CLME

CLME implements linear fixed-effects and mixed-effects models with linear inequality (order) constraints on regression coefficients for hypothesis testing and robust inference in hierarchical and longitudinal biological data.

Key Features:

• R implementation: Provided as an R package for fitting constrained linear fixed-effects and mixed-effects models.
• Inequality constraints: Enables imposition of linear order constraints on regression coefficients to reflect ordered hypotheses.
• Support for fixed and mixed effects: Applies constraints in both linear fixed-effects models and linear mixed-effects models.
• Residual bootstrap inference: Uses a robust residual bootstrap methodology for statistical inference on constrained models.
• Complex residual variance structures: Accommodates specified structures within the residual variance matrix.

Scientific Applications:

• Hierarchical and grouped data analysis: Tests ordered hypotheses in datasets with grouped or multi-level structure.
• Longitudinal studies and multi-level experiments: Applies constrained inference to repeated-measures and multi-level experimental designs.
• Genetic association studies: Evaluates ordered relationships among predictors in genetic association analyses.
• Ecological and ordered-relationship modeling: Assesses hypothesized parameter orderings in ecological and other biological models.

Methodology:

Inference is performed using residual bootstrap techniques on models with linear order constraints of regression coefficients and can accommodate structures in the residual variance matrix to maintain robustness under non-normality and heteroscedasticity.