Items | Bulk transcriptome | Single-cell transcriptome | Spatial transcriptome |
---|---|---|---|
Analytic object | Tissue | Cell | Tissue section |
Tumor cell region | Not applicable | Presumed by the algorithm | Identified directly on sections |
Dimensionality reduction | Not applicable | Principle component analysis (PCA), t-distributed stochastic neighbor embedding (t-SNE), or uniform manifold approximation and projection (UMAP) | Principle component analysis (PCA), t-distributed stochastic neighbor embedding (t-SNE), or uniform manifold approximation and projection (UMAP) |
Cluster | Not applicable | k-means, louvain or hierarchical clustering based on cell type | k-means, louvain or hierarchical clustering based on different functional areas |
Differential expression analysis | For different tissues | For cell clusters | For spatial location |
Enrichment analysis | Gene function differences between different tissues | Biological functional differences between the cell clusters | Gene function differences at different spatial locations |
Advantage | The price is low | Cell resolution | The spatial location information is retained |
Limitation | The result is the average gene expression within the tissue, and the precision is low | Missing the spatial location information | Technically, the resolution is lower than single-cell transcriptome in most cases |