Screening single-cell trajectories via continuity assessments for cell transition potential.

Area of Science:

• Computational Biology
• Genomics
• Systems Biology

Background:

• Single-cell sequencing enables inference of biological trajectories from transcriptome data.
• Trajectory analysis is crucial for understanding dynamic cellular processes like development and differentiation.
• Current methods lack metrics to assess trajectory continuity, impacting inference accuracy.

Purpose of the Study:

• To investigate the impact of trajectory discontinuity on single-cell trajectory inference accuracy.
• To develop a scoring algorithm for assessing trajectory continuity.
• To validate the continuity assessment approach in real-world biological systems.

Main Methods:

• Simulated breaks were introduced into in silico and real single-cell data to analyze discontinuity effects.
• A scoring algorithm was developed to quantify trajectory continuity.
• The algorithm was applied to intestinal stem cell development and CD8+ T cell differentiation data.
• Case studies in psoriatic arthritis and acute kidney injury were used for validation.

Main Results:

• Trajectory discontinuity significantly reduces inference accuracy.
• The developed scoring algorithm effectively assesses trajectory continuity.
• Continuity assessments successfully identified trajectories consistent with empirical knowledge in real-world data.
• The approach aids in screening inferred lineages and prioritizing trajectories for validation.

Conclusions:

• Trajectory continuity is a critical assumption in single-cell trajectory inference.
• The developed scoring algorithm provides a robust method for assessing continuity.
• This tool enhances the reliability of trajectory inference and aids in biological discovery.
• The scEGRET tool is available on GitHub for broader application.