A semi-local neighborhood-based framework for probabilistic cell lineage tracing.

Area of Science:

• Computational Biology
• Developmental Biology
• Imaging Science

Background:

• In vivo imaging advances enable complex biological process recording.
• Accurate cell tracking is computationally challenging and hinders analysis.
• In vivo analysis requires tracking numerous cells over time in 3D environments.

Purpose of the Study:

• To develop a computational framework for efficient and accurate cell lineage tracing.
• To address the challenges of computational expense and difficulty in cell tracking.
• To enable large-scale analysis of in vivo biological processes.

Main Methods:

• Focusing computational effort on difficult tracking cases identified by local cell number increases.
• Utilizing tentative cell track bifurcations to define semi-local neighborhoods.
• Employing Bayesian judgment within these neighborhoods to interpret cell behavior and correct errors.

Main Results:

• The method efficiently and accurately traces entire cell lineages.
• It correctly tracks cells through divisions and large movements, correcting detection errors.
• Demonstrated large-scale analysis of Caenorhabditis elegans development with an invariant cell lineage.

Conclusions:

• Bifurcation-based semi-local neighborhoods provide sufficient information for accurate tracking.
• Difficult tracking cases can be interpreted without global optimization.
• The method facilitates access to lineage data, enabling new statistical analyses of in vivo processes.