Influenza virus antigen dissimilarity calculation method and system

Abstract

The invention relates to an influenza virus antigen dissimilarity calculation method and system. The method comprises the following steps: acquiring a hemagglutination inhibition titer set obtained by a known hemagglutination inhibition test, wherein the hemagglutination inhibition titer set comprises known hemagglutination inhibition titers; calculating the antigen distance between strains according to the hemagglutination inhibition titer to obtain an antigen distance set; establishing an antigen dissimilar network according to the antigen distance set; mapping nodes in the antigen dissimilar network into low-dimensional dense vectors based on a network embedding technology; and quantitatively calculating the antigen diversity between every two strains according to the low-dimensional dense vector. According to the method and the system, the antigen dissimilarity is modeled based on the network structure, the nonlinear feature learning capability of deep learning on the network structure is exerted, the antigen dissimilarity between every two strains can be quantitatively calculated, the robustness of antigen dissimilarity calculation is improved, and further, antigenicity analysis except for influenza antigen specificity can be supported, and good expandability and practicability are achieved.

Description

technical field [0001] The invention relates to the technical field of bioinformatics, in particular to a method and system for calculating the dissimilarity of influenza virus antigens. Background technique [0002] Influenza virus is one of the main pathogens that endanger human health, which may lead to seasonal influenza epidemics and even global influenza outbreaks. Currently, the most effective way to prevent influenza virus infection is to vaccinate with a vaccine that is antigenically similar to the circulating strains. However, the main antigen of influenza, the surface glycoprotein hemagglutinin (HA) has a high mutation rate, and the accumulated amino acid mutations lead to antigenic drift of the virus, allowing the virus to escape immunity. Therefore, timely identification of new antigenic variants is crucial for vaccine screening and updating. However, the hemagglutinin-inhibition (HI) test to characterize antigenic differences is not only costly but also time-...

AI Technical Summary

Problems solved by technology

However, the current antigenic dissimilarity calculation method based on machine learning can only achieve qualitative antigenic dissimilarity prediction, and usually a binary classification prediction, and cannot quantitatively calculate the antigenic dissimilarity between two strains

Images