Methods:Extracting the main feature components of big data, also known as principal component analysis.

Each detected gene has an expression value (FPKM/RPKM/TPM), and the expression of all the genes is transformed into a set of vectors in a two-dimensional space, assuming that we have detected 10,000 genes this time, the spatial distribution of all the data may theoretically involve 10,000 dimensions, and according to our idea of dimensionality reduction, n points must be able to be analyzed in a k(k<n)-dimensional space, and we can finally compress the high-dimensional data into the two-dimensional plane where the first and second feature components are located by linear transformation. The idea of dimensionality reduction in PCA is involved.