DiffCorr

DiffCorr analyzes differential correlations in biological networks to identify changes in molecular correlation relationships between two experimental conditions, particularly in post-genomics data such as transcriptomics and metabolomics.

Key Features:

• Differential Correlation Analysis: Identifies pattern changes in correlation networks between two experimental conditions using association measures such as Pearson's correlation coefficient.
• Efficient and Unbiased: Provides an efficient, unbiased approach suitable for large-scale omics data analysis.
• Eigen-Molecules Identification: Calculates correlation matrices for each dataset and extracts first principal component-based "eigen-molecules" from correlation networks.
• Statistical Testing: Applies Fisher's z-test to assess differences in correlations between two groups.

Scientific Applications:

• Biological Network Analysis: Infers cellular regulatory network changes from omics data to aid understanding of the molecular basis of diseases and traits.
• Biomarker Detection: Serves as a preliminary method to identify candidate biomarkers and potential causal relationships.
• Transcriptomics and Metabolomics: Highlights biologically relevant, differentially correlated molecules in transcriptome coexpression and metabolite-to-metabolite correlation networks.

Methodology:

Computes Pearson correlation matrices per dataset, derives first principal components as eigen-molecules, and applies Fisher's z-test to evaluate differential correlations between two groups.