Likelihood-based tests for detecting circadian rhythmicity and differential circadian patterns in transcriptomic...

Area of Science:

• Genomics
• Chronobiology
• Bioinformatics

Background:

• Circadian rhythms regulate gene expression in many biological processes.
• Disruptions in circadian patterns are linked to various diseases.
• Accurate detection of circadian rhythmicity and differential patterns is crucial for understanding health and disease.

Purpose of the Study:

• To develop and validate novel likelihood-based statistical methods for detecting circadian rhythmicity (LR_rhythmicity) and differential circadian patterns (LR_diff).
• To assess the performance of these methods in controlling Type I error rates and increasing statistical power.
• To demonstrate the utility of the developed methods on diverse biological datasets.

Main Methods:

• Development of likelihood-based methods: LR_rhythmicity for detecting circadian rhythmicity and LR_diff for identifying differential circadian patterns.
• Simulation studies to evaluate Type I error control and statistical power compared to existing methods.
• Application of methods to real-world biological data, including gene expression microarray, RNA sequencing, and single-cell RNA sequencing datasets.

Main Results:

• LR_rhythmicity demonstrated superior control of Type I error rates across various simulation settings.
• LR_diff successfully controlled Type I error rates for detecting differential amplitude, phase, basal level, and fit.
• LR_diff exhibited higher statistical power in detecting differential fit compared to existing approaches.
• The methods showed robust performance in analyzing human brain aging, time-restricted feeding, and mouse suprachiasmatic nucleus datasets.

Conclusions:

• The developed LR_rhythmicity and LR_diff methods provide robust and powerful tools for analyzing circadian rhythmicity and differential gene expression patterns.
• These methods offer significant improvements over existing approaches in terms of statistical accuracy and power.
• The publicly available R package facilitates the application of these methods in diverse research areas, advancing the study of circadian biology and its role in disease.