Researchers from Australia and China have made a groundbreaking discovery using artificial intelligence, identifying approximately 162,000 new virus species. This collaborative effort involved institutions such as the University of Sydney and the Chinese National Natural Science Foundation.
According to a statement from the University of Sydney, the team utilized AI technology to analyze genetic material, specifically RNA, leading to the identification of 161,979 new virus types. They developed a deep learning algorithm called LucaProt to process genetic sequence data, including viral genomes extending up to 47,250 nucleotides.
While many of the newly discovered viruses are already documented in existing databases, the use of AI allowed researchers to organize and categorize this information more effectively. Edwards Holmes, one of the study's authors, described the findings as a "window into a hidden part of life on Earth," emphasizing that this study has revealed the highest number of new virus species in a single research effort.
Holmes noted that millions more virus species remain to be discovered, suggesting that the same AI techniques could also be applied to identify bacteria and parasites. Another author, Mang Shi, highlighted that previous bioinformatics pipelines limited the diversity of discoveries, but the new AI model enables a much deeper exploration of viral diversity, with plans for various applications.