AI and robots join forces to cook up proteins faster
Applications across chemistry, energy, and medicine await human-free acceleration
Scientists have developed a platform based around a robot guided by an AI-based computer system, which could slash the time for engineering new proteins from months to weeks.
Researchers are finding broad applications for newly engineered proteins that could find roles in medicine, chemistry, materials science, and biotechnology. However, developing the large biomolecules can be maddeningly time-consuming.
A team led by Philip Romero, assistant professor of protein engineering and computational biology at University of Wisconsin–Madison, has developed a platform designed to accelerate the process engineering new proteins without requiring human intervention.
They named the robotic platform Self-driving Autonomous Machines for Protein Landscape Exploration (SAMPLE). The researchers claim the models learn relationships between protein sequence and function, and design new proteins before sending them to a robotic system for testing. Data from the robotic system is then fed back to the AI algorithm to refine the model.
The authors tested the four SAMPLE agents' efforts to engineer glycoside hydrolase enzymes while improving their heat tolerance.
"Despite showing individual differences in their search behavior, all four agents quickly converge on thermostable enzymes. Self-driving laboratories automate and accelerate the scientific discovery process and hold great potential for the fields of protein engineering and synthetic biology," the authors said in a paper published in Nature Chemical Engineering.
- Here's how fast a spacecraft should fly to successfully detect amino acids erupting from Enceladus
- Pharma boffins sharpen hunt for target molecules using graph DB
- Beyond the hype, AI promises leg up for scientific research
- Boffins develop AI model for designing proteins to make synthetic blood plasma
Based on the study, researchers estimate that the platform could engineer these proteins in a few weeks, while human scientists would probably require six to 12 months to achieve the same results. If SAMPLE lives up to its promise, it could herald an era of proteins on demand, they said.
"Protein engineering has nearly limitless applications across chemistry, energy and medicine, but creating new proteins with improved or novel functions remains slow, labor-intensive and inefficient," the paper said.
"The powerful combination of artificial intelligence and automation is disrupting nearly every industry, from manufacturing and food preparation to pharmaceutical discovery, agriculture and waste management. Self-driving laboratories will revolutionize the fields of biomolecular engineering and synthetic biology by automating highly inefficient, time-consuming and laborious protein engineering campaigns, enabling rapid turnaround and allowing researchers to focus on important downstream applications."
The Register reported developments in the fast-advancing field in 2018 when a study showed machine learning would help robots perform chemistry experiments faster than humans.
Rather than a humanoid robot capering about the lab, the system contained a series of pumps and reactors all attached to a mass spectrometer, a nuclear magnetic resonance (NMR) spectrometer, and an infrared spectrometer. ®