Fig. 1.

Conceptual illustration of the DVC analysis based on gene expression variability and correlation. Calculation concept of DVC analysis. a First step of calculating two indexes. In each state of cells, two types of criteria, variability and correlation, is calculated. For index 1 (variability), the standard deviation (SD) of each gene expression between cells (three single cells in this example) are calculated. For index 2 (correlation), the Pearson correlation coefficient of pair of expression patterns from group of cells (three single cells in this example) is calculated between each pair of genes. By clustering, the correlated genes are grouped as co-expressed modules. b Second step of calculating state transition score (STS). The change rate between two state of cells is calculated using both indexes to obtain STS. DVC gene is the gene with high STS. An example image of DVC gene is also illustrated, which shows oscillation-like variability in the early state of cells, although settle/harmonize after the transition to the next state