bootGSEA

bootGSEA performs bootstrap resampling and rank aggregation to study the robustness of Gene Set Enrichment Analysis (GSEA) and to combine results from multiple studies or omics levels.

Key Features:

• Bootstrap GSEA: Repeats GSEA based on bootstrap samples to study the variability and robustness of results, where not all genes or proteins are involved in each bootstrap replicate.
• Rank Aggregation: Aggregates ranks from bootstrap replicates to obtain a score per gene set showing whether it gains or loses evidence compared to standard GSEA ranking.
• Multi-omics Integration: Combines GSEA results from different omics levels or from multiple independent studies at the same omics level using rank aggregation.
• Robust Gene Set Selection: Identifies gene or protein sets that are less robust when set composition changes during bootstrap analysis.

Scientific Applications:

• Cancer Transcriptomics: Applied to six independent cancer transcriptomics datasets to aid in the selection of more robust enriched gene sets.
• Spinal Muscular Atrophy Multi-omics: Applied to paired transcriptomics and proteomics data from a mouse model of spinal muscular atrophy to aggregate findings across both omics levels after obtaining robust rankings at each level.

Methodology:

The pipeline repeats GSEA based on bootstrap samples, then aggregates the ranks from bootstrap replicates to obtain a score per gene set. Rank aggregation is used to combine GSEA results from different omics levels or from multiple independent studies at the same omics level.