We are the first team to  to classify  upper respiratory viruses (UV)  family by deep learning approach (CNN). The model are implemented into a web platform as a handy tool for  sub-typing UV related sequences in high confidence and deep strain resolution.

We are the first team to develop the CNN model for classifying 11 categories of upper respiratory viruses (UV). The web application (UV hunter) we implemented is fully automatic to provide precise and rapid prediagnosis on genotype based on those long reads in third-generation sequencers with low cost. Such an approach will benefit clinical laboratories, research community for disease surveillance.

We are going to collaborate with medical institutes, hospitals, bio-industries, and research community to validate our work clinically. Meanwhile, we seek any possibility to apply this integrated solution to provide precise and rapid prognosis on the genotypes of respiratory viruses even down to specific strains, especially for those notorious strains with high virulence like EV71, D68.