Analysis | Load scRNA-seq data of 5 cancer cell lines from Tian et al., 2019. | Simplify single-cell data at the graining level $gamma = 20$ | Compute gene expression for simplified data | Map each metacell to a particular cell line | Cluster metacell data | Map clusters of metacells to cell lines | Differential expression analysis of clustered metacell data | Some additional plotting options | SuperCell graining level can be quickly chaged with supercell_rescale() function | P.S.: SuperCell to Seurat object

A combined analysis -- construction of metacells processing two samples together | Independent analysis -- construction of metacells for each sample independently (applying the same granularity gamma and the same set of features genes.use) | Combined and independent analyses do not result in the same metacell construction.

RNA velocity combining with metacells computed with SuperCell | Load data and compute metacells' spliced and un-spliced counts | Compute RNA velocity | Plot RNA velocity on a tSNE's coordinates | Set up the color scheme | Plot tSNE