SGI

SGI identifies clinical subgroups in large-scale omics datasets by integrating hierarchical clustering, association testing, and multi-omics analysis to relate subgroup structure to clinical parameters and outcomes.

Key Features:

• Implementation (R): Implemented in R.
• Hierarchical clustering trees: Constructs hierarchical clustering trees to organize samples based on omics profiles.
• Association testing: Performs association testing to evaluate relationships between identified subgroups and clinical parameters or outcomes.
• Visualization framework: Provides a visualization framework for interpreting subgroup structures and their associations.
• Multi-block extension: Supports a multi-block extension to concurrently analyze multiple omics datasets from the same samples.
• Multi-omics integration: Integrates genomics, transcriptomics, proteomics, and metabolomics data for subgroup discovery.
• Clinical parameter handling: Systematically processes and analyzes an arbitrary number of clinical parameters and outcomes.

Scientific Applications:

• Clinical subgroup discovery: Identifies clinically relevant subgroups within large-scale omics datasets.
• Multi-omics studies: Integrates multiple omics types (genomics, transcriptomics, proteomics, metabolomics) to study disease heterogeneity.
• Type 2 diabetes metabolomics: Applied to a type 2 diabetes metabolomics study.
• TCGA copy number variation: Applied to two copy number variation datasets from The Cancer Genome Atlas (TCGA).

Methodology:

Constructs hierarchical clustering trees, performs association testing between identified subgroups and clinical parameters or outcomes, employs a visualization framework, and supports a multi-block extension for concurrent analysis of multiple omics datasets; implemented in R.