lmerSeq

lmerSeq fits linear mixed effects models to transformed RNA-Seq count data to account for dependence between observations such as longitudinal sampling and enable accurate statistical inference.

Key Features:

• Linear mixed effects modeling: lmerSeq fits linear mixed effects models to transformed RNA-Seq counts, modeling fixed and random effects to account for dependence between observations.
• Support for transformations and variance structures: lmerSeq accommodates a range of count transformations to stabilize variance and supports flexible variance structures for complex experimental designs.
• Improved statistical performance: In simulation comparisons with two other methods operating on transformed counts, lmerSeq showed better control of nominal error rates and higher statistical power.

Scientific Applications:

• Longitudinal and repeated-measures studies: lmerSeq models within-subject dependence in longitudinal sampling and repeated measures designs.
• Differential expression across conditions and time points: lmerSeq enables differential expression analysis of RNA-Seq data across experimental conditions and temporal measurements.
• Gene regulation, disease progression, and treatment response: lmerSeq facilitates inference about gene expression changes relevant to gene regulation, disease progression, and response to treatments over time.

Methodology:

Raw RNA-Seq counts are transformed to stabilize variance and normalize distribution before fitting linear mixed effects models, and the approach supports various transformations and flexible variance structures.