GLMMcosinor: Flexible Cosinor Modeling to Characterize Rhythmic Time Series Using a Generalized Linear Mixed Modeling Framework

BackgroundModeling rhythmic biological processes, such as gene expression and sleep-wake cycles, is critical for understanding physiological mechanisms and their dysregulation in disease. Traditional cosinor analysis, commonly used to model rhythmic data, assumes Gaussian-distributed residuals and does not account for hierarchical data, limiting its applicability in modern biological datasets. ResultsWe present GLMMcosinor, an R package that integrates cosinor modeling into the Generalized Linear Mixed Modeling (GLMM) framework using glmmTMB. GLMMcosinor enables analysis of a broad spectrum of non-Gaussian and hierarchical data structures, including count, positive-only, and zero-inflated distributions. By incorporating mixed-effects modeling, GLMMcosinor improves parameter estimation and biological interpretability. The package includes functions for group comparisons of rhythmic parameters and visualization tools such as polar and time series plots. Additionally, GLMMcosinor is available as a Shiny app for intuitive, code-free analysis. ConclusionsGLMMcosinor significantly extends the flexibility and scope of rhythmic data analysis by incorporating GLMM functionality. It is freely available on GitHub, CRAN, rOpenSci, and the R-universe, with comprehensive documentation and reproducible examples, making it a robust tool for researchers analyzing complex rhythmic datasets.