GraphDDP: a graph-embedding approach to detect differentiation pathways in single-cell-data using prior class...

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) profiles cells to identify subpopulations via clustering.
• Clustering discretizes cell states, failing to capture continuous biological processes like cell differentiation.
• Existing methods often focus solely on either subpopulations or continuous trajectories, limiting comprehensive analysis.

Purpose of the Study:

• To develop a method that integrates both cell subpopulations and differentiation trajectories.
• To provide an intuitive visualization for exploring cell state dynamics.
• To identify complex differentiation pathways not easily detectable by other approaches.

Main Methods:

• GraphDDP utilizes a force-based graph layout approach.
• It constructs edges based on both cluster membership and density gradients to represent differentiation.
• The method starts with user-defined cluster assignments as anchor points.

Main Results:

• GraphDDP successfully integrates discrete clusters with continuous differentiation trajectories.
• The visualization reveals complex differentiation pathways in intestinal epithelial cells and myeloid progenitors.
• Identified pathways were not easily detectable using conventional methods.

Conclusions:

• GraphDDP offers a powerful, integrated approach to analyze single-cell data.
• It enhances the understanding of cell differentiation processes by combining clustering and trajectory inference.
• This method provides novel insights into biological systems through intuitive visualization.