RIYADH, March 28, 2025 – In a pioneering scientific achievement, researchers have harnessed artificial intelligence to develop a powerful new tool in the battle against microplastic pollution — microscopic protein fragments called peptides.
These AI-engineered peptides can bind to plastic particles, making the removal of microplastics from water more efficient.
Microplastics — tiny plastic particles smaller than five millimeters — are now widespread in oceans, rivers, soil, and even in the human body. These particles persist for centuries in the environment, posing a threat to ecosystems and public health.
Traditional methods of cleanup have struggled to tackle this growing issue. However, a team of scientists from Saudi Arabia, the US, and other countries has introduced a groundbreaking solution: biodegradable peptides that bind to microplastic particles with remarkable precision.
Developed using advanced deep learning models, these peptides could represent a significant turning point in the global fight against plastic pollution.
“We combined advanced biophysical simulations, showing how short proteins behave on plastic surfaces, with deep learning or AI to uncover hidden patterns in those simulations,” explained Abdulelah Al-Shehri, assistant professor of chemical engineering at King Saud University and co-author of the study.
“This allowed us to identify specific peptides that bind to microplastics up to 34 percent more effectively than traditional methods.”
Unlike conventional filtration techniques, these AI-designed peptides offer a scalable, biodegradable alternative that could revolutionize microplastic cleanup.
“AI guided us to protein sequences that traditional methods might overlook, resulting in stronger and more efficient cleanup capabilities,” Al-Shehri added.
While the discovery was initially made through computational models, laboratory tests have confirmed the peptides’ real-world effectiveness.
“We recently conducted experiments to evaluate how strongly the AI-designed peptides bind to plastic,” said Michael Bergman, a PhD candidate at North Carolina State University’s Department of Chemical and Biomolecular Engineering.
“No one has ever designed plastic-binding peptides before, and we relied solely on computational predictions for peptide design, so we were eager to see how our predictions would perform in the lab.”
“Excitingly, the AI-designed peptides performed exceptionally well, binding much more strongly to plastic than random amino acid sequences and matching the performance of our best biophysical designs.”
“This work will be published soon, and the next step is to apply these peptides in tackling microplastic pollution.”
