Figure 3.
(a) The comparison of Recall@k metrics between different models for k = 10, 20, 30, 40, and 50 for the 10x PBMC Multiome dataset (gene expression and chromatin accessibility). (b) Normalized Median Rank values for different models. (c) Cell type accuracy comparison between models. (d) Comparison of ASW score across models. (e) t-SNE embeddings of the original dataset (scRNA-seq and scATAC-seq). (f) t-SNE visualization of the sCIN’s integrated embeddings.
