On the statistical analysis of single cell lineage trees

Area of Science:

• Stem cell biology
• Developmental biology
• Computational biology

Background:

• Stem cells are crucial for tissue regeneration and repair in multicellular organisms.
• Current methods for isolating hematopoietic stem cells (HSCs) using cell-surface markers yield only 50% purity.
• Advanced microscopy enables tracking of single cells and their progeny to generate lineage trees.

Purpose of the Study:

• To develop statistical techniques for identifying true stem cells from lineage tree data.
• To assess the efficacy of these statistical methods in classifying initial cells as HSCs.

Main Methods:

• Application of novel statistical analysis to murine hematopoietic lineage trees.
• Tracking of individual cell fates and lifespans in vitro to construct lineage trees.
• Evaluation of statistical signatures within lineage trees to infer the identity of the progenitor cell.

Main Results:

• Identified a unique signature in 18% of HSC lineage trees, indicative of true stem cell origin.
• Calculated statistical power (0.35) and type-1 error rate (0.047) assuming 50% HSC purity.
• Demonstrated that statistical analysis of lineage trees can enhance cell classification beyond marker-based methods.

Conclusions:

• Statistical analysis of lineage trees offers a powerful tool to improve the identification of stem cells.
• The proposed methods are broadly applicable to lineage trees from diverse organisms, including bacteria and single-celled eukaryotes.
• This approach is expected to advance the understanding of molecular mechanisms driving cellular dynamics at the single-cell level.