Differential rhythmicity: detecting altered rhythmicity in biological data

Bioinformatics. 2016 Sep 15;32(18):2800-8. doi: 10.1093/bioinformatics/btw309. Epub 2016 May 20.

Abstract

Motivation: Biological rhythms, such as rhythms in gene expression controlled by the cell cycle or the circadian clock, are important in cell physiology. A common type of experiment compares rhythmicity in tissues or cells either kept under different conditions or having different genotypes. Such investigations provide insights into underlying mechanisms as well as functions of rhythms.

Results: We present and benchmark a set of statistical and computational methods for this type of analysis, here termed differential rhythmicity analysis. The methods detect alterations in rhythm amplitude, phase and signal to noise ratio in one set of measurements compared to another. Using these methods, we compared circadian rhythms in liver mRNA expression in mice held under two different lighting conditions: constant darkness and light-dark cycles, respectively. This analysis revealed widespread and reproducible amplitude increases in mice kept in light-dark cycles. Further analysis of the subset of differentially rhythmic transcripts implied the immune system in mediating ambient light-dark cycles to rhythmic transcriptional activities. The methods are suitable for genome- or proteome-wide studies, and provide rigorous P values against well-defined null hypotheses.

Availability and implementation: The methods were implemented as the accompanying R software package DODR, available on CRAN.

Contact: pal-olof.westermark@charite.de

Supplementary information: Supplementary data are available at Bioinformatics online.

MeSH terms

  • Animals
  • Biological Clocks*
  • Circadian Clocks
  • Circadian Rhythm*
  • Darkness
  • Gene Expression Regulation
  • Light*
  • Liver
  • Mice
  • Models, Statistical
  • Photoperiod
  • RNA / metabolism

Substances

  • RNA