EBSeq

EBSeq performs differential expression analysis of genes and isoforms from RNA-seq data using an empirical Bayesian model to identify differentially expressed features across conditions.

Key Features:

• Empirical Bayesian Framework: Uses an empirical Bayesian model to estimate expression and posterior probabilities for differential expression, accommodating estimation uncertainty inherent in isoform-level data.
• Improved Power and Performance: Increases power and reduces false discoveries for isoform differential expression compared with traditional count-based methods by modeling variability across isoforms.
• Robust Gene-Level Analysis: Applies the same modeling approach at the gene level to identify differentially expressed genes.

Scientific Applications:

• Normal development studies: Detects gene- and isoform-level expression changes relevant to developmental processes using RNA-seq data.
• Differentiation processes: Identifies differential expression of isoforms and genes during cell differentiation.
• Disease manifestation research: Characterizes gene- and isoform-level expression differences associated with disease states.

Methodology:

An empirical Bayesian approach that models variability across groups of isoforms to manage estimation uncertainty in isoform expression and improve detection of differential expression relative to count-based methods.