The invention discloses a CYP51B
copy number variation-based
antifungal drug MIC value prediction method and a CYP51B
copy number variation-based
antifungal drug MIC value prediction device, and the method comprises the following steps: obtaining second-generation
sequencing data of a plurality of
trichophyton indicum strains, the second-generation
sequencing data comprising CYP51B
gene copy number of each strain, and measuring the MIC measured value of each strain for at least one target
antifungal drug, and determining the CYP51B
copy number variation-based
antifungal drug MIC value of each strain according to the CYP51B copy number variation-based
antifungal drug MIC value. Constructing a training
data set; the copy number of the CYP51B
gene and the MIC measured value are respectively calculated to obtain a log < 2 > (copy ratio) value and a log < 2 > (MIC value); a log2 (copy ratio) value is taken as an independent variable X, log2 (MIC value) is taken as a dependent variable Y, a unary
linear regression model equation: Y = aX + b is obtained through least square fitting based on the training
data set, a is a regression coefficient, and b is an intercept. The method has the beneficial effects that the log2 (copy ratio) value of the CYP51B
gene is calculated through a
bioinformatics process only by obtaining
genome sequencing data of a strain, and the log2 (copy ratio) value can be substituted into a preset
linear regression equation to obtain an MIC predicted value in real time.