Over 160,000 Types! Chinese Scientists Utilize AI to Discover More RA Viruses

On October 9, a research team led by Professor Shi Mang from Sun Yat-sen University's Medical School, in collaboration with Li Zhaorong's team from Alibaba Cloud, published a groundbreaking paper in Cell that details the application of artificial intelligence (AI) in viral identification. The study reported the discovery of 180 viral supergroups and over 160,000 types of global RNA viruses, including viral "dark matter" that traditional methods failed to detect.

Professor Shi remarked, "AI enhances our ability to accurately, conveniently, and efficiently identify RNA viruses, including pathogens linked to human diseases and those that may cause future outbreaks." He emphasized that RNA viruses, such as the influenza and COVID-19 viruses, are characterized by their RNA-based genetic material. The diverse and complex nature of RNA viruses makes traditional identification methods limited, while AI techniques have opened new avenues for exploration.

Utilizing LucaProt, the research team analyzed 10,487 RNA sequencing data samples from various global biological environments, uncovering over 510,000 viral genomes representing more than 160,000 potential virus types and 180 RNA viral supergroups—a ninefold increase in supergroup classification. Notably, 23 of these supergroups were termed "dark matter" of the viral world, as they could not be recognized through sequence homology methods.

Shi emphasized the significance of this research, stating that it not only identifies viruses with low similarity to known strains but also discovers novel viruses that have never been documented before. This capability is crucial for monitoring potential emerging infectious diseases. "Our method can accurately detect even entirely new viruses with low homology to known ones."

Prior to this study, the International Committee on Taxonomy of Viruses (ICTV) recognized no more than 5,000 species of RNA viruses. The recent findings indicate that the actual diversity of RNA viruses is far greater, exceeding 160,000 types. "These findings reveal that the diversity of viruses in nature is beyond our imagination," Shi noted. He also pointed out that not all discovered viruses are associated with diseases. Many play an essential role in ecosystems, stressing the need for a balanced perspective on the existence of viruses rather than viewing them solely with fear.