Reconstructing cell cycle pseudo time-series via single-cell transcriptome data.

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Single-cell mRNA sequencing provides transcriptional profiles for individual cells, crucial for studying tissue and tumor development.
• Accurately resolving cell cycle stages in unsynchronized single-cell data remains a challenge for existing computational methods.

Purpose of the Study:

• To develop a novel computational method, reCAT, for reconstructing cell cycle progression from unsynchronized single-cell transcriptome data.
• To validate reCAT's accuracy and reliability across multiple datasets.
• To explore the relationship between cell cycle dynamics and methylation variations at single-cell resolution.

Main Methods:

• Developed reCAT, a computational method integrating the traveling salesman problem and hidden Markov models.
• Applied reCAT to analyze unsynchronized single-cell transcriptome data to infer pseudo-temporal cell cycle progression.
• Independently tested reCAT's performance using several independent datasets.

Main Results:

• Identified two major waves of cell cycle gene expression, correlating with G1 and G2 checkpoints.
• Demonstrated reCAT's accuracy and reliability in recovering cell cycle dynamics.
• Revealed methylation variations along the reconstructed cell cycle trajectory.

Conclusions:

• reCAT effectively reconstructs cell cycle progression from single-cell RNA sequencing data.
• The method facilitates the study of cyclic processes, including cell cycle and circadian rhythms, at single-cell resolution.
• reCAT provides a powerful tool for uncovering cell cycle-related gene expression patterns and epigenetic modifications.