SISPA

SISPA identifies and defines sample groups that share gene set enrichment profiles by integrating multiple genomic data types such as gene expression and DNA methylation to link molecular profiles to phenotypes and clinical outcomes.

Key Features:

• Integration of Multiple Genomic Data Types: Integrates datasets including gene expression and DNA methylation and avoids limitations of 'integrate by intersection' (IBI) approaches that shrink intersections as data types increase.
• Profile-Based Analysis: Defines user-specified molecular profiles or gene sets and compares their activity across multiple classes to identify concordant genes and samples.
• Ranked Gene Sets Identification: Produces ranked lists of gene sets that satisfy user-defined molecular profiles, highlighting candidate driving gene sets for each class.
• Sample Group Definition: Stratifies samples into groups with similar profile activity to associate genomic characteristics with clinical or phenotypic outcomes.

Scientific Applications:

• Oncology — Multiple Myeloma (MM): Applied to human MM cell lines to recover genes with known profiles and identify novel candidate targets.
• Clinical Cohort Stratification — coMMpass trial: Identified sample groups with significant profile activity that associate with patient outcomes in the coMMpass trial data.

Methodology:

The methodology comprises Gene Integrated Set Profile Analysis (GISPA), which defines a molecular profile and compares it across multiple classes to identify gene sets satisfying the profile as potential drivers, and Sample Integrated Set Profile Analysis (SISPA), which defines sample groups based on activity within a given gene set aligned to a predefined molecular profile.