Skip to main content
. 2021 Dec 21;50(2):704–716. doi: 10.1093/nar/gkab1236

Figure 1.

Figure 1.

Overview of CCPE approach. After normalization and pre-processing of the data, CCPE learns the discriminative helix by iteratively optimization between the original and reduced dimension until convergence. After optimization, a 3D-helix with two-gene dimensions is used to represent circular information of cell cycle phases and one dimension to represent pseudotime of cells along the cell cycle. Several downstream analyses and applications of CCPE are used to assess its performance.