AGDEX

AGDEX quantifies agreement in differential gene expression between two experiments to evaluate cross-species transcriptomic similarity and model fidelity.

Key Features:

• Metric of Agreement: Introduces a metric that quantifies similarity between differential expression results from two experiments by comparing lists of differentially expressed genes.
• Permutation-based Significance Testing: Assesses statistical significance by performing permutation tests that randomly reassign or shuffle group labels within each dataset.
• Gene-set and Meta-analytic Analyses: Performs gene-set analyses to evaluate pathway-level similarities and integrates results across studies using meta-analytic approaches.
• Cross-species Model Comparison: Enables quantitative comparison between animal models and human disease transcriptomes to assess which model aspects reflect human pathophysiology.

Scientific Applications:

• Model Validation: Validates animal models by quantifying concordance of differential expression profiles with corresponding human disease datasets.
• Disease Mechanism Elucidation: Identifies shared gene expression patterns and pathways across datasets to inform molecular underpinnings of disease.
• Cell of Origin Studies: Supports inference of cell of origin for human cancers by comparing animal model signatures to human cancer subtype profiles, as applied in brain tumor research.

Methodology:

Takes differential expression data from two experiments, computes an agreement metric by comparing lists of differentially expressed genes, assesses significance via permutation tests that randomize group labels within each dataset, and conducts gene-set analyses and meta-analytic integration.