A 'Highway' for Protein Structure Big Data Alignment Opens... Revolutionizing Disease Research [Reading Science]

Professor Martin Steinegger's SNU Team Develops Ultra-Fast Analysis Platform 'FoldMason'
Structural Alignment Up to 1,000 Times Faster Than Existing Methods... Published in Science

A domestic research team has developed an ultra-fast analysis technology capable of simultaneously tracking the evolutionary process of proteins spanning billions of years. With the advancement of artificial intelligence (AI) technologies leading to an explosion of protein structure data, the ability to precisely compare and analyze this information has now become possible. This breakthrough is expected to dramatically accelerate research into the causes of diseases and the development of new drugs.

The Ministry of Science and ICT announced that the research team led by Professor Martin Steinegger of the Department of Biological Sciences at Seoul National University has developed an ultra-fast, high-precision multiple alignment analysis platform called 'FoldMason,' based on large-scale protein structure big data. The results of this research have been published in the world-renowned journal Science.

Visualization image comparing and aligning the three-dimensional structures of proteins by superimposition. 'FoldMason,' developed by researchers at Seoul National University, can analyze large-scale protein structure data at ultra-high speed. Photo by Getty Images Bank

AI-Driven Protein Flood: 'Comparison' Was the Bottleneck

Proteins perform biological functions through their three-dimensional structures, which are formed by the folding of amino acids. Understanding how these structures have evolved is key to uncovering the causes of diseases and aging.

Recently, advances in AI-based protein structure prediction technologies have produced vast amounts of structural data. However, the speed and accuracy of comparing and aligning this data have been major challenges. In particular, the so-called 'Twilight Zone'-where protein similarities are extremely low-has long remained a limitation for research, as it was nearly impossible to analyze using conventional methods.

Combining Structure and Sequence: Up to 1,000 Times Faster

To address this bottleneck, Professor Steinegger's team developed FoldMason, a new software that integrates three-dimensional structural information and amino acid sequence data for proteins. This platform achieves speeds 100 to 1,000 times faster than existing structural alignment techniques while maintaining high accuracy.

This enables the simultaneous comparison of hundreds of thousands of protein structures, making it possible to conduct large-scale evolutionary analyses across nearly all protein families, including the previously inaccessible Twilight Zone. Using FoldMason, the research team confirmed that the structural blueprints of core virus-response proteins have been preserved for billions of years, even in life forms as distinct as humans and bacteria.

Professor Martin Steinegger, who led the research, stated, "This achievement opens a new path for large-scale tracking of protein evolution," adding, "By precisely analyzing structural differences in disease-related proteins, it may lead to the discovery of new drug targets."

This research was conducted through the Excellent Young Researcher Support Track of the Ministry of Science and ICT. The fact that global talents such as Professor Steinegger from Germany and Cameron Gilchrist, a researcher at the Korea Basic Science Institute from Australia, have produced world-class results within the domestic research environment is also highly significant.