A Robust Computational Workflow Utilizing Generalized Least Squares (GLS) and Generalized Estimating Equations (GEE-GLM) for Longitudinal Flow Cytometry Maturation Data Analysis.

GLS and GEE-GLM modeling approaches reveal significant differences in hematopoietic maturation profiles across disease states, with GEE-GLM highlighting early-stage variations.

• Why it matters: Accurate analysis of longitudinal flow cytometry data is crucial for understanding dynamic biomarker expression, but traditional tests often fail to account for data complexity, risking misleading conclusions.
• What they did: The study compared GLS and GEE-GLM methods on flow cytometry data from 25 MDS, 25 AML patients, and 50 controls, modeling cell fractions and ratios across 20 maturation stages while addressing nonlinearity, correlation, and cell count variability.
• The result: Findings show that total cell count influences population estimates, especially early stages, with GLS providing flexible modeling of ratios and GEE-GLM emphasizing early-stage differences, supporting robust, population-level insights.