saps

saps identifies gene sets associated with patient survival by combining prognostic tests and a significance framework to evaluate biological and statistical prognostic relevance in cancer genomics.

Key Features:

• Integration of prognostic tests: Combines standard prognostic tests with a significance test that stratifies patients into prognostic subtypes using random gene sets to ensure gene sets both stratify outcomes and show univariate associations with prognosis.
• Robust statistical analysis: Computes P_pure, P_random, and a pre-ranked Gene Set Enrichment Analysis (GSEA) to assess the biological significance of gene sets relative to random gene sets.
• Summary score and Q-value computation: Aggregates the three statistical measures into a summary score and optionally computes a Q-value to estimate significance.

Scientific Applications:

• Meta-analysis in cancer research: Applied in large-scale meta-analyses of prognostic pathways in breast and ovarian cancer, including the largest completed meta-analysis to date.
• Identification of prognostic signatures: Identifies prognostic gene signatures and molecular subtype-specific signatures in breast and ovarian cancers.
• Cross-cancer signature comparison: Revealed that prognostic signatures in ER-negative breast cancer are more similar to those in ovarian cancer than to ER-positive breast cancer signatures.
• Validation of biological importance: Ensures significant gene sets are enriched for genes with strong associations to patient prognosis and outperform random gene sets, supporting biological and clinical relevance.

Methodology:

Combines standard prognostic tests with a random-gene-set-based significance test that stratifies patients into prognostic subtypes; computes P_pure, P_random, and a pre-ranked GSEA; and calculates a summary score with an optional Q-value.