TrajectoryGeometry

TrajectoryGeometry uses spherical geometry to detect directionality, turning points, and divergent movements in time series or pseudo-time series of gene expression data.

Key Features:

• Directionality Detection: Identifies whether changes in gene expression exhibit directionality by detecting clustering of points on the (n-1)-dimensional sphere.
• Turning Point Identification: Detects turning points in the directionality of gene expression trajectories.
• Subset Movement Analysis: Determines whether different subsets of data move in different directions.
• Branching vs. Bifurcation Detection: Distinguishes developmental pathway branching, where a bipotential progenitor develops along a relatively straight trajectory leading to one fate while a second trajectory branches off towards another fate, from simple bifurcation models.
• Trajectory Gene Identification: Provides precise descriptions of the genes and cellular pathways involved in identified trajectories.

Scientific Applications:

• Cell Fate Decision Analysis: Tests quantitative models of multipotential progenitor cell differentiation and determines whether developmental pathways branch or bifurcate.
• Transcriptional Programme Investigation: Characterizes the transcriptional programmes associated with distinct cell fates in differentiation processes.
• Pseudo-time Series Analysis: Analyzes pseudo-time series of gene expression patterns to understand developmental dynamics.

Methodology:

The package applies spherical geometry to the first n dimensions of PCA of gene expression data, where directionality is detected as the clustering of points on the (n-1)-dimensional sphere.