Scientists continue to identify new species of plants and animals, life forms hidden in the deep sea or in intricate jungles, among other environments. But, in addition, there is a branch of life that lives in every corner of the planet and we cannot see and that artificial intelligence can help reveal. In fact, a machine learning tool has now made it possible to identify 161,979 new species of RNA viruses.
This finding has been considered by researchers to be a significant advance in mapping life on Earth and could help discover millions of additional viruses that have not yet been characterized. This study, which has been carried out by an international team of scientists and has been published in the journal Cell, constitutes the largest work of discovery of viral species in history.
“We have had access to a part of life on Earth that until now was hidden, revealing incredible biodiversity,” said Professor Edwards Holmes, member of the School of Medical Sciences of the Faculty of Medicine and Health of the University of Sydney and one of the main authors of the work.
“This is the largest number of new virus species discovered in a single study, greatly expanding our knowledge about the viruses around us,” added Holmes. “Finding so many new viruses at once is astonishing, and it is only the beginning of a world to explore. “There are still millions to be discovered, and this approach can also be applied to identify bacteria and parasites.”
RNA viruses capable of surviving in the harshest environments
Although RNA viruses are often associated with disease in humans, they are also found in extreme environments around the world, and could play critical roles in global ecosystems. In this study, hot springs and hydrothermal vents were found in the atmosphere.
“The fact that so many types of viruses exist in extreme environments is another example of their impressive diversity and ability to survive in the harshest environments, which may offer us clues about how viruses and other elemental life forms emerged,” Holmes explained. .
The researchers developed a deep learning algorithm, called LucaProt, to analyze huge amounts of genetic sequence data, including viral genomes of up to 47,250 nucleotides and complex genomic information, discovering more than 160,000 viruses.
“Most of these viruses had already been sequenced and were in public databases, but they were so different that no one knew what they were,” Holmes explained. “They were part of what is commonly called the ‘dark matter’ of the sequences. “Our AI-based methodology managed to organize and categorize all this scattered information, making sense of that ‘dark matter’ for the first time.”
“Finding so many new viruses at once is astonishing, but there are still millions to be discovered and this approach can also be applied to identify bacteria and parasites”
The AI was trained to analyze this dark matter and identify viruses based on the sequences and secondary structures of the proteins that all RNA viruses use to replicate and this allowed us to significantly accelerate the discovery of viruses, a process that with traditional methods would be much slower.
“We previously relied on tedious bioinformatics processes to discover viruses, which limited the diversity we could explore. Now, with this AI-based model, we have a much more effective tool that offers exceptional sensitivity and specificity, allowing us to delve much deeper into viral diversity. We plan to apply this model in various areas,” explained Professor Mang Shi from Sun Yat-sen University and co-author of the study.
For his part, Dr. Zhao-Rong Li, who works at Alibaba Cloud Intelligence’s Apsara Laboratory, commented: “LucaProt represents a significant integration of cutting-edge AI technology with virology, demonstrating that AI can effectively accomplish tasks in biological exploration. “This integration provides valuable insights and encourages research to decode biological sequences and decompose biological systems from a new perspective.”
“The next logical step is to train our method to uncover even more of this incredible diversity, and we don’t know what additional surprises we might find,” concludes Professor Holmes.
