SourceSet

SourceSet distinguishes primary from secondary genes driving perturbations in biological pathways by modeling omic data within a Gaussian graphical model framework.

Key Features:

• Distinguishing Primary from Secondary Dysregulation: Identifies genes that are sources of perturbation (primary dysregulation) versus genes affected by network propagation (secondary dysregulation).
• Comparative Analysis: Compares control and perturbed conditions by testing differences in mean and covariance parameters using likelihood ratio tests.
• Inference of Gene Sets: Infers a primary dysregulation set and a secondary set from evidence produced by the comparative analyses.
• High Specificity and Sensitivity: Reports high specificity and sensitivity across simulated scenarios and real biological case studies.
• Pathway Analysis Extension: Extends analysis to multiple pathways to enable joint examination of complex pathway structures.
• Graphical Outputs: Produces Cytoscape-compatible visualizations for browsing inferred gene interactions and pathway perturbations.

Scientific Applications:

• Genomics: Identifying putative driver genes in genomic perturbation studies.
• Transcriptomics: Interpreting differential gene expression perturbations by separating primary drivers from secondary responses.
• Systems Biology: Dissecting network propagation effects within pathway models.
• Biomarker and Therapeutic Target Discovery: Prioritizing primary dysregulated genes as candidate biomarkers or therapeutic targets.

Methodology:

Uses Gaussian graphical models to compare mean and covariance parameters between control and perturbed conditions via likelihood ratio tests to infer primary and secondary dysregulation gene sets and extends this approach to multiple pathways.