A probabilistic framework for cellular lineage reconstruction using integrated single-cell 5-hydroxymethylcytosine...

Area of Science:

• Developmental Biology
• Epigenetics
• Genomics

Background:

• Accurate lineage reconstruction is crucial for understanding tissue development and maintenance.
• Current methods often rely on genetically encoded reporters, which have limitations.
• Endogenous molecular markers offer a promising alternative for lineage tracing.

Purpose of the Study:

• To develop a novel computational algorithm for high-resolution endogenous lineage reconstruction.
• To utilize 5-hydroxymethylcytosine (5hmC) patterns for inferring single-cell-division resolution lineage trees.
• To enable accurate lineage tracing in early embryonic development and investigate chromosome segregation.

Main Methods:

• scPECLR: A probabilistic algorithm to infer lineage trees using 5hmC patterns.
• scH&G-seq: A technique to sequence both 5hmC and genomic DNA from single cells.
• Application to 8-cell stage pre-implantation mouse embryos for lineage tree prediction.

Main Results:

• scPECLR achieved >95% accuracy in predicting lineage trees in mouse embryos.
• Combined scPECLR with scH&G-seq (including genomic DNA data) improved lineage reconstruction for larger trees.
• Demonstrated scPECLR's utility in mapping chromosome strand segregation and testing the 'immortal strand' hypothesis.

Conclusions:

• scPECLR provides a generalized and accurate method for endogenous lineage reconstruction at single-cell-division resolution.
• The integration of 5hmC sequencing with genomic DNA analysis enhances lineage tracing capabilities.
• This approach offers new insights into developmental processes and fundamental cell biology questions.