Identifying and Interpreting Rhythms in Biological Data

Abstract

Methods for detecting and measuring biological rhythms have greatly expanded over the past decades, in parallel with the development of techniques that can collect tens of thousands of molecular measures. This chapter begins by outlining the challenge of finding and describing rhythms in noisy biological data. Using the measurement of RNA expression as a representative example, we characterize the noise and biases inherent in experimental data. We then describe the simple principles underlying several parametric and nonparametric approaches to identify rhythms in time course data, highlighting the advantages and limitations of each approach. The chapter then considers algorithms for characterizing changes in biological rhythms and moves on to methods for contextualizing and interpreting these rhythms using well-curated gene or metabolite sets. Finally, we conclude with a discussion on the emerging body of techniques developed for characterizing biological rhythms without time course data.