According toSCMP, based on information available in databases, an artificial intelligence tool (AI) of China helps scientists detect nearly 162,000 RNA virus species that have never been known for unprecedented speed -One species per second.
The general research was conducted by scientists in mainland China, Hong Kong and Australia.
Scientists discovered nearly 162,000 new RNA virus species thanks to the Chinese AI tool.(Photo: SCMP)
The team said nearly 162,000 new RNA virus species were discovered in different environments - including the atmosphere, hot springs and heat catheterization - highlighting their diversity and recovery in Articleharsh conditions, and can provide clues about how viruses and other basic living forms appear.
By analyzing the genetic sequence data that has not been identified in public databases, AI tools identify the virus based on the hidden protein structure and information that the RNA virus uses to copy.
This tool uses algorithms developed by the Alibaba Cloud Intelligence under the advice of virusists.
The team wrote in an article published in Cell magazine on October 10:"We have developed a deep -based study model, surpassing the usual methods of accuracy, efficiency and most importantly, the diversity of viruses is detected."
The group said that this is the largest new virus detection research that has been published in terms of the number of species in a report.
Team leader Shi Mang, virus expert and professor at Trung Son University Medical Academy (Guangdong, China) said that this tool not only speeds up the detection of viruses that are very time -consuming.When using traditional methods, it also allows scientists to discover the field of viruses that have never been known before.
“All viruses discovered in this study exist in the environment and have been sequenced.Our previous methods cannot identify them, making them 'dark matter' for scientists '', Mr. Shi said, referring to the previous independent or detected sequences related to the known viruses.
The professor added:“The AI tool to solve this problem for us with high accuracy is equivalent to the usual computer biological methods.It can explore the 'dark matter' chains, along with the sequences closer to the established virus groups. ”
Mr. Shi said this discovery promotes future studies by placing the foundation for the diversity of the virus."For example, this study tells us about the existence of the virus in harsh conditions, such as hot spring. This knowledge allows scientists to develop more comprehensive descriptions of ecologyIn different ecosystems ”.
The virus added:"In terms of viral ability, we will be able to study more about how the virus interacts with the host, as well as identifying viral groups that can infect a specific host."
Co -author Li Zhaorong, a biological research expert at Alibaba Cloud Intelligence's Apsara Laboratory, said the study proves that deep algorithms can effectively perform the tasks in biological discovery.
Mr. Li claims the research team will continue to update the tool with advanced AI technologies such as new training models to analyze nucleotide and protein.
"We are also reviewing the classic issues in viruses learning with AI thinking and new data, such as the structure and function of the virus, or the relationship between viruses and people and animals.", Mr. Li said.
The team expects the same methods that can be applied to identify bacteria and parasites.