TrajLens: Visual Analysis for Constructing Cell Developmental Trajectories in Cross-Sample Exploration.

Area of Science:

• Computational Biology
• Single-cell RNA Sequencing Analysis
• Systems Biology

Background:

• Inferring cell developmental trajectories is crucial for understanding cellular progression.
• Current methods for single-cell RNA sequencing (scRNA-seq) analysis are limited to within-sample trajectories.
• Manual linking of cells across samples is labor-intensive and complex for cross-sample trajectory construction.

Purpose of the Study:

• To develop an automated method for predicting cross-sample cell developmental trajectories.
• To create a visual analytics system (TrajLens) for exploring and refining these trajectories.
• To integrate spatial dynamics into the analysis of cell evolution across multiple samples.

Main Methods:

• Proposed a Graph Neural Network (GNN)-based model for predicting cross-sample cell developmental trajectories.
• Developed TrajLens, a visual analytics system with integrated multi-sample cell distribution and developmental direction features.
• Utilized contour maps superimposed on cell distribution data for intuitive exploration.

Main Results:

• Successfully predicted cross-sample cell developmental trajectories using the GNN model.
• TrajLens provides an overview of spatial evolutionary patterns and allows intuitive exploration.
• Quantitative evaluations and case studies demonstrated the system's effectiveness.

Conclusions:

• The GNN-based model and TrajLens system automate and enhance the analysis of cross-sample cell developmental trajectories.
• The approach effectively addresses the limitations of manual cross-sample analysis in scRNA-seq.
• The system aids biologists in understanding complex cellular spatial dynamics and evolutionary paths.