International Researchers Leverage AI To Engineer Smart Protein Switches For Real Time Molecular Sensing

QUT researchers use AI to create smart protein switches that detect target molecules, paving the way for low-cost medical and environmental biosensors.

By: AXL Media

Published: Apr 15, 2026, 7:37 AM EDT

Source: Information for this report was sourced from EurekAlert!

The Dawn of AI Designed Molecular Machines

In a significant leap for synthetic biology, an international collaboration has successfully used artificial intelligence to engineer "smart" proteins that act as precise molecular switches. These tiny devices are designed to remain inactive until they detect a specific target, at which point they trigger a measurable response. Professor Kirill Alexandrov of the QUT School of Biology and Environmental Science noted that while protein engineers were previously limited to modifying naturally occurring proteins, AI now allows for the creation of sensors on demand. This advancement effectively expands the toolkit available to scientists for building custom molecular machines capable of sensing environmental changes.

Challenging Traditional Views on Protein Dynamics

The study’s findings provide a major correction to a long-held belief in protein science: that sensors must undergo dramatic physical shape changes to function. Instead, the researchers discovered that AI-designed receptors can operate through subtle shifts in how the protein moves, a process that is sufficient to turn activity on or off. This insight into "subtle movement" regulation not only simplifies the design process for new biosensors but also offers a fresh perspective on how natural protein systems might be regulated within living organisms. By focusing on these refined dynamics, the team has bypassed the need for complex structural rearrangements.

Versatile Outputs for Diverse Sensing Needs

To ensure these switches are commercially and scientifically viable, the team connected the artificial receptors to enzymes capable of producing several types of outputs. These include visible color changes, light emission, and electrical signals. This versatility allows the technology to be integrated into various formats, from simple colorimetric tests to sophisticated electrochemical sensors similar to modern glucose meters. In their experiments, the researchers successfully demonstrated these switches responding to a wide array of triggers, including small molecules, peptides, and proteins, highlighting their broad potential for the diagnostic industry.