Cross-species cell-type assignment from single-cell RNA-seq data by a heterogeneous graph neural network

(Downloading may take up to 30 seconds. If the slide opens in your browser, select File -> Save As to save it.)

Click on image to view larger version.

Figure 3.
Figure 3.

Alignment comparison of cell embeddings across data sets by CAME and four other methods. (A) The UMAP plots of the cell embeddings by CAME and four integration methods on the scRNA-seq data from turtle (reference) and mouse (query) brains. Cells are colored by their cell types (the first column) or data set identities (the second column). (B) Similar settings to A. Here the reference data sets are the human pancreatic scRNA-seq data from eight batches by three different platforms, and the query is from mouse pancreas cells. The UMAP plots of the third column show the reference cells, colored by batch identities. The scores under both the figures in A and B represent the graph connectivity score for each method.

This Article

  1. Genome Res. 33: 96-111

Preprint Server