BM-DE

BM-DE infers differential gene expression from RNA-Seq position-level read counts using a Bayesian hierarchical model.

Key Features:

• Position-Level Read Count Modeling: Models position-level RNA-Seq read counts within each gene rather than aggregated gene-level summaries.
• Bayesian Framework: Employs a Bayesian hierarchical model to perform statistical inference and to produce posterior estimates and credible intervals for differential expression.
• Analysis Without Biological Replicates: Infers differential expression in experiments lacking biological replicates by leveraging multiple position-level read counts per gene.
• Validation on Empirical and Simulated Data: Demonstrated on empirical yeast datasets and simulation studies to evaluate detection sensitivity and specificity.

Scientific Applications:

• Gene Expression Studies: Identifying genes with differential expression across conditions from RNA-Seq position-level data.
• Single-Sample Analysis: Enabling differential expression analysis in single-sample or limited-replicate study designs.
• Precision Medicine: Characterizing gene expression changes associated with disease that may inform personalized treatment strategies.

Methodology:

Bayesian hierarchical modeling of position-level RNA-Seq read counts to perform statistical inference and compute posterior estimates and credible intervals, leveraging multiple position-level counts per gene to permit analysis without biological replicates; validated on empirical yeast datasets and simulation studies.