Hypergraph Dynamic Correlation

Hypergraph Dynamic Correlation constructs module-level three-way gene interaction networks using integrative uniform hypergraphs to capture global dynamic correlation patterns for analysis of dynamic regulatory systems.

Key Features:

• Dynamic Correlation Analysis: Captures changing correlations between functionally related genes across biological conditions by modeling global dynamic correlation patterns.
• Three-Way Gene Interactions: Quantifies ternary relationships among genes using the Liquid Association statistic, focusing on pairs whose correlation changes and a third gene indicative of cellular conditions.
• Module-Level Network Construction: Constructs integrative uniform hypergraphs at the module level to represent three-way interactions and system-wide dynamic correlation structure.
• False Discovery Control: Formally controls false discoveries when testing large numbers of gene triplet combinations in high-throughput datasets.

Scientific Applications:

• Omics Data Analysis: Extracts higher-order structures from omics datasets to reveal regulatory processes not apparent from pairwise analyses.
• Regulatory Mechanism Discovery: Identifies novel higher-order regulatory mechanisms by analyzing dynamic gene triplet relationships.
• Disease and Systems Biology Studies: Applies to genomics and systems biology investigations, including analyses of TCGA melanoma RNA-seq and yeast cell cycle datasets.

Methodology:

Constructs integrative uniform hypergraphs that reflect global dynamic correlation patterns and quantifies ternary relationships using the Liquid Association statistic; validates biological plausibility via real-data experiments on datasets such as TCGA melanoma RNA-seq and yeast cell cycle data.