The hottest science uses artificial intelligence t

Science: using artificial intelligence to predict the animal hosts and vectors of RNA viruses

November 3, 2018/bion/- -- many fatal and emerging viruses such as Ebola virus and Zika virus have been tested in Qujing branch of Yunnan Chihong Zinc Germanium Co., Ltd. for industrial utilization. They spread in wild animals and insect communities long before they spread to humans and cause serious diseases. Finding animal and insect hosts of different viruses from genome sequences may require years of intensive field research and laboratory work. The resulting delay means it is difficult to implement preventive measures, such as vaccinating animal sources of disease or preventing dangerous contact between species. Therefore, at present, finding these natural virus hosts in time - which may help prevent transmission to humans - poses a great challenge to scientists

now, in a new study, researchers from research institutions such as the University of Glasgow in Scotland have designed a new machine learning algorithm, which uses the viral genome sequence to predict the possible natural hosts of a series of RNA viruses, among which RNA viruses are the most common viral populations that jump from viruses to humans. The relevant research results were published in the journal Science on November 2, 2018, and the title of the paper is "predicting reservoir hosts and arthropod vectors from evolutionary signatures in RNA has its own characteristics and applicable environmental special virus genes"

digital color transmission micrograph of Zika virus. Zika virus belongs to Flaviviridae. The blue virus particles in the figure are 40nm in diameter, with an outer membrane and an inner dense core. The picture is from cdc/cynthia goldsmith

these researchers studied the genomes of more than 500 single stranded RNA viruses to train this machine learning algorithm in order to match the patterns embedded in the virus genome with the animal origin of the virus. This machine learning algorithm can accurately predict which animal host each virus comes from and whether each virus needs a blood sucking animal vector, which simplifies the operation and shortens the cycle time. If necessary, this animal vector is ticks, mosquitoes, midges or sandflies

next, these researchers applied this machine learning algorithm to Zika virus and mers CoV, as well as viruses whose hosts and animal vectors are unknown, such as Crimean Congo hemorrhagic fever virus. It has been proved that for these viruses, the host predicted by this machine learning algorithm is usually the best guess at present

surprisingly, among the four Ebola viruses believed to have bat origin, this machine learning algorithm is the same or the latest Moldex3D product release nbsp; The structural analysis of reinforced plastic composites is more inclined to the fact that the two Ebola viruses are actually of primate origin, which may indicate that non-human primates, rather than bats, are the source of some Ebola virus outbreaks

these results suggest that this new tool may help people provide preventive measures against fatal diseases. These researchers now hope that this new machine learning tool will accelerate research, monitoring and disease control activities in order to find out the correct animal origin of the virus in the wild, with the ultimate goal of preventing the deadly and dangerous virus from spreading to humans. (biological Valley)

reference:

Simon a. babayan1,2, Richard J. orton3, Daniel g. streicker science, 02 nov 2018, 362(6414):, doi:10.1126/p9072.

mark woolhouse.. science, 02 nov 2018, 362(6414):, doi:10.1126/v4265.