GAT

GAT tests whether sets of genomic intervals overlap significantly by estimating overlap significance under a null model that can account for external variables such as isochore structure and chromosome identity.

Key Features:

• Significance estimation: Provides a framework to estimate the statistical significance of overlaps between multiple sets of genomic intervals using a null model that assumes independent placement.
• Simulation-based approach: Employs simulation techniques to estimate statistical significance in complex genomic landscapes.
• Multiple testing correction: Controls the false discovery rate (FDR) to account for multiple tests.
• External-variable conditioning: Allows interval density to depend on external factors such as isochore structure or chromosome identity.

Scientific Applications:

• Functional genomics: Interprets spatial relationships between genomic features derived from ChIP-Seq and RNA-Seq experiments.
• Regulatory interaction analysis: Assesses potential regulatory interactions or co-localization of genetic elements by testing overlap significance.
• Context-adaptive genomic analysis: Adapts analyses to different genomic contexts by incorporating external variables like isochore structure and chromosome identity.

Methodology:

Uses a simulation-based framework with a null model that assumes independent placement of interval sets and permits interval density to depend on external variables (e.g., isochore structure, chromosome identity), with multiple testing controlled via FDR.