Construction of a screening and evaluation system for anti-hepatitis B virus drugs based on quadratic exponential smoothing model

Abstract

Abstract Exploring the construction of an anti-hepatitis B virus drug screening and evaluation system is to better develop anti-HBV virus drugs. In this paper, we analyzed the types of hepatitis B virus present in different hepatocytes, starting from the hepatitis B cell line model. Based on the quadratic, exponential smoothing model, a QES-LSTM viral gene detection model was constructed by introducing a long and short term memory neural network, and experimental analysis of the sensitivity and specificity of viral gene detection was conducted for this model. From the sensitivity experiments, the sensitivity of HBV DNA, DHBV DNA and DHBV cccDNA were 60 copies/ml, 60 copies/ml and 10 copies/ml, respectively. From the specificity experiments, the mean values of specificity of HBV DNA, DHBV DNA, and DHBV cccDNA were 0.489, 0.481, 0.429, respectively, 0.429, which showed positive amplification compared to other types. This indicates that effective discrimination of HBV viral genes is needed in the construction of an anti-HBV virus drug screening and evaluation system, which in turn allows targeted screening of drugs for the treatment of the HBV virus.