Spatial transcriptome data clustering method

The invention discloses a spatial transcriptome data clustering method, which comprises the following steps of: preprocessing original data, and constructing a spot spatial neighbor network at the same time; secondly, further learning low-dimensional potential representation of space information and gene expression through a graph attention automatic encoder, learning self-expression coefficient matrixes of different layers in the encoder through a multi-scale self-expression module and fusing the self-expression coefficient matrixes together, and in a depth subspace clustering module, adopting spectral clustering and feeding a clustering label back to a self-supervision module; the self-supervision module is used for guiding the learning of the potential representation by constructing a self-supervision path of a return encoder; and finally, performing biological analysis. According to the method, low-dimensional potential embedding is learned by integrating spatial information and a gene expression profile,