South Korean researchers have developed a technology that could dramatically improve the efficiency of protein analysis essential for new drug development. The breakthrough is expected to strengthen Korea's competitiveness against Google DeepMind's AlphaFold in the artificial intelligence (AI) bio-technology race.
The Ministry of Science and ICT announced Wednesday that a research team led by Professor Martin Steinegger of Seoul National University's School of Biological Sciences has developed "Foldmason," an ultra-fast, high-precision multiple alignment analysis technology based on large-scale protein structure big data. The research was published in the international journal Science on the same day.
Foldmason is software that comprehensively analyzes the three-dimensional structures of proteins and the sequence information of amino acids that comprise them. Proteins are substances that perform various biological functions in the body depending on their type. For new drug development, researchers must analyze protein abnormalities that cause diseases and discover substances that can correct them. As protein analysis is the first step in drug development, AI and software tools that can improve work efficiency, like DeepMind's AlphaFold, are being developed in succession.
Foldmason can rapidly compare hundreds of thousands of protein structures simultaneously, making it particularly effective at analyzing the so-called "twilight zone" that has been difficult to study. The twilight zone refers to regions where amino acid sequence similarity is so low that it is difficult to reliably determine homology or evolutionary relationships using conventional sequence-based methods. Protein comparison is performed through a process called "alignment," and existing sequence-based alignment methods face limitations in sequence similarity when differences between proteins are significant. Foldmason can operate 100 to 1,000 times faster than existing alignment tools.
Using Foldmason, the research team confirmed that the blueprint of core proteins that fight viruses has remained virtually unchanged for billions of years, even between completely different organisms such as humans and bacteria. Along with Foldmason's development, they uncovered an important clue to understanding where the human immune system originated.
"This research presents new possibilities for tracking protein evolution on a large scale over billions of years," Professor Steinegger said. "We will contribute to identifying functional differences in disease-related proteins and discovering new drug targets through large-scale structural variation analysis."