A New Workflow to Enhance Longitudinal Data Analysis
© 2025 EPFL
Time-series measurements are often used in biomedical research; however, existing statistical methods for analyzing the resulting longitudinal data have limitations. To address these issues, Cathrin Brisken's group at EPFL, in collaboration with the BICC, has developed a new workflow: biogrowleR.
biogrowleR is a workflow for data visualization and analysis based on frequentist and Bayesian inference, combined with hierarchical modeling. Differences between experimental conditions are interpreted using effect sizes that account for the dynamics of the longitudinal data. The workflow also includes a key randomization algorithm designed to reduce the number of laboratory animals (RRR) and associated costs. The BICC staff helped to develop the frequentist framework to derive effect sizes and confidence intervals for comparing growth rates across different conditions. To encourage broader adoption, the workflow and its associated R package were specifically designed for researchers with limited experience in R and biostatistics.
biogrowleR: Enhancing Longitudinal Data Analysis. Ronchi C, Ambrosini G, Hughes F, Flaherty RL, Quinn HM, Matvienko D, Agnoletto A, Brisken C.
J Mammary Gland Biol Neoplasia. 2025 Jun 3;30(1):9. doi: 10.1007/s10911-025-09583-7.