A revolutionary AI tool developed at the University of Missouri could significantly advance the treatment of cancer and other diseases. The tool uses cryo-electron microscopy images to reveal the structure of protein complexes, opening the door to more effective diagnostics and therapies.
In a significant breakthrough, researchers at the University of Missouri are leveraging artificial intelligence to decode the intricate dance of proteins within cells, a discovery that holds great promise for revolutionary advancements in cancer treatment and other medical fields.
Jianlin “Jack” Cheng, a Curators’ Distinguished Professor of Electrical Engineering and Computer Science, alongside his student Nabin Giri, has unveiled Cryo2Struct. This innovative computer program uses artificial intelligence to construct detailed 3D models of large protein complexes from cryo-electron microscopy (cryo-EM) images. Their findings were recently published in the journal Nature Communications.
“Cryo-EM right now is a revolutionary, key technology for determining large protein structures and assemblies in cells,” Cheng said in a news release. “But building protein structures from Cryo-EM data is labor-intensive and requires a lot of human intervention, making it time-consuming and hard to reproduce. Our technique is fully automated and generates more accurate structures than existing methods.”
Proteins are the fundamental building blocks of life, starting as simple chains of amino acids that fold into complex three-dimensional structures. These intricate shapes dictate their vital functions in the body. For decades, scientists have struggled to fully comprehend this folding process.
Cheng’s pioneering work in applying deep learning to this problem back in 2012 marked a turning point, demonstrating that AI could predict protein structures. This laid the groundwork for tools like Google’s AlphaFold, which is renowned for its accuracy in protein structure prediction.
However, understanding a single protein is only part of the challenge. In living organisms, proteins operate together in complex assemblies, functioning like molecular machines to perform critical biological tasks. Unraveling these protein interactions is pivotal for understanding disease mechanisms and developing effective treatments.
Cryo2Struct is akin to a master detective, piecing together the molecular puzzle without prior knowledge. It analyzes cryo-EM images to pinpoint individual atoms in a protein complex and assembles these into a cohesive 3D model. This comprehensive view offers profound insights into protein behavior and interactions.
“Our technology enables scientists to determine and build a structure from cryo-EM data,” Cheng added. “Once you have that structure and understand its functions, you can design drugs to counter any faulty functions of a protein complex to make it function properly.”
In an associated study published in Chemistry Communications Cheng and student Alex Morehead delved into another AI application, the diffusion model. This approach models the transformation of molecular structures from random noise to well-defined forms, potentially aiding in drug design and optimization.
“For instance, I have a drug, and I want to make it work better for some patients,” Cheng added. “Now I can use AI to change it and optimize it.”
The potential impact of Cryo2Struct and similar AI tools extends far beyond academic research, holding promising implications for the future of personalized medicine and the treatment of complex diseases like cancer. By automating and refining the study of protein structures, these innovations could lead to more effective diagnostics, targeted treatments and, ultimately, better patient outcomes.
For those in the scientific and medical communities, this breakthrough demonstrates the transformative power of AI in tackling some of the most stubborn challenges in health care and biology.