LimoRhyde2: Genomic analysis of biological rhythms based on effect sizes

Area of Science:

• Genomics
• Chronobiology
• Bioinformatics

Background:

• Genome-scale data reveal daily rhythms in many species and tissues.
• Current methods for rhythmicity assessment often focus on statistical significance, potentially missing biological relevance.
• There is a need for methods that evaluate the magnitude and uncertainty of rhythm-related effects.

Purpose of the Study:

• To develop and demonstrate LimoRhyde2, a novel method for assessing biological rhythms based on effect sizes.
• To improve the interpretation of genomic data related to daily rhythms.
• To prioritize biologically relevant rhythmic features over merely statistically significant ones.

Main Methods:

• LimoRhyde2 fits curves to genomic features using periodic splines.
• An Empirical Bayes approach (multivariate adaptive shrinkage, Mash) moderates the fits.
• Rhythm statistics, such as peak-to-trough amplitude, are calculated from moderated fits.

Main Results:

• LimoRhyde2 was applied to circadian transcriptome datasets.
• The method successfully prioritized genes with high-amplitude expression rhythms.
• Compared to a prior method (BooteJTK), LimoRhyde2 identified different sets of rhythmic genes, focusing on amplitude.

Conclusions:

• Quantifying rhythm-related effect sizes offers a more biologically relevant interpretation of genomic data.
• LimoRhyde2 provides a powerful tool for analyzing biological rhythms in large datasets.
• This approach has the potential to transform the analysis of circadian and other biological rhythms.