Fig. 5. Batch correction analysis of scRNA-seq data using SIMBA.
a, SIMBA graph construction and embedding in batch correction analysis. Overview of SIMBA’s approach to batch correction across scRNA-seq datasets. Distinct shapes indicate the type of entity (cell or gene). Colors distinguish batches or cell types. b, UMAP visualization of the scRNA-seq human pancreas dataset, with five batches of different studies before and after batch correction. Cells are colored by scRNA-seq data source and cell type, respectively. Top, UMAP visualization before batch correction. Bottom, UMAP visualization after batch correction with SIMBA. c, UMAP visualization of SIMBA embeddings of cells and genes, with batch effect removed and known marker genes highlighted.