Fig. 2

Workflow and features of the scCorr method. Data dimensional reduction and cell classification by tSNE and cell type identification using the marker gene approach in dataset 1 (A). Cell partitioning is based on a tSNE plot using scCorr with different numbers of clusters in dataset 1 (B: k = 100; C: k = 1000). scCorr partitioned different numbers of clusters for wild-type cells in dataset 2 (D: k = 50; E: k = 100; F: k = 500) and for knockout cells in dataset 2 (G: k = 50; H: k = 100; I: k = 500). J: tracing of the evolution of each partitioned cluster