Detecting Rhythmic Gene Expression in Single-cell Transcriptomics.

Area of Science:

• Chronobiology
• Genomics
• Computational Biology

Background:

• Circadian rhythms, driven by cellular transcription-translation feedback loops, regulate daily gene expression in multicellular organisms.
• Understanding circadian gene control is vital for elucidating physiological coordination.
• Single-cell RNA sequencing (scRNA-seq) offers a powerful tool for studying cell-level circadian clocks.

Purpose of the Study:

• To benchmark existing circadian detection algorithms for their reliability and efficiency on single-cell RNA sequencing data.
• To provide guidance on adapting bulk transcriptomic methods for circadian gene detection in the single-cell domain.
• To identify opportunities for improving rhythm detection in single-cell datasets.

Main Methods:

• Benchmarking of commonly used circadian detection algorithms on scRNA-seq data.
• Evaluation of algorithm performance in terms of reliability and efficiency.
• Proposal and testing of a subsampling procedure combined with harmonic regression for circadian gene detection.

Main Results:

• Established the performance characteristics of various circadian detection methods when applied to scRNA-seq data.
• Demonstrated that existing methods require adaptation for effective use with single-cell resolution.
• Identified a subsampling and harmonic regression strategy as a robust and efficient approach for single-cell circadian rhythm detection.

Conclusions:

• Guidance is provided for adapting established circadian detection algorithms to the single-cell RNA sequencing context.
• The proposed subsampling and harmonic regression method enhances the efficiency and robustness of circadian gene discovery in single-cell data.
• This work facilitates deeper understanding of cell-specific circadian regulation and its physiological implications.