a person. If Sample meets expectations, it could mark the beginning of a new era in creating customized proteins for various scientific and technological fields.
The article also emphasizes the combination of artificial intelligence and automation in various industries, including manufacturing, food production, pharmaceuticals, agriculture, and waste management. Automated laboratories are set to revolutionize biomolecular engineering and synthetic biology, automating labor-intensive and costly processes involved in protein creation, allowing scientists to focus on further applications.